Gene Therapy (2010) 17, 448–458 & 2010 Macmillan Publishers Limited All rights reserved 0969-7128/10 $32.00 www.nature.com/gt REVIEW Progress and prospects: Immunobiology of therapy for neurodegenerative disease: prospects and risks

MM McMenamin and MJA Wood Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK

Gene therapy for neurological, and in particular neurode- advances in understanding immune regulation in the central generative, disease is now a reality. A number of early phase nervous system (CNS) and with the continued development of clinical trials have been completed and several are currently new gene delivery vectors, this review critically assesses the in progress. In view of this, it is critically important to evaluate current knowledge of immunobiology within the CNS in terms the immunological risk associated with neurological gene of likely immunological risk pertaining to viral vectors and gene therapy, which has clear implications for trial safety and therapy applications for neurodegenerative disease. efficacy. Moreover, it is imperative in particular to identify Gene Therapy (2010) 17, 448–458; doi:10.1038/gt.2010.2; factors indicating potential high risk. In the light of recent published online 11 February 2010

In brief Progress Prospects

 Current understanding of the immune responses in  Outcome of Phase I/II clinical trials should inform the CNS directions for future research  Advances in the understanding of the innate immune  Better understanding of basic CNS immunobiology system and of RNA gene silencing pathways in the will shed further light on the nature of innate and CNS-implications for gene therapy vectors adaptive immune responses  Many neurodegenerative diseases and aging are  Identification of high-risk situations should inform typically associated with ongoing local innate strategies to limit the immune response immune activity that may increase immunological risk  CNS immune response to gene therapy vectors depends on vector type, vector dose and transgene  Previous exposure to the parent , the vector or reintroduction of the viral vector may suppress transgene expression and increase immunological risk  Novel viral vectors further reduce immunological risk

Keywords: viral vectors; neurodegenerative disease; immune response; CNS; re-administration

Introduction the induction of both innate and adaptive immune responses in the CNS, with consequent implications for Although immune responses to gene therapy vectors in gene therapy clinical trials, including for viral vector the central nervous system (CNS) are in general safety, transduction efficiency and transgene expression. subdued, the parameters that determine such responses Thus, immunological factors are critical determinants of are similar to those governing immune responses outside gene therapy clinical trial safety and also, ultimately, for the CNS. Importantly, results from preclinical studies in trial efficacy. There are further CNS-specific considera- animals and recent data from clinical trials have shown tions. A viable gene therapy targeting a chronic neuro- degenerative disease such as Parkinson’s disease (PD) is very likely to require vector re-administration in the long Correspondence: Dr MJA Wood, Department of Physiology, term. Moreover, such neurodegenerative diseases are Anatomy and Genetics, University of Oxford, South Parks Road, typically associated with ongoing local inflammation, Oxford OX1 3QX, UK. E-mail: [email protected] and to what extent such preexisting CNS inflammation Received 27 September 2009; revised 23 December 2009; accepted 27 increases the immunological risk for gene therapy must December 2009; published online 11 February 2010 be critically evaluated. To date, the completed and Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 449 ongoing trials have involved small numbers of patients chemokines ( inflammatory -1a, treated over short periods relative to disease duration. monocyte chemoattractant protein-1 and - Thus, the identification of specific circumstances and inducible protein) that promote the inflammatory state. patient subgroups at potentially higher immunological Similar to , microglial cells express Toll-like risk will require better understanding of disease biology receptors (TLRs) and respond to TLR ligands that initiate over the long term, and its impact on the likely success of production of pro-inflammatory mediators. In fact the a gene therapy intervention. Such evaluations of clinical discovery of pathogen recognition receptors such as risk must also be made in the light of greater knowledge the TLRs in the last decade heralded a revolution in the of immunobiology within the CNS. Recent advances in understanding of the and has the understanding of pathogen recognition have shown shed new light on how inflammatory responses are that viral vectors interact with innate immune sensing initiated and regulated within the CNS.7 Perhaps systems, and further that all neuronal and glial cell types surprisingly, all CNS cell types, including neurons, within the CNS express such pathogen recognition express these receptors and their ligation is important receptors. This has implications for the mechanism and in regulating immune privilege as it enables a rapid scope of both innate and adaptive immune responses response to infection, and provides a critical link to be anticipated in neurological gene therapy clinical between innate and adaptive immunity.8–12 After activa- trials. In addition, emerging evidence also suggests that tion of these receptors, the release of pro-inflammatory microRNA (miRNA)-mediated gene regulation repre- cytokines and chemokines activates the adaptive im- sents a fundamental layer of posttranscriptional regula- mune response.13 tion that has diverse functional tasks in viral immune The adaptive immune response involves interaction of regulation and likely in regulation of host immune T-cell receptors with antigen in the context of MHC responses to viral vectors. This review critically assesses molecule–peptide complexes in the nanometer scale gap and identifies key components of likely immunological between T cells and APCs, referred to as an immuno- risk pertaining to viral vector and gene therapy applica- logical synapse. T-cell activation also requires a costi- tions for neurological disease. mulatory signal involving interaction of CD28 on the T-cell with CD80 or CD86 (B7 family ) on the APC. The existence of mature-type immunological synapses in Current understanding of the immune the brain, indicative of T-cell responses in vivo, was responses in the CNS first described by Lowenstein and co-workers and shown to occur during an immune-mediated elimination The CNS has long been considered an immunologically of adenovirally infected cells. Anti-adenoviral CD8 privileged site. This is due to (1) the presence of the T cells infiltrated the brain and formed mature immuno- blood–brain barrier (BBB);1–3 (2) the absence of resident logical synapses with MHC class I expressing astrocytes. lymphocytes and lack of conventional lymphatics; (3) The cell–cell communication at such synapses is situa- low levels of major histocompatibility complex (MHC) tion-adapted and one of the main physiological functions class I and II molecules and a low preponderance of of immunological synapses is to direct cytokine secretion antigen-presenting cells (APCs), such as dendritic cells either directly into the synapse or in a multidirec- and macrophages4,5 and (4) the existence of local tional manner outside the synapse. In the brain immunosuppressive factors and the expression of Fas immunological synapses implement cytokine secretion ligand that induces apoptosis of activated T cells. directly into the synapse and 50% of infected astrocytes Although the brain parenchyma normally lacks dendritic are eliminated after delivery of an adenoviral vector to cells, and impairs priming of the adaptive immune the brain.14,15 response, as first shown by Lowenstein and co-workers, The challenge in the CNS is to understand a response dendritic cells do appear in the brain parenchyma on that must balance defense against invading pathogens inflammation. It is not yet clear whether these dendritic with inflicting CNS injury. Thus, the concept of immune cells develop from microglia or another brain-resident privilege has been refined and the CNS can now be more precursor, or are recruited from the blood. It is also accurately described as an immune-attenuated site, as known that in the absence of other competent APCs, B the machinery to initiate an immune response is present,7 cells and astrocytes are able to function as APCs and it is and in certain situations, robust immune responses can accepted that they have an important function in viral and do occur in response to pathogens and in auto- immunity in the brain.6 Astrocytes, in an inflammatory immune disease.16–18 environment, have the capacity to express the required MHC class II and B7 costimulatory molecules for efficient activation of naive T cells. B cells constitutively Advances in the understanding of the express MHC class II and B7 costimulatory molecules and B cells can also convert dendritic cells into antigen- innate immune system and of RNA gene specific APCs by promoting opsonization. silencing pathways in the CNS- Microglia are the main innate immune cells of the implications for gene therapy vectors brain and are the principle mediators of inflammation. In the presence of an activating stimulus, microglial cell- The innate immune response receptors are now known surface receptor expression is modified and the cells to be widely expressed in the CNS and constitute the first change from a monitoring function to one of protection line of defense against .10,19 Envelope glyco- and repair. Activated microglia secrete a variety of infla- of HSV activate cytokine responses through mmatory mediators, including cytokines (tumor TLR2. TLR9 recognizes unmethylated CpG DNA motifs necrosis factor, and interleukin (IL)-1b and IL-6) and present in the adenovirus genome and responds to

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 450 ligation by increasing pro-inflammatory cytokine secre- Neurodegenerative diseases and aging tion.20–22 Single-stranded RNA derived from activates TLR7 or TLR8 and leads to production of are typically associated with ongoing interferon-a and other cytokines in dendritic cells.23 local innate immune activity that may Although TLRs constitute the main sensors for increase immunological risk detection of viruses two other distinct families, (1) the retinoic acid-inducible gene-like helicases and (2) the Increasing evidence indicates that chronic neuroinflam- -binding oligomerization domain-like recep- mation has a critical function in the pathogenesis of most tors, act as intracellular surveillance receptors and have neurodegenerative diseases. PD, Alzheimer’s disease, been implicated in viral dsRNA recognition.24,25 Nucleo- multiple sclerosis, Huntington’s disease, amyotrophic tide-binding oligomerization domain-like receptors lateral sclerosis, tauopathies and age-related macular de- recruit the inflammatory caspase-1 into innate cytosolic generation are all associated with chronic neuroinflam- molecular complexes termed inflammasomes. Once mation and elevated levels of several cytokines.12,31–40 In activated, caspase-1 processes pro-IL-1b and pro-IL-18 addition, there are several recent reports that show to their active and secreted forms. The inflammasome normal aging is accompanied by progressive alterations has been shown to have a central function in the innate in immune function.41 immune response to adenovirus; internalized adenoviral Increased innate immune activity, microglial and DNA induces maturation of pro-IL-1b in macrophages, immune cell activation and increased pro-inflammatory which is dependent on components of the inflamma- cytokine expression/release are found in brains of some.23 The apolipoprotein B mRNA editing , patients with PD and inflammation is thought to be a catalytic polypeptide-like (APOBEC) family of cytidine major contributor to the ongoing neurodegeneration deaminases is also expressed in all the major cellular and disease progression.42,43 Patients with Alzheimer’s constituents of the CNS and inhibits virus replication. disease and a mouse model of Alzheimer’s have been APOBEC-3G/3F inhibits replication in endothelial cells shown to have increased BBB permeability.44,45 Patients of HIV-1, equine infectious anemia virus and murine have also been shown to have TLR induction,46,47 leukemia virus.26 An APOBEC3A protein hA3A targets activated microglia48 and upregulated expression of adeno-associated virus (AAV) by inhibiting its replica- cytokines, chemokines and chemokine receptors.49 In tion (for summary see Figure 1a). diseased individuals with neuroinfections, such has HIV, Emerging evidence also suggests that RNA gene the BBB has been shown to be more permeable. More silencing pathways are important in regulating host– generally, age-related immunosenescence includes in- virus interactions.27–29 Host cellular miRNAs modulate creased BBB permeability50,51 and increased microglial viral gene expression and viral gene silencing mechan- and astrocytic activity.52 isms enable host immune evasion and can exploit the Neuropathological and neuroradiological studies indi- cellular miRNA pathway to their own advantage.30 Work cate that neuroinflammatory responses may begin before on wild-type viruses has shown that viral miRNAs significant loss of neuronal populations in the progression inhibit aspects of the host-adaptive immune response, of these diseases. Although there is no evidence to sup- including antigen presentation, the innate immune port a function for any particular cytokine in the direct response including induction of apoptosis or the inter- triggering of any of these neurodegenerative conditions, feron system.28 Because most of the modified viruses cytokine-driven neuroinflammation and neurotoxicity has used for gene therapy have genes encoding miRNAs been shown to modify disease progression in a number deleted, this may be of importance in preventing them of these disorders.42–44,46–52 Regardless of the initiating from evading the immune response and therefore factor once an inflammatory response is initiated if left leaving them more susceptible to immune attack, unlike unresolved, it may contribute to death of vulnerable their wild-type counterparts. In addition, replication of neuronal populations. viruses can also be inhibited by endogenous cellular These changes in the aging brain can result in greater miRNAs but this control is unlikely to be extended to susceptibility and sensitivity to infection or immune viral vectors because genes encoding the target viral stimulants.41 A number of reports shows that a systemic transcripts will not be expressed or will be deleted in immune challenge has the ability to modulate the viral vectors. Although most of the work to date on immune response in the brain and that this is more miRNA modulation of host–virus responses has been exaggerated in an aged individual. Systemic administra- undertaken with respect to wild-type viral infections tion of a peripheral toxin such as staphylococcal outside the CNS, aspects of these findings are of enterotoxin A in an animal model shows that as well potential relevance to CNS viral infections and to the as engaging the innate arm of the immune response function of gene therapy vectors within the CNS. A full activated T cells also exert neuromodulatory effects.53 understanding of the implications of miRNAs in relation Staphylococcal enterotoxin A increased levels of IL-1b in to gene therapy vectors awaits further investigation (for the aged, but not young, mice 2 h after administration.54 summary see Figure 1b). Stimulation of the peripheral innate immune system in In summary, a complex repertoire of factors is aged Balb/c mice causes exaggerated neuroinflamma- involved in regulating and determining the immune tion.55 MHC class II expression was increased specifically response to viruses in the CNS. This immune regulation in aged microglia, causing them to maintain a reactive or is finely balanced and inappropriate activation can result primed phenotype. Peripheral lipopolysaccharide injec- in a pathogenic rather than a protective outcome.19 It is tion in mice caused microglia overexpression of TLR2 important to understand in what circumstances robust and resultant induction of pro-inflammatory cytokines immune activation can occur and what critical triggers compared to adults.56 Earlier work showed that systemic initiate it. and central endotoxin challenges exacerbate the local

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 451

Figure 1 Overview of innate immune viral sensing systems and microRNA silencing mechanisms in a brain cell. (a) Simplified schematic representation of immune system cell-surface and endosomal pattern recognition receptors that recognize viruses. TLRs (Toll-like receptors) are the main cell-surface and endosomal receptors. They signal through the adapter protein MyD88 (myeloid differentiation primary response gene) or TRIF (TIR-domain-containing adapter-inducing interferon-b) to activate transcription factors causing upregulation of cytokines and chemokines inducing inflammation and immune regulation. Other intracellular surveillance molecules RIG1 (retinoic acid- inducible gene 1)-like receptors; NLR (nucleotide-binding oligomerization domain (Nod)-like receptors), APOBEC, (apolipoprotein B mRNA editing enzyme, catalytic polypeptide)-like family of cytidine deaminases have also been implicated in viral nucleic acid recognition and in promoting the maturation of inflammatory cytokines on activation resulting in ubiquitination and degradation. For simplicity not all cellular factors involved are shown. (b) Schematic representation of miRNA-mediated gene regulation of host–virus interaction. Viral or cellular DNA encodes nonprotein coding miRNA genes. The initial product is a transcript that contains one or more stemmed looped structures, known as a primary miRNA (pri-miRNA). The nuclease Drosha cleaves the pri-miRNA stem, excising hairpin loops known as precursor miRNAs. These are exported to the cytoplasm and processed by another nuclease, Dicer, generating mature miRNAs of B22 . The miRNAs are loaded into a protein complex known as the RNA-induced silencing complex (RISC), which they then guide to the target mRNA to exert their effector function. Binding of the RISC to an mRNA bearing extensive sequence complementarity to the miRNA generally results in mRNA cleavage and degradation whereas binding to mRNAs bearing partial complementarity results mainly in translational arrest. Viruses exploit the cellular miRNA-processing pathway to generate viral miRNAs and can saturate the system causing its suppression. Viral miRNAs can inhibit the expression of cellular genes including inflammatory genes and other cellular miRNAs with anti-viral activity, including miRNA biogenesis. Cellular miRNAs can inhibit viral gene expression. Viral vectors with many or all viral genes removed will be less susceptible to these controls; in particular they are unlikely to express viral miRNAs to inhibit of cellular inflammatory genes making the vectors potentially more immunogenic. brain inflammatory response and increase neuronal neurodegenerative disease is likely to increase the risk of death during chronic inflammation.41,55 If a systemic an immune response with gene therapy vectors. immune challenge can modulate an immune response in the aged brain, it is very likely that the aged brain will be more susceptible to an immune challenge in the context CNS immune response to gene therapy of a gene therapy vector delivered directly to the brain. Given that a major focus of CNS gene therapy will be vectors depends on vector type, vector to treat neurodegenerative disease, the implications of dose and transgene increased inflammation associated with the disease process and general age-related immunosenesence As discussed above the CNS is immune attenuated and should be considered and their risk assessed. If the the innate immune/inflammatory status can be influ- immune system is already activated, preexisting inflam- enced both by patient age and whether there is ongoing mation might (1) increase the risk of an immune disease With gene therapy vectors, the likelihood of response to a gene therapy vector and/or (2) immune immune activation depends on the immune stimulus, responses elicited by gene therapy vectors might exacer- that is, the properties of the relevant gene therapy bate the target neurodegenerative disease process. In vectors and the host status. Although the main gene summary, treatment of aged individuals with ongoing therapy vectors in use for the CNS are viral vectors,

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 452 modified to be nonreplicating, their immunogenicity is widespread vector administration is required (and thus a directly dependent on vector type, vector dose and higher antigenic load delivered), such as in late infantile transgene. In view of recent and future clinical results of neuronal ceroid lipofuscinosis, an autosomal recessive, CNS gene therapy applications, it is important to identify neurodegenerative lysosomal storage disease affecting the factors or combinations of factors that contribute to the CNS.67 Although there were no unexpected serious potential high-risk situations. adverse events in this study unequivocally attributable The earliest gene therapy vectors, as shown princi- to the vector, there were serious adverse effects, the pally by the Wood and Lowenstein laboratories, elicited etiology of which could not be determined. One subject both innate and adaptive CNS responses in normal died 49 days after surgery after developing status animals. With HSV-derived vectors, the viral envelope epilepticus on day 14, but with no evidence of CNS glycoproteins were found to elicit a humoral immune inflammation. Of the 10 subjects, 4 developed a mild, response. In the case of adenovirus, an innate immune mostly transient, humoral immune response to the vector response was rapidly stimulated by viral capsid proteins. though none of the patients had detectable serum Even though the genes involved in replication and antibodies before treatment. However, caution is neces- pathogenesis were removed from these vectors, they sary because as yet there is no long-term clinical trial were still found to exhibit low levels of viral gene data available, the trials have been conducted on expression resulting in production of immunogenic viral relatively few patients and they have not all been tested gene products limiting their potential application for at higher vector doses. neurodegenerative disorder gene therapy.57,58 have also shown considerable promise In view of the immune response elicited by such early for treatment of neurodegenerative disorders. The vectors, the focus of attention in later generation vectors development of multiply deleted vectors in which has been to reduce/eliminate viral gene expression to the vast majority of the viral genome is removed has improve their clinical use. Hence more recently deve- helped to minimize the induction of a host immune loped adenoviral vectors, with all viral genes deleted, response.68,69 There are now numerous examples of AAV and lentiviral vectors are now being actively effective long-term treatment of animal models of exploited for gene delivery to the CNS. Any immune neurological disorders, such as PD, Alzheimer’s disease, response to such vectors should therefore only be elicited Huntington’s disease, motor neuron diseases, lysosomal to the input virions and potentially the transgene. storage diseases and spinal injury, using a range of Because virion-derived proteins provide only a limited therapeutic genes expressed in lentiviral vectors.68,70 As source of antigenic epitopes, any immune activation yet, however, there are no data from the clinical trials (for would be expected to be a self-limiting process. When a summary see Table 1). Although such vectors were fully deleted adenovirus with all coding sequences originally thought not to be an immunological risk, deleted was injected into the mouse brain, inflammation recent studies outside the CNS suggest that the thera- was greatly diminished at 14 days after administration peutic benefit of lentiviral-mediated gene therapy may compared with a first-generation vector and had be hindered by the activation of destructive T cells.71 In a subsided by 30 days. In a separate study, in rats the study in which GFP and luciferase reporter genes were numbers of macrophages and T lymphocytes infiltrating used to study transgene-specific T-cell activation after the brain were greatly reduced with fully deleted lentivirus-mediated gene transfer to mouse lung, both adenovirus vectors compared with earlier vectors. transgenes were shown to elicit transgene-specific T cells AAV has considerable appeal as a gene therapy in C57BL/6 and BALB/c mice.71,72 On account of this vector both in the periphery and CNS due to its reduced immunogenicity, lentiviruses are now being developed viral gene expression and low immunogenicity. as novel genetic vaccines.73–75 Although this study did Although AAV is devoid of viral-coding sequences the not deal with the immune response to lentiviral vectors cloning capacity of AAV is limited, 4.8 kb as opposed to in the CNS, caution is also warranted for CNS delivery. 35 kb with fully deleted adenovirus. A single adminis- One of the most important factors influencing the tration of AAV in the brain or the periphery of a naive immune response to viral vectors is the dose of vector animal is considered to be minimally immunogenic.59 injected. High vector doses have been shown to cause an AAV is also capable of infecting dividing and non- increased immune response and neuropathological dividing cells, and maintaining stable, long-term gene damage in the dopaminergic system of direct relevance expression in differentiated cells, especially neurons.60 to PD.57,58 With lower titers the immune response is Thus, neuronal transduction can theoretically provide greatly diminished at 14 days after administration and protein production for several years,61–63 which is a by 30 days immune activity is typically restricted to the highly advantageous attribute when considering the site of vector injection. It is likely therefore that with the treatment of long-term, progressive neurodegenera- delivery of higher doses, more vector antigen finds its tive disorders. way into the CSF or blood where it can exit the CNS to AAV serotype 2 (AAV2) has been the vector of choice local sites where antigen presentation is enabled.76 for recent PD Phase I clinical trials aimed at delivery of Immune responses to the transgene could be detri- neurturin,64 decarboxylase65,66 and human mental to the host by not only preventing further gene aromatic l-amino-acid decarboxylase62 (see summary in therapy attempts but also by preventing the administra- Table 1). These trials, while reporting encouraging safety tion of therapeutic proteins. However, few studies have data thus far, have involved only small numbers of specifically addressed the question of whether the patients and in some cases were dose escalation studies immune responses elicited by viral vectors in the CNS with only a limited number of patients receiving the are directed exclusively against the transgene or against highest vector dose.65 Greater concern may be warranted the transgene in addition to the vector. Recent work with with younger patients and in particular those where AAV in the brain has shown that the antigen that sparks

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 453 Table 1 Complete and ongoing gene therapy trials for CNS neurodegenerative disease

Disease Transgene Virus/ Clinical Status Sponsor References delivered serotype phase

Alzheimer’s NGF, Cre-110 AAV2 I Active not Ceregene American Academy of disease recruiting Neurology Annual Meeting 2007 Alzheimer’s NGF, Cre-110 AAV2 II Not yet Ceregene Update: http:// disease recruiting www.ceregene.com Batten disease CLN2, cDNA AAV2 I Recruiting Weill Medical College 67 (LINCL2) of Cornell University Nathan’s Battle Foundation Canavan’s ASPA AAV2 I Complete Canavan Research 96,97 disease Fund, Canavan (leukodystrophy) Research Foundation, Canavan Research, Illinois, NINDS Parkinson’s Neurturin AAV 2 I Complete Ceregene Results: http:// disease (Cre-120) www.ceregene.com Parkinson’s Neurturin AAV 2 II Complete Ceregene 64 disease (Cre-120) Parkinson’s GAD 65 AAV 2 I Complete Neurologix Inc. 65,98 disease Parkinson’s GAD 65 AAV 2 II Recruiting Neurologix Inc. 65,66 disease Parkinson’s hAADC AAV2 I Active Genzyme 62 disease Parkinson’s AADC, TH Lentivirus I/II Recruiting Oxford Biomedica Update: http://www. disease and GTP-CH1 oxfordbiomedica. (Prosavin) co.uk

Abbreviations: AADC, aromatic amino-acid decarboxylase; ASPA, ; CH1, cyclohydrolase 1; CLN2, ceroid lipofuscinoses, neuronal 2; LINCL2, late infantile neuronal lipofuscinosis; NGF, Nerve growth factor; TH, tyrosine hydroxylase. Sources: Clinical trials.gov and other published data. a brain immune response to AAV2 is the capsid protein exposure has taken place outside the CNS; and (2) where and not the transgene.77 When lentiviral vectors have re-administration of vector is required. been injected into the CNS there was no measurable Several studies have shown that peripheral infection immune response to the transgene. However, systemic with a gene therapy virus can severely compromise immunization with a lentiviral vector expressing the subsequent CNS gene delivery, as activated T and B same transgene caused an infiltration of inflammatory lymphocytes readily cross the BBB under such condi- cells into the CNS and it was shown that there was no tions. Where there has been previous peripheral expo- immune response to the vector itself, only to the sure to a virus such as adenovirus, a more severe transgene antigenic epitopes. Lentiviral gene delivery inflammatory response is subsequently initiated in the of nerve growth factor leading to gene expression was brain resulting in elimination of the vector, than is seen detectable in non-human primate brains for at least 1 after a local brain parenchymal injection alone. There is year after gene delivery, with no activation of inflamma- now evidence of a systemic immunization threshold for a tory markers.78 In summary, whether or not there is an dose that generates an immune response strong enough immune response to the transgene depends to some to eliminate transgene expression from the CNS. Once extent on whether the gene product is syn- or xeno- the threshold for systemic immunization is reached, the geneic, the latter being more immunogenic, and whether immune response eliminates transgene expression by or not there has been a previous exposure to the gene 490% even from brains that receive as little as 1000 product, which increases the risk of an exacerbated infectious units of adenoviral vectors.76 In contrast, inflammatory response.77 preexisting anti-adenoviral immunity is not a barrier to efficient and stable transduction of the brain by fully 76 Previous exposure to the virus deleted adenovirus vectors. Tetracycline-regulated high-capacity adenoviral vectors also sustain regulated or re-introduction of the viral vector transgene expression even in the presence of preexisting may suppress transgene expression systemic immune though the mechanisms involved here 79 and increase immunological risk are not clear. Peripheral immunization of rats with wild-type AAV Even with improved, less immunogenic vectors the resulted in high levels of serum neutralizing antibodies preexisting state of the adaptive immune system will and prevented transduction by AAV2 as assessed by influence immunological risk/outcome. Two potentially striatal GDNF levels. However, transduction was not high-risk situations should be considered: (1) previous affected when AAV5 was used for the subsequent striatal exposure to the vector antigens, in particular if the injection.80

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 454 In all gene therapy clinical trials undertaken for CNS earlier serotypes.86–88 Thus, for a re-administration neurodegenerative disease to date, there have been protocol an immune response against AAV2 capsid no reports of any adverse events associated with proteins can be circumvented by re-administration with the immune response.67 In trials where subjects had a different serotype.77 As well as human serotypes, preexisting neutralizing antibodies, no adverse immune non-human primate AAV serotypes such as AAV Rh10 events were noted.63,64,65 However, in view of the chronic have also been explored for delivery to the CNS.89 In nature of neurodegenerative disease, it is highly likely addition, by using directed evolution with a diverse that vector re-administration will need to be considered, array of novel AAV libraries, a new generation of AAV which raises the question of whether this poses any vectors capable of highly efficient delivery to astrocytes additional immunological risk. When a first-generation has been engineered.90 Self-complementary AAV vectors, adenoviral vector was re-administered to the same site in which bypass the required second-strand DNA synthesis the brains of rats one month after an initial injection, this to achieve transcription of the transgene have recently resulted in a damaging immune response and detectable been used for retinal gene therapy without adverse neuronal cell death.57 immune responses. Transgene expression lasted for over The During laboratory showed that re-administration 3.5 months in rats and 2.35 years in monkeys without of the same AAV to the CNS was possible but it depends any reported adverse effects in these studies. A number on the interval between the first and second injection. of non-HIV lentiviruses such as equine infectious anemia In subsequent studies, although low-level immune virus, simian immunodeficiency virus and feline immuno- responses were detected after intraparenchymal brain deficiency virus are also being developed. Equine in- administration of AAV, repeat administration of the same fectious anemia virus has a high transduction capacity vector serotype led to an increase in neutralizing anti- and limited adverse effects, is highly specific for gene bodies, with conflicting reports between different studies transfer to adult neural stem cells and has been of a reduction in transgene expression. In the most recent successfully used in neural tissue.91 The quiescent and study in rats, intrastriatal re-administration of AAV2 slow-dividing nature of adult stem cells was considered clearly resulted in an intensified immune response in the to be the likely basis of the affinity of these stem cells for second injection site.77 In this study, a single administra- lentivurses. Moreover, the use of miRNAs to regulate tion of AAV expressing GFP was found to be minimally vector tropism may provide a valuable tool for viral immunogenic but intrastriatal re-administration of the vectors to evade the immune system and allow repeated same vector and transgene resulted in an intensified vector administration.92 miRNA targeting involves immune response in the injection site showing that engineering the viral genome to contain miRNA target antigen presentation and priming of the immune system (miRT) elements that can then be recognized and can and does occur in the brain. There was a 70% regulated by endogenous cellular miRNAs or, possibly, reduction in transduced cells expressing GFP and the viral miRNAs. miRNA targets could possibly decrease presence of macrophages and CD8 T cells in the injection antigen presentation that could dampen an antiviral site suggested a T-cell-mediated immune response may immune response. It seems likely that viral miRNAs have been responsible for the loss in transgene expres- could be designed against many targets (both cellular sion.77 Therefore, at present vector re-administration is and viral) to enhance the performance of viruses. likely to constitute a significantly high immune risk in Potential targets that would limit natural killer-mediated CNS gene therapy applications. recognition, limit viral antigen presentation, control the interferon response, suppress inflammatory cytokine release and inhibit cellular apoptosis, could prove to be Novel viral vectors further reduce of particular importance. immunological risk In summary, the latest vectors have been refined to reduce immunogenicity by having all the viral genomic The high-risk immune situations for neurodegenerative sequences removed and can be produced safely for the disease gene therapy are likely to be where the immune clinic.68 Despite such improvements in vector design, system is already compromised, the use of immunogenic there are still reports of immune responses in certain viral vectors, high vector dose requirements, cases in situations and some of these relate to specific host which there is preexisting inflammation and ongoing factors, such as species, disease status or age, that in the progressive disease where continued treatment and clinical situation are unalterable variables. One clearly vector re-administration may be required. identifiable risk factor is previous exposure to the virus With improvements in vectorology newer vectors that or the requirement for repeat delivery; thus, caution are likely to reduce immune risk further continue to be must be observed and individuals evaluated with the engineered.81–83 The most recent generation of adenoviral knowledge of situations likely to constitute higher risk. vectors allows the use of longer tissue-specific or regulatable promoters that can modulate the immune 84 response. Capsid modification with polyethylene Prospects glycol (PEGylation) reduced the innate immune response of fully deleted vectors in the periphery and therefore Within the near future, data should emerge from the such modification also has potential for reducing the Phase I/II trials using a range of different gene therapy immune response in the CNS.85 To evade an immune vectors for neurodegenerative disease. As well as response to AAV vectors, capsid modification also holds providing possible efficacy data, these trials should promise. In addition, there are now several different inform directions required for further research into gene serotypes of AAV in use for the CNS, which have greater therapy vectors, and for the CNS in particular, any and more widespread transduction efficiencies than adverse immune events caused by the vectors that

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 455 Table 2 High-risk immune situations for gene therapy for neurodegenerative disease

High-risk factors Circumstances of risk Immune manifestation

Age Old age Increased immune activation in senescent brain Disease status Active neurodegenerative disease Ongoing active inflammation Vector type Vectors intrinsically immunogenic, in Vectors with larger more complex genomes are particular early generation vectors more immunogenic Vector dose High dose including multiple sites may High dose more likely to trigger an immune be necessary to achieve effect response—may be threshold effect Transgene Exposure to antigen to which immune Naive immune response to novel transgene- system is not previously tolerized derived antigens Previous exposure to vector Many individuals previously exposed to Immune response exacerbated by a previously viruses used for gene therapy primed immune system Multiple administrations of vector Chronic ongoing disease may require Repeat administration leads to an increased repeated administration adaptive immune response should be addressed. The major factors influencing showed preferential expansion of activated CD4 cells, immune risk are summarized in Table 2. A better under- increasing the number of virus targets and leading to standing of basic CNS immunobiology to include greater higher susceptibility to HIV infection.95 Thus, for the knowledge of the acute innate immune response, and in purposes of virus delivery to people with a previous particular the function of pathogen recognition receptors exposure to the vector, the problem may be overcome by will shed further light on the nature of both the innate using vectors from rarer serotypes. For re-administra- and adaptive immune responses in the CNS. TLRs are tion, it will almost certainly be necessary to use a particularly important as they are targeted by viral different vector serotype for subsequent deliveries or to immune evasion strategies,93and identifying the target include an immunosuppressive agent within the treat- host proteins of such viral inhibitors could provide ment protocol. Where the transgene contributes to valuable insights into how to engineer gene therapy immunogenicity, the potential deleterious effects could vectors successfully to evade the immune response. The be minimized with the use of inducible promoters and use of novel vector serotypes and novel chemical formu- tight regulation. lations or methods for immunosuppression are all potentially important in this regard. An increased understanding of miRNA-mediated gene regulation Conclusion should also provide insight into virus–host interactions and the regulatory control of the innate immune res- There has been much progress in recent years in ponse of direct relevance to developing safer gene reducing the immunogenicity of viral vectors for therapy vectors and protocols. Improved understanding delivery to the CNS. However, there remain potential should potentially provide therapeutic strategies for risks, in chronic disease states in aged individuals immune response regulation and targets for pharmaco- with activated immune systems, in particular where logical intervention. By incorporating target sites for a there has been earlier exposure to the vector or where re- specific miRNA into a transgene or the genome of a administration is required. Although systematic studies virus, transgene expression becomes susceptible to in animal models are useful, it is not always possible regulation in cells in which that miRNA is expressed to extrapolate from these to the clinical situation and and may provide a novel strategy to prevent transgene thus a critical awareness of potential high-risk situations expression in nontarget cells, for example APCs.94 is imperative in the design and conduct of future The major immunological risks for CNS gene therapy clinical trials. applications are (1) where the immune system is already activated in an aged individual or in the case of ongoing chronic disease, (2) where there has been previous Conflict of interest exposure to the virus or vector and (3) instances or protocols where vector re-administration is required. The authors declare no conflict of interest. Individuals may respond differently depending on their disease and immune status, which may require tailoring treatments accordingly. Strategies to limit the immune Acknowledgements response therefore need to be designed and tested with these findings in mind. Where patients have preexisting We thank members of the Wood laboratory for critical antibodies or cytotoxic T cells to a vector from a previous input and in particular Yiqi Seow for help in the wild-type exposure, these levels should be carefully preparation of Figure 1. MJAW receives funding from monitored in view of data from peripheral gene therapy the Medical Research Council UK and the Parkinson’s trials, as both innate immune responses to adenovirus Disease Society. challenge and preexisting adenovirus immunity repre- sent fundamental problems, with great implications for both safety and efficacy of adenovirus vector gene References therapy applications.85,95 In a recent clinical trial invol- ving an adenoviral-based vaccination against HIV 1 Galea I, Bechmann I, Perry VH. What is immune privilege (not)? individuals with preexisting immunity to adenovirus Trends Immunol 2007; 28: 12–18.

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 456 2 Bechmann I, Galea I, Perry VH. What is the blood-brain barrier 24 Saito T, Owen DM, Jiang F, Marcotrigiano J, Gale Jr M. Innate (not)? Trends Immunol 2007; 28: 5–11. immunity induced by composition-dependent RIG-I recognition 3 Glezer I, Simard AR, Rivest S. Neuroprotective role of the innate of hepatitis C virus RNA. Nature 2008; 454: 523–527. immune system by microglia. Neuroscience 2007; 147: 867–883. 25 Sabbah A, Chang TH, Harnack R, Frohlich V, Tominaga K, Dube 4 Perry VH. Stress primes microglia to the presence of systemic PH et al. Activation of innate immune antiviral responses by inflammation: implications for environmental influences on the Nod2. Nat Immunol 2009; 10: 1073–1080. brain. Brain Behav Immun 2007; 21: 45–46. 26 Argyris EG, Acheampong E, Wang F, Huang J, Chen K, Mukhtar 5 Ransohoff RM, Perry VH. Microglial physiology: unique stimuli, M et al. The interferon-induced expression of APOBEC3G in specialized responses. Annu Rev Immunol 2009; 27: 119–145. human blood-brain barrier exerts a potent intrinsic immunity to 6 Barcia C, Thomas CE, Curtin JF, King GD, Wawrowsky K, block HIV-1 entry to central nervous system. Virology 2007; 367: Candolfi M et al. In vivo mature immunological synapses 440–451. forming SMACs mediate clearance of virally infected astrocytes 27 Sullivan CS. New roles for large and small viral in from the brain. J Exp Med 2006; 203: 2095–2107. evading host defences. Nat Rev Genet 2008; 9: 503–507. 7 Kielian T. Overview of toll-like receptors in the CNS. Curr Top 28 Gottwein E, Cullen BR. Viral and cellular microRNAs as Microbiol Immunol 2009; 336: 1–14. determinants of viral pathogenesis and immunity. Cell Host 8 Aravalli RN, Hu S, Lokensgard JR. Toll-like receptor 2 signaling Microbe 2008; 3: 375–387. is a mediator of apoptosis in herpes simplex virus-infected 29 Moens U. Silencing viral microRNA as a novel antiviral therapy? microglia. J Neuroinflammation 2007; 4:11. J Biomed Biotechnol 2009; 2009: 419539. 9 Tang SC, Arumugam TV, Xu X, Cheng A, Mughal MR, Jo DG 30 Ghosh Z, Mallick B, Chakrabarti J. Cellular versus viral et al. Pivotal role for neuronal Toll-like receptors in ischemic microRNAs in host-virus interaction. Nucleic Acids Res 2009; brain injury and functional deficits. Proc Natl Acad Sci USA 2007; 37: 1035–1048. 104: 13798–13803. 31 Wahner AD, Sinsheimer JS, Bronstein JM, Ritz B. Inflammatory 10 Vercammen E, Staal J, Beyaert R. Sensing of viral infection and cytokine gene polymorphisms and increased risk of Parkinson activation of innate immunity by toll-like receptor 3. disease. Arch Neurol 2007; 64: 836–840. Clin Microbiol Rev 2008; 21: 13–25. 32 Heneka MT, O’Banion MK. Inflammatory processes in 11 Pedras-Vasconcelos J, Puig M, Verthelyi D. TLRs as therapeutic Alzheimer’s disease. J Neuroimmunol 2007; 184: 69–91. targets in CNS inflammation and infection. Front Biosci (Elite Ed) 33 Infante J, Garcia-Gorostiaga I, Sanchez-Juan P, Sanchez-Quintana 2009; 1: 476–487. C, Gurpegui JL, Rodriguez-Rodriguez E et al. Inflammation- 12 van Noort JM, Bsibsi M. Toll-like receptors in the CNS: related genes and the risk of Parkinson’s disease: a multilocus 15 implications for neurodegeneration and repair. Prog Brain Res approach. Eur J Neurol 2008; : 431–433. 34 Lassmann H. Mechanisms of inflammation induced tissue injury 2009; 175: 139–148. in multiple sclerosis. J Neurol Sci 2008; 274: 45–47. 13 Falsig J, van Beek J, Hermann C, Leist M. Molecular basis for 35 O’Brien K, Fitzgerald DC, Naiken K, Alugupalli KR, Rostami AM, detection of invading pathogens in the brain. J Neurosci Res 2008; Gran B. Role of the innate immune system in autoimmune 86: 1434–1447. inflammatory demyelination. Curr Med Chem 2008; 15: 1105–1115. 14 Barcia C, Sanderson NS, Barrett RJ, Wawrowsky K, Kroeger KM, 36 Rezai-Zadeh K, Gate D, Town T. CNS infiltration of peripheral Puntel M et al. T cells’ immunological synapses induce immune cells: D-day for neurodegenerative disease? JNeuroimmune polarization of brain astrocytes in vivo and in vitro: a novel Pharmacol 2009; 4: 462–475. astrocyte response mechanism to cellular injury. PLoS One 2008; 37 Owens T. Toll-like receptors in neurodegeneration. Curr Top 3: e2977. Microbiol Immunol 2009; 336: 105–120. 15 Barcia Jr C, Gomez A, Gallego-Sanchez JM, Perez-Valles A, 38 Nimmo AJ, Vink R. Recent patents in CNS drug discovery: the Castro MG, Lowenstein PR et al. Infiltrating CTLs in human management of inflammation in the central nervous system. glioblastoma establish immunological synapses with tumori- Recent Pat CNS Drug Discov 2009; 4: 86–95. Am J Pathol 175 genic cells. 2009; : 786–798. 39 Rivest S. Regulation of innate immune responses in the brain. 16 Weiss HA, Millward JM, Owens T. CD8+ T cells in inflammatory Nat Rev Immunol 2009; 9: 429–439. demyelinating disease. J Neuroimmunol 2007; 191: 79–85. 40 Frank-Cannon TC, Alto LT, McAlpine FE, Tansey MG. Does 17 Frischer JM, Bramow S, Dal-Bianco A, Lucchinetti CF, Rauschka neuroinflammation fan the flame in neurodegenerative diseases? H, Schmidbauer M et al. The relation between inflammation and Mol Neurodegener 2009; 4: 47. neurodegeneration in multiple sclerosis brains. Brain 2009; 132: 41 Abraham J, Johnson RW. Central inhibition of interleukin-1beta 1175–1189. ameliorates sickness behavior in aged mice. Brain Behav Immun 18 Koedel U. Toll-like receptors in bacterial meningitis. Curr Top 2009; 23: 396–401. Microbiol Immunol 2009; 336: 15–40. 42 Hirsch EC, Hunot S. Neuroinflammation in Parkinson’s 19 Suh HS, Brosnan CF, Lee SC. Toll-like receptors in CNS viral disease: a target for neuroprotection? Lancet Neurol 2009; 8: infections. Curr Top Microbiol Immunol 2009; 336: 63–81. 382–397. 20 Iacobelli-Martinez M, Nemerow GR. Preferential activation of 43 Reale M, Iarlori C, Thomas A, Gambi D, Perfetti B, Di Nicola M Toll-like receptor nine by CD46-utilizing adenoviruses. J Virol et al. Peripheral cytokines profile in Parkinson’s disease. 2007; 81: 1305–1312. Brain Behav Immun 2009; 23: 55–63. 21 Zhu J, Huang X, Yang Y. Innate immune response to adenoviral 44 Banks WA, Farr SA, Morley JE, Wolf KM, Geylis V, Steinitz M. vectors is mediated by both Toll-like receptor-dependent and - Anti-amyloid beta protein antibody passage across the blood- independent pathways. J Virol 2007; 81: 3170–3180. brain barrier in the SAMP8 mouse model of Alzheimer’s 22 Cerullo V, Seiler MP, Mane V, Brunetti-Pierri N, Clarke C, disease: an age-related selective uptake with reversal of learning Bertin TK et al. Toll-like receptor 9 triggers an innate immune impairment. Exp Neurol 2007; 206: 248–256. response to helper-dependent adenoviral vectors. Mol Ther 2007; 45 Del Valle J, Duran-Vilaregut J, Manich G, Camins A, Pallas M, 15: 378–385. Vilaplana J et al. Time-course of blood-brain barrier disruption in 23 Muruve DA, Petrilli V, Zaiss AK, White LR, Clark SA, senescence-accelerated mouse prone 8 (SAMP8) mice. Int J Dev Ross PJ et al. The inflammasome recognizes cytosolic microbial Neurosci 2009; 27: 47–52. and host DNA and triggers an innate immune response. 46 Landreth GE, Reed-Geaghan EG. Toll-like receptors in Alzheimer’s Nature 2008; 452: 103–107. disease. Curr Top Microbiol Immunol 2009; 336: 137–153.

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 457 47 Avagyan H, Goldenson B, Tse E, Masoumi A, Porter V, Wiedau- 65 Kaplitt MG, Feigin A, Tang C, Fitzsimons HL, Mattis P, Pazos M et al. Immune blood biomarkers of Alzheimer disease Lawlor PA et al. Safety and tolerability of gene therapy with patients. J Neuroimmunol 2009; 210: 67–72. an adeno-associated virus (AAV) borne GAD gene for 48 Cashman JR, Ghirmai S, Abel KJ, Fiala M. Immune defects in Parkinson’s disease: an open label, phase I trial. Lancet 2007; Alzheimer’s disease: new medications development. 369: 2097–2105. BMC Neurosci 2008; 9 (Suppl 2): S13. 66 Feigin A, Kaplitt MG, Tang C, Lin T, Mattis P, Dhawan V et al. 49 Reale M, Iarlori C, Feliciani C, Gambi D. Peripheral chemokine Modulation of metabolic brain networks after subthalamic gene receptors, their ligands, cytokines and Alzheimer’s disease. therapy for Parkinson’s disease. Proc Natl Acad Sci USA 2007; J Alzheimers Dis 2008; 14: 147–159. 104: 19559–19564. 50 Pelegri C, Canudas AM, Del Valle J, Casadesus G, Smith MA, 67 Worgall S, Sondhi D, Hackett NR, Kosofsky B, Kekatpure MV, Camins A et al. Increased permeability of blood-brain barrier on Neyzi N et al. Treatment of late infantile neuronal ceroid the hippocampus of a murine model of senescence. Mech Ageing lipofuscinosis by CNS administration of a serotype 2 adeno- Dev 2007; 128: 522–528. associated virus expressing CLN2 cDNA. Hum Gene Ther 2008; 51 Hafezi-Moghadam A, Thomas KL, Wagner DD. ApoE deficiency 19: 463–474. leads to a progressive age-dependent blood-brain barrier 68 Valori CF, Ning K, Wyles M, Azzouz M. Development and leakage. Am J Physiol Cell Physiol 2007; 292: C1256–C1262. applications of non-HIV-based lentiviral vectors in neurological 52 Godbout JP, Johnson RW. Age and neuroinflammation: a lifetime disorders. Curr Gene Ther 2008; 8: 406–418. of psychoneuroimmune consequences. Immunol Allergy Clin 69 Dissen GA, Lomniczi A, Neff TL, Hobbs TR, Kohama SG, North Am 2009; 29: 321–337. Kroenke CD et al. In vivo manipulation of gene expression in 53 Urbach-Ross D, Crowell B, Kusnecov AW. Relationship of non-human primates using lentiviral vectors as delivery varying patterns of cytokine production to the anorexic and vehicles. Methods 2009; 49: 70–77. neuroendocrine effects of repeated Staphylococcal enterotoxin A 70 Lundberg C, Bjorklund T, Carlsson T, Jakobsson J, Hantraye P, exposure. J Neuroimmunol 2008; 196: 49–59. Deglon N et al. Applications of lentiviral vectors for biology and gene 54 Kohman RA, Crowell B, Urbach-Ross D, Kusnecov AW. therapy of neurological disorders. Curr Gene Ther 2008; 8: 461–473. Influence of age on behavioral, immune and endocrine 71 Pluta K, Kacprzak MM. Use of HIV as a gene transfer vector. responses to the T-cell superantigen staphylococcal enterotoxin Acta Biochim Pol 2009; 56: 531–595. A. Eur J Neurosci 2009; 30: 1329–1338. 72 Limberis MP, Bell CL, Heath J, Wilson JM. Activation of 55 Henry CJ, Huang Y, Wynne A, Hanke M, Himler J, Bailey MT transgene-specific T cells following lentivirus-mediated gene et al. Minocycline attenuates lipopolysaccharide (LPS)-induced delivery to mouse lung. Mol Ther 2009; 18: 143–150. neuroinflammation, sickness behavior, and anhedonia. 73 He Y, Falo Jr LD. Lentivirus as a potent and mechanistically J Neuroinflammation 2008; 5: 15. distinct vector for genetic immunization. Curr Opin Mol Ther 56 Henry CJ, Huang Y, Wynne AM, Godbout JP. Peripheral 2007; 9: 439–446. lipopolysaccharide (LPS) challenge promotes microglial hyper- 74 Lopes L, Dewannieux M, Gileadi U, Bailey R, Ikeda Y, Whittaker activity in aged mice that is associated with exaggerated C et al. Immunization with a lentivector that targets tumor induction of both pro-inflammatory IL-1beta and anti-inflam- antigen expression to dendritic cells induces potent CD8+ and matory IL-10 cytokines. Brain Behav Immun 2009; 23: 309–317. CD4+ T-cell responses. J Virol 2008; 82: 86–95. 57 Gonzalez SC, McMenamin MM, Charlton HM, Goodman J, 75 Liu Y, Peng Y, Mi M, Guevara-Patino J, Munn DH, Fu N et al. Lantos T, Simpson C et al. Readministration of adenoviral gene Lentivector immunization stimulates potent CD8 delivery to dopamine neurons. Neuroreport 2007; 18: 1609–1614. responses against melanoma self-antigen tyrosinase-related 58 Lee MB, McMenamin MM, Byrnes AP, Charlton HM, Wood MJ. protein 1 and generates antitumor immunity in mice. J Immunol Th1 cytokines are upregulated by adenoviral vectors in the 2009; 182: 5960–5969. brains of primed mice. Neuroreport 2008; 19: 1187–1192. 76 Barcia C, Jimenez-Dalmaroni M, Kroeger KM, Puntel M, 59 Lowenstein PR, Mandel RJ, Xiong WD, Kroeger K, Castro MG. Rapaport AJ, Larocque D et al. One-year expression from high- Immune responses to adenovirus and adeno-associated vectors capacity adenoviral vectors in the brains of animals with pre- used for gene therapy of brain diseases: the role of immunolo- existing anti-adenoviral immunity: clinical implications. gical synapses in understanding the cell biology of neuroim- Mol Ther 2007; 15: 2154–2163. mune interactions. Curr Gene Ther 2007; 7: 347–360. 77 Peden CS, Manfredsson FP, Reimsnider SK, Poirier AE, Burger 60 Mandel RJ, Burger C, Snyder RO. Viral vectors for in vivo gene C, Muzyczka N et al. Striatal readministration of rAAV vectors transfer in Parkinson’s disease: properties and clinical grade reveals an immune response against AAV2 capsids that can be production. Exp Neurol 2007; 209: 58–71. circumvented. Mol Ther 2009; 17: 524–537. 61 Hadaczek P, Forsayeth J, Mirek H, Munson K, Bringas J, 78 Nagahara AH, Bernot T, Moseanko R, Brignolo L, Blesch A, Pivirotto P et al. Transduction of non-human primate brain with Conner JM et al. Long-term reversal of cholinergic neuronal adeno-associated virus serotype 1: vector trafficking and decline in aged non-human primates by lentiviral NGF gene immune response. Hum Gene Ther 2008; 20: 225–237. delivery. Exp Neurol 2009; 215: 153–159. 62 Eberling JL, Jagust WJ, Christine CW, Starr P, Larson P, 79 Xiong W, Candolfi M, Kroeger KM, Puntel M, Mondkar S, Bankiewicz KS et al. Results from a phase I safety trial of Larocque D et al. Immunization against the transgene but not the hAADC gene therapy for Parkinson disease. Neurology 2008; 70: TetON switch reduces expression from gutless adenoviral 1980–1983. vectors in the brain. Mol Ther 2008; 16: 343–351. 63 Eberling JL, Kells AP, Pivirotto P, Beyer J, Bringas J, Federoff HJ 80 Madsen D, Cantwell ER, O’Brien T, Johnson PA, Mahon BP. et al. Functional effects of AAV2-GDNF on the dopaminergic AAV-2 induces cell mediated immune responses directed nigrostriatal pathway in parkinsonian rhesus monkeys. against multiple epitopes of the capsid protein VP1. J Gen Virol Hum Gene Ther 2009; 20: 511–518. 2009; 90 (Pt 11): 2622–2633. 64 Marks Jr WJ, Ostrem JL, Verhagen L, Starr PA, Larson PS, Bakay 81 Seiler MP, Cerullo V, Lee B. Immune response to helper RA et al. Safety and tolerability of intraputaminal delivery of dependent adenoviral mediated liver gene therapy: challenges CERE-120 (adeno-associated virus serotype 2-neurturin) to and prospects. Curr Gene Ther 2007; 7: 297–305. patients with idiopathic Parkinson’s disease: an open-label, 82 Malecki M, Wozniak A, Janik P. [Adeno-associated viruses phase I trial. Lancet Neurol 2008; 7: 400–408. (AAV)]. Postepy Biochem 2008; 54: 57–63.

Gene Therapy Immunobiology of gene therapy for neurodegenerative disease MM McMenamin and MJA Wood 458 83 Brunetti-Pierri N, Ng P. Progress and prospects: gene therapy for 91 Jacquet BV, Patel M, Iyengar M, Liang H, Therit B, Salinas- genetic diseases with helper-dependent adenoviral vectors. Gene Mondragon R et al. Analysis of neuronal proliferation, migration Therapy 2008; 15: 553–560. and differentiation in the postnatal brain using equine infectious 84 Segura MM, Alba R, Bosch A, Chillon M. Advances in helper- anemia virus-based lentiviral vectors. Gene Therapy 2009; 16: dependent adenoviral vector research. Curr Gene Ther 2008; 8: 1021–1033. 222–235. 92 Kelly EJ, Russell SJ. MicroRNAs and the regulation of vector 85 Zaiss AK, Machado HB, Herschman HR. The influence of innate tropism. Mol Ther 2009; 17: 409–416. and pre-existing immunity on adenovirus therapy. J Cell Biochem 93 Stack J, Bowie AG. Characterisation of viral proteins that inhibit 2009; 108: 778–790. Toll-like receptor signal transduction. Methods Mol Biol 2009; 517: 86 Taymans JM, Vandenberghe LH, Haute CV, Thiry I, Deroose CM, 217–235. Mortelmans L et al. Comparative analysis of adeno-associated 94 Brown BD, Naldini L. Exploiting and antagonizing microRNA viral vector serotypes 1, 2, 5, 7, and 8 in mouse brain. Hum Gene regulation for therapeutic and experimental applications. Ther 2007; 18: 195–206. Nat Rev Genet 2009; 10: 578–585. 87 Howard DB, Powers K, Wang Y, Harvey BK. Tropism and 95 Benlahrech A, Harris J, Meiser A, Papagatsias T, Hornig J, Hayes toxicity of adeno-associated viral vector serotypes 1, 2, 5, 6, 7, 8, P et al. Adenovirus vector vaccination induces expansion of and 9 in rat neurons and glia in vitro. Virology 2008; 372: 24–34. memory CD4 T cells with a mucosal homing phenotype that are 88 Lawlor PA, Bland RJ, Mouravlev A, Young D, During MJ. readily susceptible to HIV-1. Proc Natl Acad Sci USA 2009; 106: Efficient gene delivery and selective transduction of glial cells in 19940–19945. the mammalian brain by AAV serotypes isolated from non- 96 Janson C, McPhee S, Bilaniuk L, Haselgrove J, Testaiuti M, Freese human primates. Mol Ther 2009; 17: 1692–1702. A et al. Clinical protocol. Gene therapy of Canavan disease: 89 Klein RL, Dayton RD, Tatom JB, Henderson KM, Henning PP. AAV-2 vector for neurosurgical delivery of aspartoacylase gene AAV8, 9, Rh10, Rh43 vector gene transfer in the rat brain: effects (ASPA) to the human brain. Hum Gene Ther 2002; 13: 1391–1412. of serotype, promoter and purification method. Mol Ther 2008; 97 McPhee SW, Janson CG, Li C, Samulski RJ, Camp AS, Francis J 16: 89–96. et al. Immune responses to AAV in a phase I study for Canavan 90 Koerber JT, Klimczak R, Jang JH, Dalkara D, Flannery JG, disease. J Gene Med 2006; 8: 577–588 Schaffer DV. Molecular evolution of adeno-associated virus 98 Luo J, Kaplitt MG, Fitzsimons HL, Zuzga DS, Liu Y, Oshinsky for enhanced glial gene delivery. Mol Ther 2009; 17: ML et al. Subthalamic GAD gene therapy in a Parkinson’s 2088–2095. disease rat model. Science 2002; 298: 425–429.

Gene Therapy