|Home | About | Journals | Submit | Contact Us | Français|
Viral vector–mediated gene delivery is an attractive procedure for introducing genes into the brain, both for purposes of basic neuroscience research and to develop gene therapy for neurological diseases. Replication-defective adenoviruses possess many features which make them ideal vectors for this purpose—efficiently transducing terminally differentiated cells such as neurons and glial cells, resulting in high levels of transgene expression in vivo. Also, in the absence of anti-adenovirus immunity, these vectors can sustain very long-term transgene expression within the brain parenchyma. This unit provides protocols for the stereotactic injection of adenoviral vectors into the brain, followed by protocols to detect transgene expression or infiltrates of immune cells by immunocytochemistry or immunofluorescence. ELISPOT and neutralizing antibody assay methodologies are provided to quantitate the levels of cellular and humoral immune responses against adenoviruses. Quantitation of adenoviral vector genomes within the rat brain using qPCR is also described. Curr. Protoc. Neurosci. 50:4.24.1–4.24.49. © 2010 by John Wiley & Sons, Inc.
Viral vector–mediated gene delivery is an attractive procedure for introducing genes into the brain, both for purposes of neurobiological research and for gene therapy of neurological diseases. Replication-defective adenoviruses possess many features that make them ideal vectors for this purpose. Adenoviruses are easily purified to the high titers required for in vivo administration, and they are efficient in transducing terminally differentiated cells such as neurons and glial cells, resulting in high levels of transgene expression. At doses of 1 × 108 IU or less, highly purified, LPS-free adenoviral vectors delivered to the brain parenchyma cause minimal inflammation or toxicity, and vector-mediated transgene expression is stable for up to 12 months and spatially restricted to the region of vector administration or areas that project to the injection site (Thomas et al., 2001a; Barcia et al., 2006a).
This unit focuses on methods that have been tailored to the injection of viral vectors into the brain, and the subsequent detection of transgene expression and local inflammation. Basic Protocol 1 here complements Basic Protocol 1 in UNIT 3.10, and describes a procedure for the direct injection of adenovirus into the CNS using stereotactic guidance, while Support Protocol 1 in this unit is similar to Basic Protocol 2 in UNIT 1.1, and describes preparation of brain sections for immunohistochemical analysis. However, adenovirus-specific modifications are emphasized for these two procedures. Several methods for evaluating gene transfer to the brain following in vivo administration of the adenoviral vectors are also presented in this unit. A method for detecting vector-mediated transgene expression by horseradish peroxidase (HRP)–based immunohisto-chemical staining is described (see Basic Protocol 2). Fluorescence-based immunohis-tochemical staining to detect more than one marker antigen within a single brain section is described in Alternate Protocol 1.
Adenoviruses are immunogenic and have been used to vaccinate against infection in humans. Adenovirus-derived vectors are also immunogenic and can elicit inflammatory and cytotoxic effects in the brain that are vector dose dependent (Thomas et al., 2001a; Barcia et al., 2006a; Zirger et al., 2006). In addition to their potentially harmful side effects, inflammatory, immune, and cytopathogenic responses also have a profound impact upon the efficacy of gene transfer to the brain (Thomas et al. 2001a,b; Barcia et al., 2006a,b; King et al., 2008). It is therefore necessary to evaluate vector-mediated inflammatory responses and toxicity concomitantly with the evaluation of vector-mediated gene transfer. HRP-based immunohistochemistry (see Basic Protocol 2) or fluorescence-based immunohistochemistry (see Alternate Protocol 1) can be used with the appropriate antibodies to visualize the acute and/or chronic infiltration of immune and inflammatory cells, the activation of brain microglia and astrocytes, the loss of glial cell or neuronal markers (indicating cell loss through acute or chronic cytotoxicity), and/or the integrity of brain myelin after in vivo administration of adenovirus vectors to the brain. More global visualizations of brain structure and infiltrating cells can be obtained by staining cell nuclei with cresyl violet, which is described for Vibratome brain sections (see Basic Protocol 3), together with an alternative method for staining brain myelin with Luxol fast blue. Staining semi-thin plastic-embedded sections with toluidine blue (see Basic Protocol 4) allows even better resolution of the cellular structure of the brain. The preparation of APES-coated slides required for these applications is described in Support Protocol 2.
If vector-mediated transgene expression is observed to decrease or disappear over time, it is usually a result of inflammatory and/or immune responses against the vector or transgene. In such cases, it may be desirable to determine whether the vector genome has been eliminated from the brain (Barcia et al., 2006b; Puntel et al., 2006). Quantitative PCR–based detection of vector genomes from rat brains is described for this purpose (see Basic Protocol 5).
Adenovirus vectors injected into the brain parenchyma (particularly vectors encoding proinflammatory molecules) can cause disruption of the blood-brain barrier (BBB), facilitating the infiltration of inflammatory cells from the circulation. A method for determining the extent of breakage of the blood-brain barrier following administration of adenovirus vectors to the brain is described (see Basic Protocol 6).
Adenovirus vectors administered to the brain parenchyma are able to persist for long periods of time (at least up to 12 months, compared with 2 to 3 weeks after delivery of identical vectors to peripheral tissues such as the liver). The persistence of vector in the brains of naive animals compared to peripheral tissues is thought to be due to the failure to elicit an effective anti-adenoviral T cell response through intraparenchymal vector injection. However, an effective immune response capable of eliminating vector or transgene expression from the brain can be elicited by priming the immune response through injection of adenoviral vectors into the skin (see Basic Protocol 7). The induction of an anti-adenovirus humoral immune response can be assessed using an adenovirus-specific neutralizing antibody assay (Basic Protocol 8) and the induction of an anti-adenovirus cellular immune response can be assessed using an adenovirus specific ELISPOT assay (Basic Protocol 9).
Lastly, since collaborative efforts between different laboratories often involve the transport of valuable vector stocks from one lab to another, a protocol is provided for the optimal transport of viral vector without loss of vector titer (see Support Protocol 3).
The majority of current gene-therapy protocols utilizing adenoviral vectors have involved “first-generation” vectors. These are recombinant vectors which are rendered nonreplicative by deletion of the E1 region from the viral genome. Detailed protocols for the generation and purification of first-generation adenoviral vectors are presented in UNIT 4.23. Before any vector preparation is used in vivo, it must be accurately titered and subjected to the stringent quality control tests described in UNIT 4.23.
NOTE: All protocols using live animals must first be reviewed and approved by an Institutional Animal Care and Use Committee (IACUC) and must follow officially approved procedures for the care and use of laboratory animals.
UNIT 3.10 describes administration of inoculum to a precise location within the adult mouse CNS using stereotactic guidance. In the protocol below, the technique is adapted to viral vector injection in the adult rat brain using coordinates from Paxinos and Watson (1998). The stereotactic frame, the overall setup, and the sequence of events in this protocol are modified from the procedure described in UNIT 3.10. Briefly, the rat is manually restrained, followed by intraperitoneal injection of anesthetics [ketamine (75 mg/kg), and dexmedetomidine (0.25 mg/kg)]. Following anesthesia and prior to surgery, the analgesic Carprofen (5 mg/kg) is administered subcutaneously.
At the appropriate timepoint, euthanize the rat via terminal perfusion.
There are many different methods for evaluating adenovirus-mediated gene transfer in an in vivo paradigm. Immunohistochemical detection of the transgene product within brain sections is a widely used procedure, UNIT 1.2 describes a method in which serial brain sections are processed using “free-floating” horseradish peroxidase-based immunohisto-chemistry to detect the transgene product. In the protocol below, a modification of the ABC method described in UNIT 1.2 is presented, in which nickel chloride is added to the substrate solution to generate a blue/black precipitate (which is easier to photograph than the reddish/brown precipitate obtained with DAB alone). Further, an additional initial blocking step is incorporated whereby the sections are incubated with H2O2 to inactivate endogenous peroxidase activity. Triton X-100 is also included throughout the incubations to improve the staining of various epitopes. Also, glucose oxidase is incorporated into the staining solution to generate H2O2 from the oxidation of glucose, thus allowing staining to proceed more slowly (see Fig. 4.24.6).
Adenovirus vectors are inflammatory and immunogenic, and can be cytotoxic at high multiplicities of infection. It is therefore expedient (particularly when adenovirus vectors are used for neurological gene therapy applications) to evaluate the inflammatory, immune, and cytotoxic responses to vector delivery, simultaneously with any evaluation of vector-mediated gene transfer. The HRP-based immunohistochemistry protocol described here can be used with appropriate antibodies to visualize the acute and/or chronic infiltration of immune and inflammatory cells, the activation of brain microglia and astrocytes, the loss of glial cell or neuronal markers (indicating cell loss through acute or chronic cytotoxicity), and/or the integrity of brain myelination after in vivo administration of adenovirus vectors to the brain (see Fig. 4.24.12 and Fig. 4.24.13 for some examples). Table 4.24.1 shows some of the antibodies that have been used to investigate host responses to adenoviral vector administration in the rat striatum. All antibodies listed work well at the given dilutions, using Vibratome sections from brains that have been post-fixed in 4% paraformaldehyde solution for 48 hr.
For simultaneous analysis of the expression of two different epitopes (e.g., a cell type-specific marker and a transgene marker), fluorescence-based immunohistochemistry is often more appropriate (see Alternate Protocol 1). Microscopic analysis of fluorescently labeled cells requires tissue sections (optimally ~40 to 60 µm), generated by Vibratome sectioning, UNIT 1.1 describes cryostat sectioning of frozen brain tissue, which can be used to generate sections as thin as 5 µm. The choice of fixed versus frozen brain tissue sections will depend on the characteristics of the primary antibodies used.
UNIT 1.2, as well as Basic Protocol 2 in this unit, describe the immunohistochemical localization of a single marker antigen within tissue sections. Sometimes it may be desirable to detect two different marker antigens simultaneously within a single brain section (e.g., to determine which subpopulation of brain cells has been transduced by codetecting a cell type–specific marker with the transgene marker). The following protocol describes double-labeling of tissue sections using two different fluorescently labeled antibodies. To accurately determine the subpopulation of cells which is double-labeled, confocal microscopy will greatly improve the resolution and quality of data. This protocol describes free-floating immunofluorescent labeling of Vibratome sections; however the method can be adapted for use on frozen sections.
A global visualization of infiltrating cells and brain structure can be provided by staining cell nuclei with cresyl violet. The following protocol describes a staining procedure wherein Vibratome brain sections are histologically stained with both cresyl violet and Luxol fast blue for the simultaneous detection of cell nuclei (to visualize areas of inflammation) and myelin (to detect areas of demyelination). If desired, each stain can be used individually for the separate detection of cell nuclei and Nissl substance (cresyl violet) or myelin (Luxol fast blue). Staining of nuclei with cresyl violet is useful for the visualization of the total overall pattern of distribution of inflammatory cells infiltrating the brain after injection of adenovirus, whereas Luxol fast blue staining enables the detection of demyelination, which may occur concomitantly with adenovirus-induced inflammation. Luxol fast blue staining will be reduced in tissues displaying either demyelination or edema. When examining a novel experimental paradigm, it is important to confirm actual demyelination using specific myelin antibodies (see Table 4.24.1). Edema and many aspects of cellular and axonal integrity can also be analyzed by staining semi-thin plastic-embedded sections with toluidine blue (see Basic Protocol 4).
Toluidine blue staining of semi-thin plastic embedded sections (standard preparation for electron microscopy analysis; see UNIT 1.2) containing the injection site enables an evaluation of the integrity of the structure of neural cell bodies and processes, and allows a determination of the presence of extracellular edema within the tissue (Dewey et al., 1999).
It is possible to quantitate the levels of vector genomes within the brain or other organs. This protocol describes a quantitative PCR (qPCR)–based method for detecting adenoviral genomes within the rat brain (Puntel et al., 2006). A primer/probe set is presented with qPCR reaction parameters for quantitation of the vector genomes present in the total DNA extracted from brain tissue. The primer/probe set described below identifies a region of the L3 gene of the human adenovirus serotype 5 genome. It is essential to isolate vector genomes from brains of animals perfused with oxygenated Tyrode’s solution in absence of fixative. It is also possible to quantitate vector genomes in other peripheral organs such as liver, lung, and testes.
Due to the high sensitivity of this technique, particular care should be taken when setting up a PCR reaction to avoid cross-contamination between samples; use of aerosol-resistant filter pipet tips will reduce this risk. All tubes, tips, and solutions should be designated for this procedure only and should not be stored in close proximity to any vector stocks. Predesignated PCR-only pipets and racks should be used.
|1 cycle:||2 min||50°C|
|1 cycle:||10 min||95°C|
|40 cycles:||15 sec||95°C|
Adenoviruses are immunogenic and will cause a dose-dependent brain inflammation (see Critical Parameters). Adenovirus vectors per se injected into the brain parenchyma, or the use of adenoviruses encoding proinflammatory molecules (Bell et al., 1996), can cause disruption of the blood-brain barrier (BBB) with concomitant infiltration of inflammatory cells from the circulation. While certain cell types (e.g., activated T lymphocytes and antigen-specific B cells) are able to cross an intact BBB, other cell types can only infiltrate the brain once the BBB is broken or opened. Mechanical or vector-induced disruptions to the integrity of the BBB are usually repaired within a few days (although this could depend upon the transgene being expressed by the vector, i.e., if transgenes are pro-inflammatory (M.G. Castro and P.R. Lowenstein, unpub. observ.). Any assessment of vector-induced BBB permeability should therefore be performed within the first few days after surgery.
The following protocol (modified from Hanker et al., 1977) involves injection of the plant enzyme horseradish peroxidase (HRP) into the bloodstream of the rat via the tail vein. Under normal conditions, the intact BBB will exclude circulating HRP from the brain. A permeabilized BBB will, however, allow the HRP to enter the brain within minutes after administration. The extent of BBB permeabilization can subsequently be visualized by staining Vibratome-cut brain sections according to the Hanker-Yates method, i.e., by incubating the sections with a horseradish peroxidase (HRP) substrate solution, the components of which are metabolized to form a colored reaction product. Typical results using two different doses of first-generation adenovirus vector-expressing β-galactosidase are shown in Figure 4.24.9. Injection of 1 to 5 µg of lipopolysaccharide (LPS) into the rat striatum provides an excellent positive control for BBB breakage.
Alternatively, clinically used blood pool agents such as gadolinium-derived compounds (e.g., gadoteridol) given as an IV bolus injection and imaged using magnetic resonance angiography have shown benefits over traditionally used extracellular contrast media, and have gained increasing usage in the laboratory setting (Prohance, Bracco Diagnostics). Moreover, intravenous injections of variably sized dextrans conjugated to a number of different fluorochromes such as Texas Red, tetramethylrhodamine, and FITC (available from Invitrogen) have been successfully employed in order to label the intraluminal vascular space in studies requiring the visualization of parenchymal blood vessels through fluorescence microscopy.
Adenovirus vectors injected into the brains of naive animals are able to sustain transgene expression for at least 1 year (Barcia et al., 2007), compared with those injected into peripheral organs such as the liver, where transgene expression is generally eliminated within 2 to 3 weeks. The relative persistence of vector-mediated expression in the brain is attributed to the failure to elicit an effective anti-adenoviral T cell response following the intraparenchymal vector injection. A strong anti-adenoviral T cell response can be primed through peripheral infection with adenovirus, either prior to or after the intraparenchymal injection. After peripheral exposure to vector, activated anti-adenoviral T cells can cross the blood-brain barrier and facilitate the elimination of adenoviral vector-mediated transgene expression in the brain. Use of animals peripherally primed against adenovirus (rather than naive animals) may provide a more clinically relevant model for neurological gene therapy applications using adenovirus vectors, since most of the human population will have been exposed to, or will be at risk of, a peripheral infection with adenovirus. The following protocol describes a method for activating a strong antiviral immune response by intradermal injection of vector. This protocol has been used to prime animals 30 to 60 days after, or 2 weeks before, intracranial injection of vector (Thomas et al., 2000; Thomas et al., 2001b; Xiong et al., 2006; Barcia et al., 2006b, 2007, 2008; King et al., 2008). Although peripheral anti-adenoviral immune priming before the intracranial injection is a paradigm more likely to represent a clinical scenario, immunizing 30 to 60 days after the intracranial injection allows any early acute brain inflammation elicited by the initial intracranial injection to subside before analyzing the effect of a peripherally induced anti-viral immune response upon vector-mediated transgene expression in the brain.
Systemic administration of adenovirus by intradermal injection elicits both humoral and cell-mediated immune responses specific for adenovirus. Before conducting large experiments with multiple timepoints, it is important to confirm the induction of an anti-adenovirus humoral immune response following immunization. Below is a protocol to assess the levels of anti-adenovirus circulating neutralizing antibodies (Thomas et al., 2000; Barcia et al., 2007; Candolfi et al., 2007; King et al., 2008). This assay should be conducted at 14 days post-immunization during the peak production of neutralizing antibodies.
It is often necessary to evaluate whether individual immunized animals successfully mounted an anti-adenoviral immune response. Below is a protocol to assess the frequency of anti-adenovirus-specific IFN-γ-secreting lymphocytes from the spleens of immunized animals (Barcia et al., 2007; King et al., 2008). This assay will give the most robust signal when performed at 7 days post-immunization; however the assay may be performed up to 2 weeks post-immunization (Xiong et al., 2008). This technique may also be used to assess the frequency of IFN-γ-secreting lymphocytes specific for other antigens such as transgenes, tumor antigens, etc.
Note that alternative methods exist for assessing the frequency of anti-adenovirus-specific IFN-γ-secreting lymphocytes from the spleens of immunized animals by using peptide-MHC tetramers to analyze in vivo precursor frequency and evaluate intracellular cytokine production of antigen-specific T cells.
Perform procedures under sterile conditions in a tissue culture hood.
Perform procedures under sterile conditions in a tissue culture hood.
From this point on, procedures can be performed under nonsterile conditions.
UNIT 1.1 describes a method for perfusion fixation of animals with saline and fixative. In the protocol below, the animal is instead perfused with oxygenated Tyrode’s solution containing heparin prior to perfusing the fixative. Use of oxygenated Tyrode’s solution maintains the supply of oxygen to the brain and tissues and avoids clotting during blood clearance, thus preventing cellular death prior to perfusion of the fixative. Additionally, oxygen improves the chemical fixation reaction. If a peristaltic pump is not available, it can be replaced with the apparatus shown in Figure 4.24.10, which uses gravity to facilitate the flow of Tyrode’s and fixative. Also, the descending aorta is clamped just below the liver prior to starting the perfusion of fixative, to improve the flow of fixative to the brain. Once the perfused-fixed brains have been removed from the animal as described in UNIT 1.1, a period of post-fixing is required before the brains can be sectioned using a Vibratome, or stored in PBS/0.01% azide at 4°C. Post-fixing for 5 hr is optimal for subsequent immunohistochemical analysis (see Basic Protocol 2 and Troubleshooting). Under-fixing the brains can make subsequent sectioning very difficult, whereas over-fixing can lead to the masking or destruction of epitopes to be detected by immunohistochemical analysis, or of nucleic acids for molecular analysis. Since paraformaldehyde leaches out from the fixed brains into the PBS-azide, it is advisable to change the storage solution at regular intervals if the brains are to be kept at 4°C for long periods before processing. The duration of post-fixing required will depend upon the quality of initial fixation. If the brains are to be frozen for cryostat sectioning, then infiltration with sucrose is required, as described in UNIT 1.1. Sucrose infiltration is not required for Vibratome sectioning. This protocol describes a procedure for collecting and storing Vibratome-cut brain sections for subsequent immunohistochemical analysis. Cryostat sectioning is described in UNIT 1.1.
Gelatin-coated slides (see UNIT 1.1) can be used for most of the immunohistochemical protocols described in the current unit. For applications where slide-mounted sections are stained (e.g., Luxol fast blue staining), APES-coated slides are more appropriate since gelatin is water-soluble and the brain sections can become detached when the slides are incubated in solution for long periods of time.
Clinical trials and collaborations between research groups using recombinant adenoviral vectors require the use of dry-ice transport between laboratories in order to keep the vector stocks frozen. Transport on dry ice is essential when the vector is to be used immediately (without retitration), as the titer used for the experiments will be the one determined by the laboratory that manufactured and shipped the vector. It is recommended that the vector be retitrated after shipment (UNIT 4.23), as titration protocols (and therefore results) vary from laboratory to laboratory. This is especially important when comparing the effects of two or more viruses against each other.
It has recently been observed that significant loss of viral titer occurs during dry ice transport (Nyberg-Hoffman and Aguilar-Cordova, 1999). Thawing of the dry ice over 24 hr results in the increased production of CO2 gas in the transport container, which can penetrate the tubes containing the viruses, thereby decreasing the pH of the vector solution. The viral capsid is unstable at acid pH, and virus aliquots in small microcentrifuge tubes placed directly into dry ice for 48 hr can lose up to 6 log units of infectious titer. A common method of vector transport currently utilized is to place the microcentrifuge tubes containing the vector aliquots into 50-ml polypropylene centrifuge tubes before placing in dry ice. However, this precaution is still not sufficient for maintaining titer; a 2-log decrease in titer was observed when vector was stored in this manner in dry ice for 48 hr. A combination of sealing the aliquot in Parafilm, placing this into a 50-ml centrifuge tube, then placing the tube into two polythene bags, maintains the adenoviral vector titer after storage in dry ice for 48 hr. Alternative methods are described in the original manuscript by Nyberg-Hoffman and Aguilar-Cordova (1999).
NOTE: Each step performed below is done individually at short intervals in the −80°C freezer to keep the aliquots frozen. Freeze/thawing of the aliquot may affect the viral titer.
Use deionized, distilled water in all recipes and protocol steps. For common stock solutions, see APPENDIX 2A; for suppliers, see SUPPLIERS APPENDIX.
First-generation adenoviruses have been used to successfully deliver genes into a wide variety of non-CNS tissues and organs in animal models of human disease and in several human phase I clinical trials (Hedman et al., 2009; Li et al., 2009; Opyrchal et al., 2009; Xing et al., 2009). These studies have shown that although gene transfer is efficient, expression of the transgene is consistently transient (rarely sustained beyond 3 weeks; Chen et al., 2000; Jeong et al., 2008; UNIT 4.23) and administration of vector is associated with the development of local inflammation. The vector-induced immune response outside of the brain is biphasic and consists of an initial innate response which rapidly eliminates many transduced cells, followed by adaptive cellular and humoral mechanisms. The precise delineation of the components of the anti-vector immune response has been confounded by differences in host backgrounds, routes of vector administration, and, particularly, by strong immune responses against certain transgenes. It is widely accepted, however, that a cytotoxic T lymphocyte response directed against transduced cells presenting vector antigens in conjunction with surface MHC class I molecules is a principal cause of the decline in transgene expression observed after adenovirus-mediated gene transfer to peripheral tissues (Yang et al., 1994, 1995). The use of high-capacity vectors can reduce the acute inflammation and provide extended transgene expression (Morral et al., 1998; Kreppel et al., 2002; Brunetti-Pierri et al., 2009). High-capacity adenoviral vectors are thought to be less inflammatory because of their total absence of any original genomic adenoviral sequences. In addition, once the high-capacity adenoviral vectors have un-coated and delivered the vector genome to the cell’s nucleus, the vector genome essentially becomes invisible to the host’s immune system, as the vector genome will not express any viral protein. Further, high-capacity adenoviral vectors can be engineered to encode molecules that inhibit inflammatory and immune responses, and their viral capsids can be made less evident to the innate immune system by chemical modification of the adenovirus capsids with synthetic polymers.
In contrast to the transient transgene expression observed in the periphery, adenovirus vectors delivered to the CNS of naive animals can mediate prolonged transgene expression. The first demonstrations of adenoviral transduction of the brain parenchyma showed substantial vector-mediated transgene expression for at least 2 months (Akli et al., 1993; Davidson et al., 1993; Le Gal La Salle et al., 1993). More recent studies from our lab and others have now shown that transgene expression from Ad vectors carefully delivered to the brain parenchyma can last between 6 and 18 months (Byrnes et al., 1995; Kajiwara et al., 1997; Ideguchi et al., 2000; Thomas et al., 2000; Zou et al., 2000; Zermansky et al., 2001; Amalfitano and Parks, 2002; Glover et al., 2003). Thus, with present highly purified and LPS-free vector preparations, adenoviral-mediated gene delivery is long term. This difference can be ascribed to the distinct immunological status of the CNS as compared with other organs. The brain parenchyma is commonly described as having an immune-privileged status; allografts can survive for long periods of time without immunosuppression, and lipopolysaccharide or pro-inflammatory cytokines injected into the parenchyma induce only minimal inflammation compared to other organs (Montero-Menei et al., 1994). The CNS is not, however, devoid of immune reactions; rather it utilizes a complex interplay between different mechanisms to raise the threshold for and to regulate the progression of immune responses. Communication between the CNS and the immune system is retarded by the lack of a conventional lymphatic drainage system or resident professional antigen-presenting dendritic cells within the brain parenchyma, and by the physical obstruction to leukocyte surveillance presented by the blood-brain barrier (which in rodents is formed by a tight endothelial barrier that strictly regulates the entry of substances and cells into the brain). Activated T cells and monocytes are, however, able to cross the intact blood-brain barrier, and there is evidence to suggest that, despite the lack of conventional lymphatics, proteins can drain from the brain into deep cervical lymph nodes (Cserr and Knopf, 1997; Weller et al., 2009). Thus, these mechanisms alone do not constitute the immune privilege. Recent evidence suggests that the CNS possesses its own innate immune system composed of microglia and astrocytes, which can actively suppress or shape the immune response by direct glia:immune cell interactions or by producing specific cytokine and chemokine profiles (Matyszak, 1998; Carson and Sutcliffe, 1999; Lowenstein, 2002; Mi et al., 2004; Bechmann et al., 2007; Galea et al., 2007; Popovich and Longbrake, 2008).
Adenovirus vectors injected into the parenchyma stimulate an inflammatory response; local microglia and astroctyes are activated strongly, and inflammatory and immune cells infiltrate across the blood-brain barrier to the site of infection (Byrnes et al., 1995, 1996a; Dewey et al., 1999; Zirger et al., 2006). As previously mentioned, lysis of adenovirus-infected cells by CTLs is a principal cause of the transient transgene expression seen after vector delivery to the periphery. Depletion of CD4+ or CD8+ T cells before adenovirus administration to the brain reduces the level of inflammation seen from 6 days after vector injection, but such depletion has no effect on the level or longevity of transgene expression (Byrnes et al., 1996b). Thus, even if CD4+ or CD8+ cells mediate, in part, the inflammation, persistent adenovirus-mediated transgene expression in the brain could be a result of an ineffective T cell response which fails to clear transduced cells (Wood et al., 1996).
Although low doses of adenovirus vectors injected into the brain parenchyma do not seem able to elicit an effective systemic T cell response, a strong T cell response can be stimulated by subsequent peripheral exposure to adenovirus. In such cases, transgene expression in the brain declines rapidly and is accompanied by severe inflammation, characterized by T cell and dendritic cell infiltration, intense microglial activation, and demyelination (Byrnes et al., 1996a; Thomas et al., 2000). If animals are peripherally sensitized to adenovirus prior to injection of vector into the brain, transgene expression is transient and is again accompanied by severe inflammation (Ohmoto et al., 1999; Barcia et al., 2006a,b, 2007, 2008).
In this unit, vector toxicity is based on titers expressed as infectious units. Vector toxicity will also depend on the total number of physical viral particles administered. This is measured by optical absorbance (see UNIT 4.23). In the authors’ laboratory, vector stocks routinely contain 20 physical viral particles per infectious unit. This ratio must be determined for each individual viral stock prepared, in order to accurately gauge vector quality and related toxicity.
In contrast to the brain parenchyma, efficient immune priming occurs after vector infection of the CSF. Thus, the ventricles and the brain parenchyma represent immunologically distinct compartments, and the immune responses to intracerebral vector infection depend on the anatomical site to which the vector is delivered (Stevenson et al., 1997; Larocque et al., submitted). Adenoviral vectors injected into the lateral ventricle of adult rats elicit a rapid febrile response, which is not seen after vector delivery to the striatum. The fever response to adenoviruses administered to the CSF occurs concurrently with an increase in the concentrations of the proinflammatory cytokines TNF-α, IL-1β, and IL-6, although it has been shown that IL-1 is the key mediator of adenovirus-induced fever (Cartmell et al., 1999). Thus, it is important to carefully monitor whether vectors injected into the brain parenchyma reach the ventricles.
There have many advances over the last 10 years in the construction and use of adenoviral vectors for gene transfer into the brain, and their use for gene therapy of brain diseases, described in detail elsewhere throughout this chapter. Improvements in adenoviral vectors that facilitate their use for gene transfer to the brain include more stringent quality controls for vector preparations before their inoculation into animals, as described herein. It is important to note that in our laboratory, unless they are deemed virtually free of LPS and RCA, vectors are discarded and never used for gene-transfer experiments. Using such high-quality viral vector preparations, and keeping the vector dose below the threshold of 1 × 108 IU/injection site, we and others have achieved reliable transgene expression for up to 12 months in naive animals (Barcia et al., 2007). In addition, the use of high-capacity adenoviral vectors allows long-term expression in the brain, even in the presence of preexisting immune responses against adenovirus (Barcia et al., 2007).
Adenoviral vectors have now also been harnessed for the treatment of human brain diseases, including brain tumors such as glioblastoma multiforme grade IV, the most aggressive of brain tumors. Various approaches have been taken into the clinic since 2000. For example, Ylä-Herttuala and co-workers have pioneered the use of adenoviral vectors expressing HSV-1 derived thymidine kinase for the treatment of glioblastoma multiforme grade IV (Määttä et al., 2009), including the recent completion of an encouraging Phase III multicentric clinical trial. Further vectors being tested in the context of ongoing and future clinical trials for glioblastoma multiforme grade IV are oncolytic adenoviruses (Jiang et al., 2006; Nandi and Lesniak, 2009), and the combination of adenoviruses expressing TK and Flt3L (Candolfi et al., 2009).
The applications of adenovirus-mediated gene transfer into the brain are diverse. For many applications, long-term expression of the transgene of interest with minimal vector-associated toxicity (immune-mediated or direct vector-mediated) is desirable. The following parameters, in addition to those discussed above, can affect the longevity, stability, and/or toxicity of adenovirus vector-mediated transgene expression:
The balance between achieving high-level, widespread and sustained transgene expression is precarious with most vector backbone–promoter–transgene combinations that are easily available at the present time. Injection of 108 to 109 infectious units of adenovirus vector into the adult rat caudate generally results in extensive transgene expression throughout the ipsilateral striatum; however this is transient, with little expression remaining by 30 days post-vector-administration (see Figure 4.24.7). The eventual loss of transgene expression from high doses of vector is likely to be a consequence of direct vector-mediated acute cytotoxicity caused by infection of neuronal cells at high multiplicities of infection, and also the elicitation of severe chronic inflammation (Thomas et al., 2000). Administration of 106 to 107 infectious units generally results in more spatially restricted, but prolonged, transgene expression (Figure 4.24.7), with accompanying inflammation that is transient and does not decrease transgene expression. Administration of fewer than 106 infectious units of vector results in very little or no transduction (unless a very strong promoter is used; see discussion below and Fig. 4.24.11). All vector preparations should therefore be accurately titered for infectious units before in vivo use, as described in UNIT 4.23. It is also necessary to determine the concentration of physical particles and the ratio of particles to infectious units for each batch of vector (also described in UNIT 4.23). Since vector titers may vary across different laboratories, each individual lab should determine the dose of vector resulting in optimal gene transfer, based on their own determination of titer. As discussed below, the dose for optimal long-term transgene expression will depend upon the constituents of the vector genome (i.e., vector backbone, promoter and transgene). Generally, the dose should be sufficient to obtain maximal spread of vector-mediated transgene expression in the complete absence of acute direct vector-mediated cytotoxicity. Cytotoxic lesions can be identified by the acute absence of GFAP (glial fibrillary acidic protein: astrocytes) or NeuN (neuronal nuclei) immunoreactivity within the region corresponding to the area of viral spread (see Fig. 4.24.12), or through the appearance of reactive astrogliosis. The severity of the lesion increases with increasing viral dose (Thomas et al., 2000; Zirger et al., 2006). Note that damage caused by inserting the needle will also result in GFAP and NeuN loss, restricted to the needle track.
As discussed above, long-term transgene expression is possible from first-generation adenovirus vectors only in the brains of naive animals. Peripheral immunization with adenovirus activates a strong anti-adenoviral immune response that is able to eliminate vector-mediated expression from the brain. Since most humans will have been subjected to a peripheral infection with adenovirus, use of peripherally primed animals (see Basic Protocol 7), rather than naive animals, may provide a more clinically relevant model for neurological gene therapy applications using adenovirus vectors.
In contrast to first-generation vectors, new-generation, high-capacity adenovirus vectors injected into the rat brain parenchyma are able to sustain prolonged transgene expression even in the presence of an anti-adenoviral T cell response elicited through intradermal injection of adenovirus (Thomas et al., 2000; Barcia et al., 2006a). High-capacity adenoviral vectors are deleted of all viral genes, retaining only the minimal elements of the viral genome required for replication in vitro and packaging (Morsy and Caskey, 1999). Thus, at least at moderate doses of vector (107 pfu), the vector backbone has profound consequences upon the stability of adenoviral vector-mediated transgene expression in the brain.
Ohmoto et al. (1999) have shown that a “clean” environment reduces vector-induced inflammation and improves longevity of transgene expression.
Transgenes encoding known immunemodulatory molecules will have an effect upon the nature of the immune and inflammatory responses to the vector (Bell et al., 1996); however, other transgene products can also contribute to vector toxicity. It is therefore important to characterize the “side effects” of each different vector used in vivo. The importance of thoroughly investigating long-term as well as short-term toxicity is illustrated by Dewey et al. (1999) and Cowsill et al. (2000), who showed that coadministration of adenovirus vectors expressing HSV-1 thymidine kinase and ganciclovir to the brains of rats for glioma gene therapy caused unexpected ongoing chronic inflammation and brain cytotoxicity in long-term survivors.
The choice of promoter driving the expression of a transgene of interest will greatly affect the levels of transgene expression achievable from any given dose of adenovirus vector. Some applications of adenovirus-mediated gene transfer to the brain may require the restriction of transgene expression to within a certain subpopulation of neural cells. This can be achieved by “transcriptional targeting” through the use of cell-type-specific promoters. Cell-type-specific promoters are generally much weaker than the strong “promiscuous” promoters derived from viruses—e.g., cytomegalovirus (CMV) or Rous Sarcoma virus (RSV)—which are able to drive high levels of transgene expression within many cell types (Smith-Arica et al., 2000). Different virus-derived “strong” promoters also display markedly different efficiencies. Figure 4.24.11 shows the levels of β-galactosidase expression achievable from two doses of adenovirus vector expressing the transgene under the control of the major intermediate early human CMV promoter (Ad-hCMV-lacZ; Shering et al., 1997). When the same transgene is expressed from a different vector containing the much stronger major intermediate early murine CMV promoter (Ad-mCMV-lacZ), substantial amounts of β-galactosidase expression can be detected even at the very low vector dose of 1 × 104 infectious units. No transgene expression is detectable from the same dose of Ad-hCMV-lacZ—i.e., it appears that only cells infected with this vector at a high multiplicity will produce detectable levels of transgene product (Gerdes et al., 2000). Hence, the use of very strong promoters can facilitate substantial and widespread expression of transgene at vector doses that elicit negligible inflammatory and toxic side effects. The choice of promoter is crucial to the efficiency of viral vector-mediated gene transfer to the brain.
As previously discussed, the brain parenchyma and ventricles constitute immunologically distinct compartments, and immune priming can take place after vector infection of the CSF (but not the parenchyma). Vector administered to the striatum in large volumes may leak into the CSF; therefore, the vector preparation should be of sufficiently high titer that the required dose can be delivered in a volume not exceeding 2 to 4 µl (keeping the volume closer to 2 µl if vector titer allows). Vector administered to rat ventricles can be delivered in larger volumes, but should still be kept as close to 2 to 4 µl as possible.
Vector purity should be assessed by calculating the ratio of absorbance at 260 nm to that at 280 nm (pure adenovirus type 5 has a 260:280 ratio of 1:3; a protocol for this determination is given in UNIT 4.23). In addition to determining vector purity by optical absorbance, it is imperative to ensure that the preparation is free of contamination from both replication competent adenovirus (RCA) and lipopolysaccharide (LPS or endotoxin). Emergence of RCA during propagation of first-generation adenoviral vectors on 293 cells is common, and injection of RCA-contaminated vector into the brain can cause severe inflammation and tissue damage, as illustrated in Figure 4.24.13. Alternative cell lines to the standard 293 cell line have recently been developed to reduce the probability of generating RCA during vector propagation (Fallaux et al., 1996). LPS is a component of the cell wall of Gram-negative bacteria and is an extremely potent stimulator of the mammalian immune system. Many studies have used injection of LPS into the brain parenchyma to investigate inflammatory responses within the CNS. Unfortunately, LPS is another frequent contaminant of adenoviral vector preparations, and it has been shown that its toxicity to primary cultures of human cells is increased in the presence of adenovirus (Cotten et al., 1994). The level at which contaminating LPS is toxic when administered in vivo with adenovirus has not been determined, but the authors do not use vector stocks containing more than 2 endotoxin units per ml for in vivo experiments.
In a nonsensitized animal, adenovirus vectors can mediate prolonged transgene expression, but levels of expression decline over time. The cause of this decline is not clear, but a putative mechanism may be the gradual inactivation of the viral promoters—i.e., cytomegalovirus (CMV) or Rous Sarcoma virus (RSV) elements—that are generally used to drive transgene expression in adenovirus vectors. The inactivation of viral promoters used in genetically modified skin fibroblasts transplanted in vivo has been documented (Palmer et al., 1991). In further support of the hypothesis of promoter inactivation, adenoviral vector genomes have been detected in transduced muscle tissue long after transgene expression has disappeared (Chen et al., 1999). Use of CNS-specific housekeeping promoters may ameliorate shut-down of transgene expression, and, indeed, use of such promoters confers the additional advantage of targeting expression to a specific subpopulation of cells within the CNS. Such promoters include the glial fibrillary-acidic protein (GFAP) promoter specific for astrocytes, the neuron-specific enolase (NSE) promoter specific for neurons, and the myelin basic protein (MBP) promoter for targeting oligodendrocytes (Morelli et al., 1999; Smith-Arica et al., 2000). Importantly, most cell-type specific promoters used to date are weaker than most virus-derived promoters (Smith-Arica et al., 2000). Other factors that may explain the diminution of transgene expression over time include cell turnover and direct toxicity of viral or transgene products to infected cells.
Poor surgical technique can generate lesions within the brain that will jeopardize the experiment and will certainly affect the quality of the data. It is therefore essential to practice the procedure on dead animals to gain competence before starting an experiment. The Hamilton needle should be examined under the microscope before using, as bent needles can also cause lesions within the brain. For very precise injections, it may be desirable to grind the tip of the needle to remove the bevel, or to use superfine glass needles, fixed with wax to the tip of the Hamilton needle.
When setting up a stereotactic apparatus, it is advisable to perform preliminary experiments injecting adenovirus expressing β-galactosidase, to ensure that the correct area of the CNS is targeted using a particular set of coordinates. The animal can be sacrificed 2 to 5 days after surgery and brain sections analyzed by Xgal histochemistry.
Adenovirus vectors injected into the brain parenchyma will infect neurons, astrocytes, and microglial cells surrounding the injection site. In addition, vector taken up by neuronal axons in the injection site will be transported in a retrograde manner to neuronal cell bodies in distant sites. This phenomenon can be exploited to specifically transduce neuronal populations in areas of the brain which are difficult to target by direct injection. Such delivery confers the additional advantage of avoiding undesirable side effects associated with tissue damage at the site of interest. Retrograde transport of vector is particularly suited to studies of neurodegeneration in rat models of Parkinson’s disease (see UNIT 9.4); adenovirus vectors encoding neurotrophic factors can be injected into the striatum, from where they will be transported along dopaminergic neuron axons to the substantia nigra. The efficiency of this procedure should be tested with each new vector, and in each neuronal system.
It should be possible to generate good-quality intact Vibratome sections: 40 to 60 µm from properly fixed material using a high-quality Vibratome. Underfixed tissue will disintegrate during sectioning. Refixing the brains for 2 to 3 hr in 4% paraformaldehyde will generally ameliorate the problem; however, it is important not to overfix the material as this will compromise the quality of the subsequent samples for immunohistochemistry and PCR. Lesions within the brain also tend to disintegrate during sectioning; in these cases it may be necessary to cut thicker sections in order to preserve the integrity of the lesioned area.
Problems encountered during immunohistochemical tissue staining generally fall into two categories: absence of staining or high levels of nonspecific background staining. Each primary antibody used should be titrated to determine the dilution required for optimal staining, using tissue sections that are known to be positive for the epitope of interest (these sections may be derived from other organs such as spleen or liver). If a range of optimal dilutions is not recommended by the supplier, then try a range of dilutions from 1:100 to 1:2000 as a starting point. When performing immunohistochemistry on brain sections, it is expedient to include a positive control for each marker under investigation to ensure that absence of staining within the brain tissue is not the result of omission of one of the reagents. Fixation or tissue processing (e.g., paraffin embedding) can mask some antigenic processes, particularly those on the cell surface. Incubation of sections in 0.01 M sodium citrate (pH 7.6) for one minute at 60°C (heated in a microwave) can retrieve some epitopes without severe tissue damage.
If the optimal dilution of each antibody has been determined in a preliminary experiment, the levels of background staining should be very low. High background staining may result from insufficient washing, or, more likely, from overdeveloping with the DAB reagent. Since the color reaction proceeds rapidly, it is advisable to stain each well of brain sections individually (if multiple samples are being processed simultaneously) and to carefully monitor color development by eye (rather than with a timer) to avoid overstaining.
The results of adenovirus vector-mediated gene transfer into the brain will depend upon the experimental design and the factors discussed in Critical Parameters. Several figures illustrate examples of results obtained with different doses of vector (Fig. 4.24.7, Fig. 4.24.9, Fig. 4.24.11, and Fig. 4.24.12), different promoters (Fig. 4.24.11), and vector stocks of variable purity (Fig. 4.24.13).
An in vivo experiment involving stereotactic delivery of adenovirus to the brain is a major undertaking which requires considerable preparation and labor input. The duration of the experiment and the time taken to generate data will depend upon the questions being posed and the number of animals required to answer the question in a statistically significant manner. Generation and titration of the viral vectors requires several weeks, and all the quality-control tests (e.g., LPS/RCA assays) must be completed for each batch of vector before in vivo administration. Stereotactic delivery of the adenoviral vector is in itself a labor-intensive process. With practice, the surgical procedure should take no more than 40 min per animal, and hence several animals can be operated upon in 1 day. The time required to process the data is again dependent upon the number of animals involved in the experiment and the type of analysis undertaken.
This work is supported by National Institutes of Health/National Institute of Neurological Disorders and Stroke (NIH/NINDS) Grant 1R01 NS44556.01, Minority Supplement NS445561.01; 1R21-NSO54143.01; 1UO1 NS052465.01, R01NS057711-01A2 to M.G.C.; NIH/NINDS Grants 1 RO1 NS 054193.01; R01NS061107-02, U54 NS045309-01, and 1R21 NS047298-01 to RR.L.; the Bram and Elaine Goldsmith and the Medallions Group Endowed Chairs in Gene Therapeutics to RR.L. and M.G.C., respectively; and the Linda Tallen & David Paul Kane Foundation Annual Fellowship and the Board of Governors at CSMC. The authors would like to acknowledge Mr. Ivan Di Stefano for the photographic work done for this manuscript.