Delivering efficient liver-directed AAV-mediated gene therapy

Adeno-associated virus vectors (AAV) have become the leading technology for liver-directed gene therapy.1 After the pioneering trials using AAV2(ref. 2) and AAV8(ref. 3) to treat haemophilia B, D’Avola et al.4 recently reported the first-in-human clinical trial of adeno-associated virus vector serotype 5 (AAV5) in acute intermittent porphyria (AIP). Treatment was reported as safe, but the main surrogate biomarkers of AIP, porphobilinogen (PBG) and delta-aminolevulinate (ALA) were unchanged. This lack of efficacy contrasts with results from the haemophilia B trial using AAV8 capsid by Nathwani et al.,3 which showed a significant and long-lasting improvement of the clinical phenotype. Haemophilia B is an amenable target for successful gene therapy as raising expression of plasma factor IX (FIX) level above 1% can modify the phenotype from severe to moderate.3 Development of a variety of capsids for clinical application is useful to overcome pre-existing neutralising antibodies. The differences in cell-specific transduction by different AAV serotypes are primarily owing to specificities in cellular uptake or post cell-entry processing. Indeed AAV5 presents several theoretical advantages as an alternative capsid to AAV8 for liver-directed gene therapy: suitable liver tropism, less off-target biodistribution,5 low seroprevalence in humans and minimal cross-reactivity with other serotypes.6

Reliability of animal models in capsid testing

The reliability of the available animal models for comparison of transduction of the liver by different AAV serotypes has been questioned.7 In the AIP trial,4 the high-dose group received 1.8 × 1013 vg kg−1, which is equivalent to the therapeutic threshold needed to achieve a correction of the murine phenotype (1.25 × 1013 vg kg−1),8 but lower than that required for supra-physiological enzymatic activity in Rhesus macaques (5 × 1013 vg kg−1).5 AAV5 is currently used in a clinical trial for haemophilia B with the same transgene cassette used by Nathwani et al. ( Nine months post infusion, the low-dose group, who received 5 × 1012 vg kg−1, showed a plasma FIX of 5.4% (range 3.1–6.7%; n=5) ( In), which is similar to the level observed in the high-dose group of the AAV8 trial receiving 2 × 1012 vg kg−1 (plasma FIX of 5.1%, range 2.9–7.1%; n=6) 4 months post infusion.3 These results suggest that, to obtain similar plasma FIX levels to those achieved in AAV8 trial, administration of 2.5-fold more AAV5 vector is necessary.

Although this assumption is made on the basis of a small number of treated subjects, and confounded by different methods of production, titration and purification, it supports data obtained after intravenous injection in different animal models:

  1. i

    In murine models of AIP, AAV5 resulted in 10-fold less liver transduction compared with AAV8.8

  2. ii

    In Gunn rats, AAV5 vector was inefficient at restoring metabolic activity and achieved three times lower copy number compared with AAV8.6

  3. iii

    In Rhesus macaques, AAV5 vector produced slightly lower plasma FIX in adult animals with slower kinetics compared with AAV8,9 lower hepatocyte transduction after fetal intrahepatic venous injection and lower plasma FIX 2 months post injection (<1 μg ml−1 (n=3) versus 5 μg ml−1 (n=1)).10

  4. iv

    In Fah−/−/Rag2−/−/Il2rg−/− (FRG) mice, AAV5 achieved transduction of 10-fold fewer of human hepatocytes than AAV8 (0.1 versus 1.1%, respectively).11

Further results from larger human trials will provide further information on the reliability of animal data, which will accelerate the development of liver-directed gene therapy.

Episomal versus endogenous gene expression

D’Avola et al.4 are the first to report data from human liver biopsies after AAV treatment. Interestingly, the liver vector copy number 1 year post injection did not correlate with the escalating doses of vector received. This finding is in contrast with the studied tissues from animal models5, 8 or plasma FIX levels in haemophilia B trial.3 In liver biopsies with high vector copy number of the transgene codon-optimised PBG deaminase (coPBGD) (patients 2, 5 and 7), coPBGD mRNA expression compared with endogenous PBGD (normalised by DNA copy number) was lower by 45%, 76% and 36%, respectively.4 In AAV-mediated gene therapy, most of the transgene DNA copies persist as non-integrated episomes. Different episomal expression compared with the endogenous gene of interest underpins results observed in an ornithine transcarbamylase12-deficient Spfash mouse model. Untreated Spfash mice with a 5–7% wild-type residual ornithine transcarbamylase activity become hyperammonaemic after a short hairpin RNA (shRNA)-mediated knockdown of the endogenous ornithine transcarbamylase activity to 0–2.5%. In shRNA-injected Spfash mice, the level of AAV-encoded ornithine transcarbamylase activity required to normalise ammonaemia was threefold higher than the residual ornithine transcarbamylase activity in untreated Spfash mice.13 An AAV pattern of transduction not reproducing the physiological metabolic zonation of the liver might have had an additional role. Although these findings rely on a small cohort and require caution in interpretation, various explanations might account for a different episomal expression such as inadequate chromatinisation, incomplete circularisation of the AAV genome altering the constitution of the open reading frame for transgene expression or inverted terminal repeats recombination. The exact mechanism for this phenomenon is yet to be identified.

Functional metabolic assays as efficacy end points in clinical trials

Finally, the use of metabolite levels as primary end point for trials in metabolic diseases can be questioned. These surrogate markers often reflect a static picture, and remain indirect assessments of the metabolic flux and its environmental or epigenetic regulation. Indeed, haem biosynthesis is mainly regulated by haem-mediated inhibitory feedback of the transcription of ALA-synthetase, but other parameters can exert an influence such as glucose intake, stress, drugs, circadian rhythm,14 and may potentially affect ALA and PBG results. Thus, whenever feasible, stable isotope studies would be better indicators of the in vivo dynamics of the pathway. For example, oral administration of N15-labelled glycine can monitor the biosynthesis of haem and its intermediate compounds in physiology and patients with inherited porphyrias.15 This approach has been successfully used in other metabolic pathways like the urea cycle to assess ureagenesis utilising either N15-labelled urea in animal models after AAV-mediated gene therapy16, 17 or C13-labelled acetate in humans for accurately stratifying the disease severity in ornithine transcarbamylase deficiency.18 Furthermore, the use of clinically relevant end points would not only provide better assessment of the effect of therapy, but may be viewed more favourably by regulatory bodies.


  1. 1

    Dolgin E . Early clinical data raise the bar for hemophilia gene therapies. Nat Biotechnol 2016; 34: 999–1001.

    CAS  Article  Google Scholar 

  2. 2

    Manno CS, Pierce GF, Arruda VR, Glader B, Ragni M, Rasko JJ et al. Successful transduction of liver in hemophilia by AAV-Factor IX and limitations imposed by the host immune response. Nat Med 2006; 12: 342–347.

    CAS  Article  Google Scholar 

  3. 3

    Nathwani AC, Reiss UM, Tuddenham EG, Rosales C, Chowdary P, McIntosh J et al. Long-term safety and efficacy of factor IX gene therapy in hemophilia B. N Engl J Med 2014; 371: 1994–2004.

    Article  Google Scholar 

  4. 4

    D'Avola D, Lopez-Franco E, Sangro B, Paneda A, Grossios N, Gil-Farina I et al. Phase I open label liver-directed gene therapy clinical trial for acute intermittent porphyria. J Hepatol 2016; 65: 776–783.

    CAS  Article  Google Scholar 

  5. 5

    Paneda A, Lopez-Franco E, Kaeppel C, Unzu C, Gil-Royo AG, D'Avola D et al. Safety and liver transduction efficacy of rAAV5-cohPBGD in nonhuman primates: a potential therapy for acute intermittent porphyria. Hum Gene Ther 2013; 24: 1007–1017.

    CAS  Article  Google Scholar 

  6. 6

    Montenegro-Miranda PS, Paneda A, ten Bloemendaal L, Duijst S, de Waart DR, Gonzalez- Aseguinolaza G et al. Adeno-associated viral vector serotype 5 poorly transduces liver in rat models. PLoS One 2013; 8: e82597.

    Article  Google Scholar 

  7. 7

    Wang L, Bell P, Somanathan S, Wang Q, He Z, Yu H et al. Comparative study of liver gene transfer with AAV vectors based on natural and engineered AAV capsids. Mol Ther 2015; 23: 1877–1887.

    CAS  Article  Google Scholar 

  8. 8

    Unzu C, Sampedro A, Mauleon I, Alegre M, Beattie SG, de Salamanca RE et al. Sustained enzymatic correction by rAAV-mediated liver gene therapy protects against induced motor neuropathy in acute porphyria mice. Mol Ther 2011; 19: 243–250.

    CAS  Article  Google Scholar 

  9. 9

    Nathwani AC, Gray JT, McIntosh J, Ng CY, Zhou J, Spence Y et al. Safe and efficient transduction of the liver after peripheral vein infusion of self-complementary AAV vector results in stable therapeutic expression of human FIX in nonhuman primates. Blood 2007; 109: 1414–1421.

    CAS  Article  Google Scholar 

  10. 10

    Mattar CN, Nathwani AC, Waddington SN, Dighe N, Kaeppel C, Nowrouzi A et al. Stable human FIX expression after 0.9G intrauterine gene transfer of self-complementary adeno-associated viral vector 5 and 8 in macaques. Mol Ther 2011; 19: 1950–1960.

    CAS  Article  Google Scholar 

  11. 11

    Vercauteren K, Hoffman BE, Zolotukhin I, Keeler GD, Xiao JW, Basner-Tschakarjan E et al. Superior in vivo transduction of human hepatocytes using engineered AAV3 capsid. Mol Ther 2016; 24: 1042–1049.

    CAS  Article  Google Scholar 

  12. 12

    Inagaki K, Piao C, Kotchey NM, Wu X, Nakai H . Frequency and spectrum of genomic integration of recombinant adeno-associated virus serotype 8 vector in neonatal mouse liver. J Virol 2008; 82: 9513–9524.

    CAS  Article  Google Scholar 

  13. 13

    Cunningham SC, Kok CY, Dane AP, Carpenter K, Kizana E, Kuchel PW et al. Induction and prevention of severe hyperammonemia in the spfash mouse model of ornithine transcarbamylase deficiency using shRNA and rAAV-mediated gene delivery. Mol Ther 2011; 19: 854–859.

    CAS  Article  Google Scholar 

  14. 14

    Besur S, Hou W, Schmeltzer P, Bonkovsky HL . Clinically important features of porphyrin and heme metabolism and the porphyrias. Metabolites 2014; 4: 977–1006.

    Article  Google Scholar 

  15. 15

    London IM . The use of stable isotopes in biological and medical research. J Clin Invest 1949; 28 (6 Pt 1): 1255–1270.

    CAS  Article  Google Scholar 

  16. 16

    Cunningham SC, Spinoulas A, Carpenter KH, Wilcken B, Kuchel PW, Alexander IE . AAV2/8-mediated correction of OTC deficiency is robust in adult but not neonatal Spf(ash) mice. Mol Ther 2009; 17: 1340–1346.

    CAS  Article  Google Scholar 

  17. 17

    Hu C, Tai DS, Park H, Cantero G, Chan E, Yudkoff M et al. Minimal ureagenesis is necessary for survival in the murine model of hyperargininemia treated by AAV-based gene therapy. Gene Ther 2015; 22: 111–115.

    CAS  Article  Google Scholar 

  18. 18

    Opladen T, Lindner M, Das AM, Marquardt T, Khan A, Emre SH et al. In vivo monitoring of urea cycle activity with (13)C-acetate as a tracer of ureagenesis. Mol Gene Metab 2016; 117: 19–26.

    CAS  Article  Google Scholar 

Download references


JB is supported by a Clinical Starter Research Grant from Great Ormond Street Hospital for Children Charity.

Author contributions

JB wrote the manuscript. SNW, IEA and PG contributed and revised the manuscript. All authors read and approved the final manuscript.

Author information



Corresponding author

Correspondence to J Baruteau.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Baruteau, J., Waddington, S., Alexander, I. et al. Delivering efficient liver-directed AAV-mediated gene therapy. Gene Ther 24, 263–264 (2017).

Download citation

Further reading


Quick links