Breaching the blood-brain barrier: AAV triggers dose-dependent toxicity in the brain.

Since in vivo transduction of neurons and glia by adeno-associated virus (AAV) vectors was first demonstrated in rat brain,1Kaplitt M.G. Leone P. Samulski R.J. Xiao X. Pfaff D.W. O'Malley K.L. During M.J. Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain.Nat. Genet. 1994; 8: 148-154Google Scholar it has been shown that AAV can efficiently transduce neurons, microglia, astrocytes, and oligodendrocytes across multiple species, and AAV vectors have become a widely used tool in basic and therapeutic research targeting the nervous system.2Peters C.W. Maguire C.A. Hanlon K.S. Delivering AAV to the Central Nervous and Sensory Systems.Trends Pharmacol. Sci. 2021; 42: 461-474Google Scholar However, recent observations have shown that the administration of AAV vectors at high doses can cause local or widespread neurotoxicity in the central (CNS) and peripheral (PNS) nervous systems, and the long-term consequences of this for human therapies targeting nervous tissue remain to be determined. In this issue of Molecular Therapy Methods and Clinical Development, Guo and colleagues investigate the mechanisms by which AAV mediates neurotoxicity following intraparenchymal injection into mouse brain.3Guo Y. Chen J. Ji W. Xu L. Xie Y. He S. Lai C. Hou K. Li Z. Chen G. Wu Z. High-titer AAV Disrupts Cerebrovascular Integrity and Induces Lymphocyte Infiltration in Adult Mouse Brain.Mol. Ther. Methods Clin. Dev. 2023; 31https://doi.org/10.1016/j.omtm.2023.08.021Google Scholar Their observation that high doses of AAV can cause localized disruption of the blood-brain barrier (BBB), which allows an influx of blood and serum factors into brain parenchyma (Figure 1), provides important mechanistic insights into our understanding of how neurotoxicity at the site of injection is mediated by infiltrating lymphocytes. Although endemic AAV infections have not been reported in nervous tissue, vectorized AAV shows broad neurotropism across many species. Generally, AAV is considered to be safe, and hundreds of studies have used AAV vectors to target the nervous system with few reporting neurotoxicity. Nevertheless, reports of neurotoxicity following delivery of high AAV doses have raised concerns about the long-term impact AAV therapies may have on neurological functions.4Stone D. Aubert M. Jerome K.R. Adeno-associated virus vectors and neurotoxicity-lessons from preclinical and human studies.Gene Ther. 2023; Google Scholar The mechanisms by which AAV triggers toxicity and/or cell death in nervous tissues are poorly understood, although immune-mediated mechanisms and apoptosis following activation of the unfolded protein response have been implicated in the death of AAV transduced neurons within the dorsal root ganglion (DRG).5Hordeaux J. Buza E.L. Jeffrey B. Song C. Jahan T. Yuan Y. Zhu Y. Bell P. Li M. Chichester J.A. et al.MicroRNA-mediated inhibition of transgene expression reduces dorsal root ganglion toxicity by AAV vectors in primates.Sci. Transl. Med. 2020; 12eaba9188Google Scholar In the brain, little is known about vector-mediated neurotoxicity, even though it has been observed in brains of mice, rats, and non-human primates (NHPs) (Table 1) and, more recently, in a clinical trial for late infantile Battens disease.6Sondhi D. Kaminsky S.M. Hackett N.R. Pagovich O.E. Rosenberg J.B. De B.P. Chen A. Van de Graaf B. Mezey J.G. Mammen G.W. et al.Slowing late infantile Batten disease by direct brain parenchymal administration of a rh.10 adeno-associated virus expressing CLN2.Sci. Transl. Med. 2020; 12eabb5413Google Scholar Importantly, vector-mediated neurotoxicity across different species has been observed when similar AAV doses are administered relative to the recipient brain size (6.43 × 108–2.5 × 1010 vector genomes [vg] per gram of brain tissue, Table 1).Table 1Approximate AAV dose relative to brain weight for studies reporting localized neurotoxicity after intra-parenchymal delivery of AAV vectorsStudySpecies – strainAAV dose administered (vg/brain)Assumed brain weight (g)AAV dose (vg/g of brain)Guo et al.3Guo Y. Chen J. Ji W. Xu L. Xie Y. He S. Lai C. Hou K. Li Z. Chen G. Wu Z. High-titer AAV Disrupts Cerebrovascular Integrity and Induces Lymphocyte Infiltration in Adult Mouse Brain.Mol. Ther. Methods Clin. Dev. 2023; 31https://doi.org/10.1016/j.omtm.2023.08.021Google Scholarmice – C57BL/61 × 10100.432.33 × 1010Johnston et al.10Johnston S. Parylak S.L. Kim S. Mac N. Lim C. Gallina I. Bloyd C. Newberry A. Saavedra C.D. Novak O. et al.AAV ablates neurogenesis in the adult murine hippocampus.Elife. 2021; 10e59291Google Scholarmice – C57BL/63 × 1090.437 × 109Klein et al.11Klein R.L. Dayton R.D. Leidenheimer N.J. Jansen K. Golde T.E. Zweig R.M. Efficient neuronal gene transfer with AAV8 leads to neurotoxic levels of tau or green fluorescent proteins.Mol. Ther. 2006; 13: 517-527Google Scholarrats – Sprague Dawley5 × 101022.5 × 1010Rosenberg et al.12Rosenberg J.B. Chen A. De B.P. Dyke J.P. Ballon D.J. Monette S. Ricart Arbona R.J. Kaminsky S.M. Crystal R.G. Sondhi D. Safety of Direct Intraparenchymal AAVrh.10-Mediated Central Nervous System Gene Therapy for Metachromatic Leukodystrophy.Hum. Gene Ther. 2021; 32: 563-580Google ScholarAfrican green monkey1.5 × 101264.3/61.1, M/F2.33–2.45 × 109Samaranch et al.13Samaranch L. Sebastian W.S. Kells A.P. Salegio E.A. Heller G. Bringas J.R. Pivirotto P. DeArmond S. Forsayeth J. Bankiewicz K.S. AAV9-mediated expression of a non-self protein in nonhuman primate central nervous system triggers widespread neuroinflammation driven by antigen-presenting cell transduction.Mol. Ther. 2014; 22: 329-337Google ScholarMacaca fascicularis6 × 101168.4/60.4, M/F8.8–9.9 × 109Golebiowski et al.7Golebiowski D. van der Bom I.M.J. Kwon C.S. Miller A.D. Petrosky K. Bradbury A.M. Maitland S. Kühn A.L. Bishop N. Curran E. et al.Direct Intracranial Injection of AAVrh8 Encoding Monkey beta-N-Acetylhexosaminidase Causes Neurotoxicity in the Primate Brain.Hum. Gene Ther. 2017; 28: 510-522Google ScholarMacaca fascicularis3.2 × 101268.4/60.4, M/F4.7–5.3 × 1010Keiser et al.9Keiser M.S. Ranum P.T. Yrigollen C.M. Carrell E.M. Smith G.R. Muehlmatt A.L. Chen Y.H. Stein J.M. Wolf R.L. Radaelli E. et al.Toxicity after AAV delivery of RNAi expression constructs into nonhuman primate brain.Nat. Med. 2021; 27: 1982-1989Google ScholarMacaca mulatta>6 × 101186.1/96.1, M/F>6.24–6.97 × 109Sondhi et al.6Sondhi D. Kaminsky S.M. Hackett N.R. Pagovich O.E. Rosenberg J.B. De B.P. Chen A. Van de Graaf B. Mezey J.G. Mammen G.W. et al.Slowing late infantile Batten disease by direct brain parenchymal administration of a rh.10 adeno-associated virus expressing CLN2.Sci. Transl. Med. 2020; 12eabb5413Google Scholarhuman9 × 10111,000aApproximate brain weight of a 1- to 5-year-old child.6.43 × 108The lowest dose at which parenchymal neurotoxicity was seen is reported.a Approximate brain weight of a 1- to 5-year-old child. Open table in a new tab The lowest dose at which parenchymal neurotoxicity was seen is reported. In this manuscript, Guo et al. investigate the events that trigger localized neurotoxicity in mice at the site of injection after administration of two AAV doses (1 × 108 and 1 × 1010 vg/brain, equivalent to 2.33 × 108 and 2.33 × 1010 vg/gram of brain). They first show a dose threshold effect for activation of microglia and astrocytes at the higher dose around the injection site, and for activation of genes associated with anti-viral responses, recruitment of immune cells, BBB integrity, and inflammatory responses, via bulk RNAseq. They then show that localized disruption of vascular endothelial tight junctions, astrocytic perivascular endfeet, and pericyte-endothelial cell interactions occur at the higher dose between 10 and 30 days post injection (dpi), which permeabilizes the BBB and enables blood cells or serum proteins such as platelets and IgG to enter the parenchyma. BBB permeability at the site of AAV injection was also demonstrated after fluorescent cadaverine and dextran tracers were administered intravenously. Next, the authors investigated whether BBB disruption facilitates immune cell infiltration and acute-neuronal cytotoxicity at the site of high-dose AAV administration. CD4+, CD8+, and CD19+ lymphocytes infiltrated the injection site between 6 and 10 dpi, persisting until 30 dpi, and CD4+/GranzymeB+ T cells were found at the injection site at 20 and 30 dpi along with TUNEL+/NeuN+ neurons. After high-dose AAV administration, progressive loss of NeuN+ cells was observed at the injection site from 10 to 30 dpi. BBB disruption and immune cell influx was also seen when high AAV doses were delivered to the striatum and hippocampus, following AAV5 or AAV9 delivery, and when different GFAP promoters were utilized, but not when empty capsids were administered. Interestingly, the loss of NeuN+ cells was independent of transgene expression in neurons since expression was restricted to astrocytes/glia by the GFAP promoter. In fact, the authors showed that expression of the GFP transgene was not required for BBB disruption, immune cell infiltration, or loss of NeuN+ cells, which were all seen at the injection site in both wild-type and GFAP-Cre transgenic mice administered a Cre-lox AAV vector system that requires Cre expression to turn on transgene expression. Finally, the authors investigated the effect of immune cell depletion on responses to high AAV doses in brain parenchyma. Treatment with anti-CD4 or anti-CD8 depleting antibodies inhibited both the influx of CD4+ and CD8+ T cells and the depletion of NeuN+ cells at the site of AAV injection. Similarly, treatment with an inhibitory anti-GranzymeB antibody partially attenuated loss of NeuN+ cells at the site of injection. These data suggest that infiltrating lymphocytes play a significant role in vector-mediated brain neurotoxicity and provide a potential approach to prevent unwanted neurotoxicity in future studies targeting brain parenchyma. While these observations offer us a basic understanding of the biological processes that lead to neuronal death in the brain, much remains unknown about this phenomenon or its clinical significance. It is unclear how AAV triggers disruption of the BBB—is BBB disruption a direct or indirect consequence of AAV administration? And, why is BBB disruption not seen until after 6 dpi? The consequences of vector-mediated neurotoxicity on long-term neurological functions are also uncertain, both in the CNS and PNS. Abnormal neurological behaviors or deficits have been seen in some NHP with loss of neurons in the DRG or brain after systemic or parenchymal AAV administration respectively.7Golebiowski D. van der Bom I.M.J. Kwon C.S. Miller A.D. Petrosky K. Bradbury A.M. Maitland S. Kühn A.L. Bishop N. Curran E. et al.Direct Intracranial Injection of AAVrh8 Encoding Monkey beta-N-Acetylhexosaminidase Causes Neurotoxicity in the Primate Brain.Hum. Gene Ther. 2017; 28: 510-522Google Scholar,8Hordeaux J. Buza E.L. Dyer C. Goode T. Mitchell T.W. Richman L. Denton N. Hinderer C. Katz N. Schmid R. et al.Adeno-Associated Virus-Induced Dorsal Root Ganglion Pathology.Hum. Gene Ther. 2020; 31: 808-818Google Scholar,9Keiser M.S. Ranum P.T. Yrigollen C.M. Carrell E.M. Smith G.R. Muehlmatt A.L. Chen Y.H. Stein J.M. Wolf R.L. Radaelli E. et al.Toxicity after AAV delivery of RNAi expression constructs into nonhuman primate brain.Nat. Med. 2021; 27: 1982-1989Google Scholar However, signs of neurotoxicity in the PNS or CNS were not associated with abnormal neurological behaviors or deficits in most animals. It is worth noting that abnormal DRG pathology in NHPs is largely not detectable by 6 months post AAV administration,8Hordeaux J. Buza E.L. Dyer C. Goode T. Mitchell T.W. Richman L. Denton N. Hinderer C. Katz N. Schmid R. et al.Adeno-Associated Virus-Induced Dorsal Root Ganglion Pathology.Hum. Gene Ther. 2020; 31: 808-818Google Scholar whereas the long-term persistence of vector-associated neurotoxicity in NHP brain has not been studied in depth. The relevance of these observations to the clinical setting should also be determined, since lower AAV doses that are not neurotoxic may show therapeutic efficacy. Although limited, there is evidence that parenchymal neurotoxicity occurs in the human brain. In the trial for late infantile Battens disease, disease progression was slowed, but some trial participants had AAV injection site abnormalities detectable by MRI that persisted for up to 18 months, and adverse events including seizures, dystonia, and abnormal movements were seen in some participants.6Sondhi D. Kaminsky S.M. Hackett N.R. Pagovich O.E. Rosenberg J.B. De B.P. Chen A. Van de Graaf B. Mezey J.G. Mammen G.W. et al.Slowing late infantile Batten disease by direct brain parenchymal administration of a rh.10 adeno-associated virus expressing CLN2.Sci. Transl. Med. 2020; 12eabb5413Google Scholar Interestingly, trial participants received a slightly lower dose per brain weight than other species with reported vector-mediated parenchymal neurotoxicity (Table 1), demonstrating that the dose threshold for AAV-mediated BBB dysregulation may vary by species. This is highly relevant since calculations of the maximum permissible “safe” dose based on studies conducted in animals may not be accurate. Ultimately, the clinical significance of localized BBB disruption at high AAV doses should be studied further in animals and humans across a broader range of doses, and it will be important to longitudinally monitor neurological behaviors of patients receiving intra-cranial AAV administrations as therapy. Research in our laboratory is funded by NIH grants R21AI117519, R01AI132599, UM1 AI126623, amfAR, the Caladan Foundation, over 1600 individual donors, and, in part, by a developmental grant from the University of Washington Center for AIDS Research (CFAR), an NIH-funded program under award number P30 AI 027757, which is supported by the following NIH Institutes and Centers (NIAID, NCI, NIMH, NIDA, NICHD, NHLBI, NIA, NIGMS, and NIDDK), and in part by NIH/NCI Cancer Center Support Grant P30 CA015704. D.S. and M.A. have sponsored research agreements with Excision Biosciences. K.R.J. holds equity in Excision Biosciences and Caladan Therapeutics.