Optimization of Adeno-Associated Viral Vector-Mediated Gene Delivery to the Hypothalamus

Abstract

To efficiently deliver genes and short hairpin RNAs to the hypothalamus we aimed to optimize the transduction efficiency of adeno-associated virus (AAV) in the rat hypothalamus. We compared the transduction efficiencies of AAV2 vectors pseudotyped with AAV1, AAV8, and mosaic AAV1/2 and AAV2/8 coats with that of an AAV2 coated vector after injection into the lateral hypothalamus of rats. In addition, we determined the transduction areas and the percentage of neurons infected after injection of various titers and volumes of two AAV1-pseudotyped vectors in the paraventricular hypothalamus (PVN). Successful gene delivery to the hypothalamus was achieved with AAV1-pseudotyped AAV vectors. The optimal approach to transduce an area, with the size of the PVN, was to inject 1 × 10⁹ genomic copies of an AAV1-pseudotyped vector in a volume of 1 μl. At a radius of 0.05 mm from the injection site almost all neurons were transduced. In addition, overexpression of AgRP with the optimal approach resulted in an increase in food intake and body weight when compared with AAV-GFP.

Introduction

D ysregulation of energy balance involves genetic, environmental, and cognitive factors and may result in obesity or eating disorders (de Krom et al., 2009). These disorders are associated with substantial morbidity, mortality, and socioeconomic burden. Although genes expressed in the hypothalamus have repeatedly been shown to contribute to obesity (Bolze and Klingenspor, 2009; Renström et al., 2009; Shafrir and Ziv, 2009; Willer et al., 2009), gene therapy thus far has hardly been considered as a therapeutic strategy (Cao et al., 2009). Animal studies with spontaneous gene mutations, which resulted in obesity and/or diabetes, have greatly contributed to our knowledge about the role certain genes play in energy balance (Shafrir and Ziv, 2009). The importance of genes for regulation of energy balance was confirmed by knockout mice, although these mice may display the capacity to compensate for the lack of that gene during development rather than the function of a certain gene in the adult stage. One of the neuropeptides involved in feeding is Agouti-related protein (AgRP) (Graham et al., 1997). AgRP is a peptide produced in the arcuate nucleus of the hypothalamus and is secreted in many brain regions, including multiple hypothalamic nuclei such as the paraventricular nucleus (PVN) and the lateral hypothalamic nucleus (LH) (Broberger et al., 1998; Haskell-Luevano et al., 1999). In the projection areas, AgRP can bind the melanocortin receptors and reduce their activity. Reduction of melanocortin signaling is suggested to induce hyperphagia and lower energy expenditure, resulting in obesity in rodents and humans (reviewed in Farooqi and O'Rahilly, 2006; Bolze and Klingenspor, 2009).

In view of the development of gene therapy to modify gene expression in the hypothalamus, it is important to determine the local effect of genes in the hypothalamus. We have chosen to use adeno-associated viral (AAV) vectors for local overexpression and knockdown of genes in the brain, because AAV vectors transduce nondividing neurons and gene cassettes are expressed for long periods of time, up to 25 months in rats (Klein et al., 2002). AAV vectors have been reported to transduce specific cell types in vitro and in vivo by using cell type-specific promoters and/or by packaging the AAV DNA in various serotype capsids or in capsids that have been genetically or chemically altered (Shevtsova et al., 2005; Büning et al., 2008; Cearley et al., 2008; Howard et al., 2008; Kwon and Schaffer, 2008). AAV serotype 2 (AAV2) vectors were the first to be used for gene transfer (Hermonat and Muzyczka, 1984). Since the 1980s many other serotypes have been identified and investigated for their cell tropism (Cearley et al., 2008; Kwon and Schaffer, 2008). Until now, the most widely used AAV serotype to modify gene expression in the hypothalamus has been AAV2 (Garza et al., 2008; McCrimmon et al., 2008; Yang et al., 2009). However, studies have shown that serotypes other than AAV2 are more efficient in transducing rat primary cortical and hippocampal cultures (Howard et al., 2008; Royo et al., 2008). In addition, AAV1- and AAV8-coated vectors have been shown to transduce more neurons in vivo, for example, in the rat striatum (Burger et al., 2004; Reimsnider et al., 2007), hippocampus (Burger et al., 2004; Klein et al., 2006, 2008), and midbrain (Burger et al., 2004; McFarland et al., 2009).

Therefore, in this study, we aimed to characterize how various capsids affect AAV-mediated transduction of selected areas in the hypothalamus. We packaged an AAV-CBA-GFP-WPRE vector, containing AAV2 inverted terminal repeats (ITRs), with single AAV1 and AAV8 capsids or with mosaic capsids AAV1/2 and AAV2/8. This was done to compare the transduction efficiencies of AAV1 and AAV8, which most often show higher transduction efficiencies than AAV2, with the “gold standard” AAV2 after injection into the lateral hypothalamus. The mosaic vectors were tested because some studies have shown that a combination of two serotype capsids may confer (synergistic) effects of both parental strains (Hauck et al., 2003; Rabinowitz et al., 2004). In addition, we investigated the area of transduction after administration of various titers (1 × 10⁸, 1 × 10⁹, and 1 × 10¹⁰ genomic copies [GC]) and volumes (1 vs. 0.3 μl) of AAV1-coated vectors aimed at the paraventricular nucleus (PVN) of the hypothalamus. From these results we obtained an indication of the optimal titer and volume to transduce a nucleus of a certain size in the hypothalamus. Our major findings were that AAV1-pseudotyped AAV vector was more efficient than AAV vectors pseudotyped with AAV2, AAV8, AAV1/2, and AAV2/8 in neuronal transduction of the LH. In addition, a titer of 1 × 10⁹ GC, in a volume of 1 μl, transduced an area of 0.5 mm² successfully. At a distance of 0.05 mm from the injection site almost all neurons were transduced when AAV1-pseudotyped AAV vectors were used. As a proof of principle of gene delivery to induce a phenotype we injected AAV1-AgRP into the PVN.

Materials and Methods

Cell lines and plasmids

Human embryonic kidney (HEK) 293T cells were maintained at 37°C with 5% CO₂ in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and penicillin–streptomycin.

The helper plasmids used to produce AAV vectors were pDP1, pDP2, and pAR-8 in combination with pXX6. The pDP1 and pDP2 helper plasmids have been described (Grimm et al., 2003) and were obtained from PlasmidFactory (Bielefeld, Germany). The packaging plasmid pAR-8 and the transgene vector pAAV-CBA-GFP-WPRE, with AAV2 inverted terminal resolution sites, were both previously described (Broekman et al., 2006) and were a kind gift from M. Sena-Esteves (University of Massachusetts Medical School, Worcester, MA). pAR-8 lacks the adeno-helper genes, and therefore pXX6 is necessary as an additional helper plasmid. pXX6 has been described (Xiao et al., 1998). We constructed pAAV-CBA-AgRP-ires-gfp by removing the gene encoding green fluorescent protein (GFP) from pAAV-CBA-GFP-WPRE with the restriction enzymes AgeI and BsrGI. Subsequently, agouti-related protein (AgRP)-ires-GFP was isolated from pIRES2-AgRP with AfeI and BsrGI, and was ligated to pAAV-CBA-GFP-WPRE, from which the GFP gene was removed. This ligation generated pAAV-CBA-AgRP-ires-GFP.

AAV production

To produce AAV-GFP with single capsids of AAV1, AAV2, and AAV8, we cotransfected pAAV-CBA-GFP-WPRE with the individual helper plasmid pDp1, pDp2, or pAR-8 + pXX6, at a molar ratio of 1:1 (AAV vector plasmid:helper plasmid). Mosaic coated AAV-GFP with AAV1/2 and AAV2/8 capsids were produced by mixing the packaging vectors (pDp1 or pAR-8 with pDp2) at a molar ratio of 1:1. Subsequently, pAAV-CBA-GFP-WPRE was cotransfected, at a 1:1 molar ratio, with either the pDP1 + pDP2 or pAR-8 + pDP2 mix. pAAV-CBA-AgRP-ires-GFP was cotransfected only with pDP1. Each AAV production was performed with 15 dishes of 80–90% confluent 293T cells on the day of transfection. Two hours before transfection, the 10% FCS–DMEM was replaced with 2% FCS–DMEM. The transfections were performed with polyethylenimine (PEI) as described by Reed and colleagues (2006). The transfection mix remained on the cells until the next day, and then the 2% FCS–DMEM was refreshed. The purification was performed as described by Zolotukhin and colleagues (2002). Briefly, 60 hr after transfection, the cells were harvested in their medium, centrifuged, and washed with phosphate-buffered saline (PBS) containing 5 mM ethylenediaminetetraacetic acid (EDTA). Last, the cells were collected in 12 ml of ice-cold lysis buffer (150 mM NaCl, 50 mM 2-amino-(hydroxymethyl)-1,3-propanediol [Tris]; pH 8.4) and stored at −20°C until further use. Usually after 3 days the cells were freeze-thawed twice, incubated for 30 min with Benzonase (50 units/ml; Sigma, Zwijndrecht, The Netherlands) at 37°C and centrifuged. After centrifugation, the supernatant was loaded onto an iodixanol gradient (60, 40, 25, and 15%, supernatant) (OptiPrep; Lucron Bioproducts, Sint-Martens-Latem, Belgium) in Quick-Seal tubes (Beckman Coulter, Woerden, The Netherlands). After 1 hr of ultracentrifugation (70,000 rpm at 20°C) in a Ti70 rotor (Beckman Coulter), the 40% layer was extracted. This 40% layer was used for ion-exchange chromatography with 5-ml HiTrap Q HP columns (GE Healthcare, Brussels, Belgium). The AAV-positive fractions, determined by polymerase chain reaction (PCR), were pooled and concentrated on Centricon Plus-20 Biomax-100 concentrator columns (Millipore, Amsterdam, The Netherlands). The titer, in genomic copies per milliliter (GC/ml), was determined by quantitative PCR (qPCR) with SYBR green mix in a LightCycler (Roche, Mannheim, Germany) (Veldwijk et al., 2002). The qPCR primers were designed to detect the bovine growth hormone (BGH) poly(A) signal and included BGHpolyA_F (5′-CCTCGACTGTGCCTTCTAG) and BGHpolyA_R (5′-CCCCAGAATAGAATGACACCTA).

Animals

Male Wistar rats (strain Crl:WU), weight ranging from 220 to 250 g, were purchased from Charles River (Sulzfeld, Germany). All rats were individually housed in filtertop cages with ad libitum access to food (CRM pellets; Special Diet Services, Whitham, Essex, UK) and water. Animals were kept in a temperature- and humidity-controlled room (21 ± 2°C) with a 12-hr light/dark cycle (lights on at 7:00 a.m.). All experimental procedures were approved by the Committee for Animal Experimentation of the University of Utrecht (Utrecht, The Netherlands).

Surgical procedures

In study 1 the transduction efficiency of various capsids surrounding AAV-CBA-GFP-WPRE after injection into the lateral hypothalamus was studied. Twenty rats were anesthetized by intramuscular injection of fentanyl–fluanisone (Hypnorm, 0.1 ml/100 g; Janssen Pharmaceutica, Beerse, Belgium) and intraperitoneal injection of midazolam (Dormicum, 0.05 ml/100 g; Hoffman-LaRoche, Mijdrecht, The Netherlands). After this, the rats were injected bilaterally via the lateral hypothalamus (LH) (coordinates: anteroposterior [AP], −1.8; mediolateral [ML], +1.7; dorsoventral [DV], −8.6) with 1.5 μl of AAV-CBA-GFP-WPRE coated with, respectively, AAV2, AAV1, AAV8, AAV1/2, and AAV2/8. The injected volume contained 1 × 10⁸ GC of AAV vector and was delivered at a rate of 0.2 μl/min. After the injection the needle remained in the injection site for 10 min. After surgery, for postoperative pain control the rats received carprofen (5 mg/kg; Vericore, Dundee, UK). Four weeks after injection the rats were anesthetized and perfused with 4% paraformaldehyde (PFA) containing 0.05% glutaraldehyde. After perfusion the brains were isolated and were placed in 4% PFA overnight at 4°C. The next morning the brains were placed in PBS and stored at 4°C until further use. The perfused brains were sectioned (thickness, 40 μm) with a Vibratome in series of 10. The first and sixth series of sections were combined and used for double immunohistochemistry against GFP and the neuronal marker NeuN (neuronal nuclei).

In study 2 the effect of various titers of AAV1-coated AAV vectors on transduction efficiency was explored. Twenty-four rats received bilaterally 1 μl of AAV-CBA-GFP-WPRE or AAV-CBA-AgRP-ires-GFP coated with AAV1, aimed at the paraventricular nucleus (PVN) (coordinates: AP, −1.8; ML, +1.7; DV, −8.1; angle, 10 degrees). The injections contained, respectively, 1 × 10¹⁰ GC (two rats), 1 × 10⁹ GC (two or three rats), or 1 × 10⁸ GC (three of four rats) of the AAV vectors. Each rat was injected via both sides of the brain and these sides were analyzed separately. After 4 weeks the rats were decapitated and the brains were removed, quickly frozen on dry ice, and stored at −80°C until they were sectioned with a cryostat (Leica, Rijswijk, The Netherlands) at 20 μm. The sections were used for in situ hybridization with GFP–digoxigenin-labeled mRNA probe to show the transduction area of the AAV vectors.

In study 3 the effect of an injection volume smaller than 1 μl, of AAV1-coated vectors, was investigated. Twenty rats were injected with 0.3 μl of AAV-CBA-GFP-WPRE or AAV-CBA-AgRP-ires-GFP-WPRE, coated with AAV1, containing 1 × 10⁹ GC (three rats), 5 × 10⁸ GC (four rats), or 1 × 10⁸ GC (three rats) aimed at the PVN (coordinates: AP, −1.8; ML, +1.7; DV, −8.1; angle, 10 degrees). After 31 days the animals were decapitated and processed as described in study 2.

Immunohistochemistry

Sections (thickness, 40 μm) from study 1 were used for GFP–NeuN staining (series 1 and 6 were pooled). The free-floating sections were washed three times with PBS, permeabilized for 30 min in PBS supplemented with 0.5% Triton X-100 at room temperature, blocked for 1 hr in PBS with 1.5% normal goat serum (NGS) at room temperature, and incubated overnight in PBS supplemented with mouse monoclonal anti-NeuN (diluted 1:2000; Chemicon/Millipore, Temecula, CA), rabbit polyclonal anti-GFP (diluted 1:1500; Invitrogen, Carlsbad, CA), and 1.5% NGS at 4°C. The next morning sections were washed three times for 10 min with PBS and incubated for 1 hr with secondary antibodies: Alexa 555-conjugated goat anti-mouse (diluted 1:1000) and Alexa 488-conjugated goat anti-rabbit (diluted 1:1000) (both from Invitrogen) in 1.5% NGS at room temperature. After washing three times (10 min each) with PBS, the sections were transferred to microscope slides and kept overnight in the dark to dry. All sections were embedded in 90% glycerol and stored flat at 4°C.

In situ hybridization

The brains from study 2 and 3 were sectioned (thickness, 20 μm) with a cryostat and were used for in situ hybridization with a 720-base pair (bp)-long digoxigenin (DIG)-labeled enhanced green fluorescent protein (eGFP) riboprobe (antisense to NCBI gene DQ768212). The in situ hybridization was performed as described by Schaeren-Wiemers and Gerfin-Moser (1993) with small modifications in the fixation procedure and hybridization temperature. Briefly, sections were fixed in 4% PFA for 20 min and rinsed three times for 3 min each in PBS. After acetylation for 10 min (0.25% acetic anhydride in 0.1 M triethanolamine), the sections were washed three times in PBS for 5 min each and prehybridized at room temperature in hybridization solution, containing 50% deionized formamide, 5 × saline–sodium citrate (SSC), 5 × Denhardt's solution, baker's yeast tRNA (250 μg/ml), and sonicated salmon sperm DNA (500 μg/ml). After 2 hr 150 μl of hybridization mixture containing DIG-labeled riboprobe (400 ng/ml) was applied per slide, covered with Nescofilm (Karlan, Cottonwood, AZ), and hybridized overnight at 72°C. The next morning the slides were quickly washed in 2 × SSC followed by 0.2 × SSC for 2 hr; both wash steps were performed at 72°C. DIG was detected with an alkaline phosphatase-labeled antibody (diluted 1:5000; Roche), using nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (NBT/BCIP; Sigma) as a substrate. After overnight incubation at room temperature with the NBT/BCIP mixture, the sections were quickly dehydrated in ethanol, cleared in xylene, and mounted using Entellan (Electron Microscopy Sciences, Hatfield, PA).

Imaging, quantitation, and statistical analysis

The MCID system (MCID, Linton, Cambridge, UK) was used to digitize pictures from sections containing the injection site and a section 400 μm more posterior. The images were analyzed blind with ImageJ (National Institutes of Health, Bethesda, MD; http://rsb.info.nih.gov/ij/). This program was used to transform the digitized, color pictures into binary (black and white) pictures (Fig. 2A and 2B), the scale was set, and total black area was measured. The total black area was used as a value for the total area that was transduced by the different serotypes in one section (Fig. 1B). In addition, ImageJ was used to obtain an indication of the percentage of area transduced at a certain distance from the injection site, as a measure of spread from injection site (Fig. 1B). This was done by using the macro: measure and label tool circle. Circles of different radiuses were placed at the site of injection and total black area and percent black area in a certain-sized circle were measured. From this the percentage transduced per circle radius was calculated.

FIG. 1.

Transduction by AAV1-GFP or AAV2-GFP in vivo. (A) Colocalization of GFP (green) and NeuN (red) at the injection site, 4 weeks after injection with 1 × 10⁸ GC of AAV1-GFP. (B) Colocalization of GFP and NeuN after injection of titer-matched AAV2-GFP. Color images available online at www.liebertonline.com/hum.

FIG. 2.

Effect of AAV serotype capsids on transduction efficiency in the lateral hypothalamus. (A) GFP immunostaining of a 40-μm slice containing the lateral hypothalamus, 4 weeks after injection with AAV1-GFP at a titer of 1 × 10⁸ GC in 1.5 μl. (B) A binary picture made with ImageJ through conversion of the original picture (Fig. 2A) to a binary picture. Subsequently, ImageJ was used to measure the total area that is GFP positive (black). (C) Total area transduced (in mm²) by AAV-GFP coated with various serotype capsids. (D) A graphical representation of the relation between percentage transduced area and the number of neurons transduced in an area of 0.115 mm² by AAV1-GFP (closed circles) and AAV2-GFP (open squares). This area contains approximately 19 or 20 neurons. (E) The percentage of area transduced at certain radii from the injection site. *Significant difference compared with AAV2-GFP, p < 0.05. Color images available online at www.liebertonline.com/hum.

SPSS 15.0 (SPSS, Chicago, IL) was used for statistical analysis. The total area transduced and percentage transduced area per circle were compared by analysis of variance (ANOVA). The post-hoc Bonferroni correction was used when three or more groups were compared.

Results

AAV1 transduces the LH most efficiently

To compare the efficiency of various AAV serotypes in transducing the LH, we packaged pAAV-GFP with capsids derived from serotypes 1, 2, and 8 or combined packaging plasmids to obtain mosaic capsids AAV1/2 and AAV2/8. These pseudotyped AAV vectors were bilaterally injected into the LH of rats (n = 2–4) with 1 × 10⁸ GC in a volume of 1.5 μl. Free-floating brain sections (thickness, 40 μm) were used for double immunohistochemistry against GFP (green signal) and the neuronal marker NeuN (red signal). Colocalization of GFP and NeuN showed that all AAV serotypes transduced primarily neurons (>95%), because there were only a few GFP-stained cells that lacked staining for the neuronal marker NeuN (Fig. 1A and B).

The area transduced by AAV vectors was quantified with imaging software (Fig. 2A and B). Analysis of brains injected with AAV-GFP coated with various serotypes and injected into the LH revealed that the total area transduced was 0.025 mm² with AAV2-GFP. The total area transduced with serotypes 8, 1/2, and 2/8 was not significantly different from that of AAV2 (respectively, 0.05, 0.03, and 0.005 mm²; p > 0.05) (Fig. 2C). However, AAV1-GFP showed a significantly larger total transduction area than AAV2-GFP (0.312 mm²; p = 0.042). In addition to the total area of transduction, we also investigated whether the percentage of transduced cells at various radii from the injection site was different for the different AAV serotypes. The percentage of the area transduced was positively correlated with the number of neurons transduced (Fig. 2D). The results in Fig. 2E show that serotypes 2, 8, and 1/2 transduced approximately 50% of the neurons at a distance of 0.05 mm from the injection site at a titer of 1 × 10⁸ GC. AAV2/8-GFP transduced an area of only approximately 20% at a distance of 0.05 mm. Again, AAV1-GFP appeared to be superior, transducing almost 90% of the neurons at 0.05 mm. In addition, AAV1-transduced neurons could be found at distances further from the injection site, namely up to 1 mm instead of 0.5 mm (Fig. 2E). AAV1 also showed a significantly larger spread in the rostral-to-caudal direction. Ten sections (∼2 mm; p = 0.002 compared with AAV2) were positive for GFP protein in the AAV1 group compared with 4, 3, 5, and 6 sections for AAV2, AAV8, AAV1/2, and AAV2/8, respectively (p > 0.05 for AAV8, AAV1/2, and AAV2/8 compared with AAV2). These results together suggest that AAV1-coated AAV vectors are more efficient in transducing neurons after injection into the LH than AAV2, AAV8, or the mosaic AAV1/2- and AAV2/8-coated AAV vectors.

Effect of titer on transduction area and on percentage transduction

To investigate the effect of various titers, we tested two AAV genomes packaged with AAV1 capsids, namely, pAAV-GFP and pAAV-AgRP. These AAV vector preparations were diluted in PBS to obtain 1 × 10⁸, 1 × 10⁹, and 1 × 10¹⁰ GC/μl. One microliter of these diluted preparations was bilaterally injected into the PVN. As expected, an increase in titer of AAV1-AgRP (Fig. 3A) and AAV1-GFP (data not shown) augmented the total transduction area as determined by GFP in situ hybridization (ISH). Increasing the amount of viral particles from 1 × 10⁸ to 1 × 10⁹ GC of AAV1-AgRP resulted in a total transduction area that was eight times larger; it increased from 0.08 to 0.65 mm² (p = 0.004). An additional 10-fold rise in total viral particles, from 1 × 10⁹ to 1 × 10¹⁰ GC, did not transduce a significant larger area than with 1 × 10⁹ GC. The total transduced area increased only from 0.65 to 0.77 mm² (p > 0.05). Figure 3B shows that the percentage of transduced area at a certain distance from the injection site with 1 × 10⁹ GC was not significantly different from that with 1 × 10¹⁰ GC, whereas results with both 1 × 10⁹ and 1 × 10¹⁰ GC were significantly different from the result with 1 × 10⁸ GC (Fig. 3B). The total area transduced and the percentage transduced by AAV1-GFP (Fig. 3C) did not significantly differ from the results with AAV-AgRP at comparable titers (1 × 10⁹, p > 0.05; 1 × 10¹⁰, p > 0.05).

FIG. 3.

Effects of various titers of AAV1-coated vectors on transduction efficiency in the paraventricular hypothalamus. (A) Total area transduced (in mm²) by AAV1-AgRP at various titers in the rat PVN. (B) Percentage transduced area per circle with AAV1-AgRP. (C) Percentage transduced area per circle with AAV-GFP. *Significant difference compared with AgRP at 1 × 10⁸ GC.

Effects of injection volume on transduction area and percentage transduction

In addition to variation in titer, we also varied the injection volume to investigate the effect on transduction area. We injected AAV1-GFP and AAV1-AgRP at similar genomic copy numbers (1 × 10⁸ and 1 × 10⁹ GC), but in different end volumes (1 and 0.3 μl), into the PVN. As can be seen in Fig. 4A and B, the injection volume of 0.3 μl resulted in significantly smaller transduction areas (p = 0.029). In addition, 0.3 μl resulted in a lower percentage transduced area per circle than injections with similar amounts of viral particles, but in volumes of 1 μl. We observed more often a GFP-positive injection track in 0.3-μl groups than in the 1-μl groups (Fig. 4C and D).

FIG. 4.

Effects of various volumes of AAV1-coated vectors on transduction efficiency in the paraventricular hypothalamus. (A) Total transduced area with injection volumes of 0.3 and 1 μl. Both volumes contained 1 × 10⁹ GC of AAV1-AgRP. (B) Percentage area transduced per circle with AAV1-AgRP. (C) GFP ISH picture with injection volume of 0.3 μl. (D) GFP ISH picture with injection volume of 1 μl. The 0.3-μl injection volume more often resulted in GFP-positive injection tracks than did 1-μl injection volumes. Color images available online at www.liebertonline.com/hum.

Behavioral effects of AAV-AgRP overexpression in the PVN

Injection of AAV1-AgRP and AAV1-GFP at the optimal dose of 1 × 10⁹ GC, in 1 μl, into the PVN resulted in an increase in body weight in the AAV1-AgRP rats compared with AAV1-GFP rats. On day 28 AAV1-AgRP rats had accumulated 147 ± 5.4% (p = 0.0002) more weight than did the AAV1-GFP rats (Fig. 5A). The increased body weight of AAV1-AgRP rats was at least partly caused by an increase in daily caloric intake in the AAV1-AgRP rats (p = 0.0001) (Fig. 5B).

FIG. 5.

Effects of AAV1-AgRP on body weight (BW) and caloric food intake. (A) Cumulative body weight gain of rats receiving AAV1-AgRP or AAV1-GFP in the PVN. The body weight gain is presented from day 0 until day 28 after injection. Both vectors were injected at a titer of 1 × 10⁹ GC in 1 μl. (B) Total caloric food intake per day 1 day before injection and 28 days after injection of AAV1-AgRP or AAV1-GFP into the PVN. Significant difference compared with AAV1-GFP: **p < 0.01; ***p < 0.001.

Discussion

Studies have shown that AAV vectors coated with serotypes other than AAV2 are more efficient in the transduction of several brain areas, such as striatum, hippocampus, and substantia nigra (Burger et al., 2004; Klein et al., 2006, 2008; Reimsnider et al., 2007; McFarland et al., 2009). However, most studies using AAV vectors for transduction of the hypothalamus have been conducted with AAV2-coated vectors (Garza et al., 2008; McCrimmon et al., 2008; Yang et al., 2009). Therefore, we compared the transduction efficiency of AAV1, AAV8, or mosaic capsids AAV1/2 and AAV2/8 with that of AAV2 in the hypothalamus. We injected AAV-GFP coated with various capsid proteins into the LH at a titer of 1 × 10⁸ GC. The results showed that AAV1-GFP transduced a significantly larger total area than AAV2. However, AAV-GFP coated with AAV8, AAV1/2, and AAV2/8 did not significantly differ from AAV2-GFP in the total area transduced after injection into the LH. In addition, AAV1-GFP transduced more neurons at a certain radius from the injection site in comparison with the other serotypes (AAV2, AAV8, AAV1/2, and AAV2/8); AAV1 also showed a significantly larger dispersion in the rostral-to-caudal direction. It is possible that the mosaic capsids transduced a different set of neurons; however, we did not investigate this. We cannot exclude that other AAV serotypes, such as AAV5, AAV6, AAV9, and AAV10, are more efficient than AAV1 in transduction of the hypothalamus. To our knowledge this is the first study to determine the efficacy of various AAV serotypes in the adult hypothalamus.

In this study, we determined which titer was required to transduce a brain area of a certain size. We injected two vectors, namely AAV1-GFP and AAV1-AgRP, at various titers (1 × 10⁸, 1 × 10⁹, and 1 × 10¹⁰ GC in 1 μl), aimed at the PVN. These results showed that a titer of 1 × 10⁹ GC was able to transduce an area of approximately 0.5 mm². There were some differences in total transduction areas between AAV1-GFP and AAV1-AgRP. AAV1-AgRP at 1 × 10⁹ GC transduced a total area of 0.65 mm² (SEM, ±0.07) whereas AAV1-GFP at 1 × 10⁹ GC transduced 0.38 mm² (SEM, ±0.18); at 1 × 10¹⁰ GC the transduction areas were 0.77 mm² (SEM, 0.24) for AAV1-AgRP and 1.08 mm² (SEM, 0.09) for AAV1-GFP. However, these differences were not significant. A reason for this variation may be that there is a difference in the ratio of transducing particles and genomic particles between the different AAV preparations. We determined the titers by qPCR and this resulted in a value representing the amount of genomic particles. The determination of transducing particles was not feasible, because the receptors and pathways by which AAV1, AAV8, and other serotypes enter cells are still (largely) unknown. The mechanisms for uptake and trafficking of AAV to the cell nucleus are probably different for each serotype (Vihinen-Ranta et al., 2004; Akache et al., 2006). In addition, Rabinowitz and colleagues demonstrated that one dose of AAV vector coated with AAV1/2 yielded different percentages of GFP expression in different cell lines (Rabinowitz et al., 2004). We also confirmed that different serotypes with similar genomic particles resulted in different transducing particle values in HEK293T, HeLa, and HT1080 cell lines (data not shown). Thus, comparison of different AAV-serotyped AAV vectors in cell lines most probably does not reflect transduction efficiency in the brain and therefore we determined GC values for comparison between different coats. However, transduction of a cell line with AAV vectors can be used to determine whether different batches of AAV vectors, with identical serotype capsids, have comparable ratios of transducing particles to genomic particles; this to ensure consistent AAV production over time.

When the titer of AAV1-AgRP and AAV1-GFP was increased from 1 × 10⁹ to 1 × 10¹⁰ GC we did not see a significant increase in the total area transduced. This ceiling effect was previously reported by Klein and colleagues (2002). They showed that AAV-NSE-GFP (AAV2 coat) at a titer of 3 × 10¹⁰ particles was not significantly different from a titer of 5 × 10⁹ particles. In addition, they demonstrated that a minimal dose of 5 × 10⁷ AAV particles was necessary to obtain GFP expression in rat hippocampus (Klein et al., 2002). Together, these data suggest that injection of different amounts of AAV genomic particles, to transduce a brain region, can be described by an S-shaped dose–response curve.

Besides the effect on the total area transduced, we also investigated the effect of the titer on the percentage of neurons that were transduced at a certain radius from the injection site. The results showed that with an increase in titer, the percentage transduced area at a certain radius increased. This is in concordance with the larger total area that was transduced. Thus, more neurons in a certain area were transduced when the titer was increased.

We also studied the effect of different injection volumes, 1 versus 0.3 μl, on the total area transduced. As expected, an injection volume of 1 μl transduced a larger total area. In addition, the 1-μl injection volume showed a trend toward a higher percentage transduced area per circle. This difference was probably due to a difference in diffusion, because the viral preparation, titer, and flow rate were comparable.

AAV-mediated AgRP overexpression in the PVN increased food intake and body weight, as would be expected from studies in the literature, which have shown an increase in food intake and body weight after acute and chronic intracerebroventricular (ICV) administration of AgRP (Small et al., 2003; Semjonous et al., 2009). In addition, overexpression of Agouti with AAV2-pseudotyped vector in the PVN also resulted in an increase in food intake and body weight (Kas et al., 2004). Similar amounts of AAV-AgRP that missed the PVN did not result in an increase in food intake or body weight (data not shown), showing that local delivery at precise locations is necessary for AAV vectors to result in a phenotype.

Taken together, the results we obtained regarding titer, volume, and type of coat to transduce a certain area in the hypothalamus are similar to those obtained in the striatum, hippocampus, and substantia nigra (Burger et al., 2004; Klein et al., 2006, 2008; Reimsnider et al., 2007; McFarland et al., 2009) and may therefore be informative for AAV-mediated transduction of other brain regions. AAV vectors are effective tools with which the exact role of specific genes and brain areas in the dysregulation of energy balance can be dissected. The further optimization of AAV technology for transduction of the hypothalamus, which we showed here, will be helpful in designing experiments in this field. Most investigators in this area of research still use AAV2-coated vectors, whereas we have shown here that an AAV1-pseudotyped vector transduces hypothalamic neurons more efficiently. In the future, AAV-mediated gene therapy may be considered as a new approach to treat obesity (Cao et al., 2009). In particular, in view of the burden and high mortality associated with bariatric surgery and the limited efficacy that is obtained with lifestyle adaptations, such as diets and exercise (Tice et al., 2008), local injection of viral vectors to downregulate or to deliver genes to the human hypothalamus may be considered as an alternative therapeutic strategy.

In summary, this study showed that AAV1-GFP was most efficient for transduction of the lateral hypothalamus when compared with AAV2, AAV8, AAV1/2, and AAV2/8-pseudotyped vectors. We were able to transduce an area of approximately 0.5 mm² with AAV1-pseudotyped vectors at a titer of 1 × 10⁹ GC, in an injection volume of 1 μl. An area of 0.5 mm², 4–5% of the total hypothalamic area in the analyzed plane, corresponds to a brain nucleus such as the PVN. These results also imply that for transduction of larger areas, multiple injections are required with distances of 0.5–1 mm between injection sites.

Footnotes

Acknowledgments

M.W.A. de Backer was supported by the Netherlands Organization for Scientific Research (NWO grant 90339175) and by the Foundation De Drie Lichten in the Netherlands. This work was performed in Utrecht, The Netherlands.

Author Disclosure Statement

No competing financial interests exist.

References

Akache

, Grimm

, Pandey

, Yant

S.R.

, Xu

, Kay

M.A.

2006. The 37/67-kilodalton laminin receptor is a receptor for adeno-associated virus serotypes 8, 2, 3, and 9. J. Virol., 80:9831–9836.

Bolze

, Klingenspor

2009. Mouse models for the central melanocortin system. Genes Nutr., 4:129–134.

Broberger

, Johansen

, Johansson

, Schalling

, Hökfelt

1998. The neuropeptide Y/agouti gene-related protein (AGRP) brain circuitry in normal, anorectic, and monosodium glutamate-treated mice. Proc. Natl. Acad. Sci. U.S.A., 95:15043–15048.

Broekman

M.L.

, Comer

L.A.

, Hyman

B.T.

, Sena-Esteves

2006. Adeno-associated virus vectors serotyped with AAV8 capsid are more efficient than AAV-1 or −2 serotypes for widespread gene delivery to the neonatal mouse brain. Neuroscience, 138:501–510.

Büning

, Perabo

, Coutelle

, Quadt-Humme

, Hallek

2008. Recent developments in adeno-associated virus vector technology. J. Gene Med., 10:717–733.

Burger

, Gorbatyuk

O.S.

, Velardo

M.J.

, Peden

C.S.

, Williams

, Zolotukhin

, Reier

P.J.

, Mandel

R.J.

, Muzyczka

2004. Recombinant AAV viral vectors pseudotyped with viral capsids from serotypes 1, 2, and 5 display differential efficiency and cell tropism after delivery to different regions of the central nervous system. Mol. Ther., 10:302–317.

Cao

, Lin

E.J.

, Cahill

M.C.

, Wang

, Liu

, During

M.J.

2009. Molecular therapy of obesity and diabetes by a physiological autoregulatory approach. Nat. Med., 15:447–454.

Cearley

C.N.

, Vandenberghe

L.H.

, Parente

M.K.

, Carnish

E.R.

, Wilson

J.M.

, Wolfe

J.H.

2008. Expanded repertoire of AAV vector serotypes mediate unique patterns of transduction in mouse brain. Mol. Ther., 16:1710–1718.

de Krom

, Bauer

, Collier

, Adan

R.A.

, la Fleur

S.E.

2009. Genetic variation and effects on human eating behavior. Annu. Rev. Nutr., 29:283–304.

10.

Farooqi

, O'Rahilly

2006. Genetics of obesity in humans. Endocr. Rev., 27:710–718.

11.

Garza

J.C.

, Kim

C.S.

, Liu

, Zhang

, Lu

X.Y.

2008. Adeno-associated virus-mediated knockdown of melanocortin-4 receptor in the paraventricular nucleus of the hypothalamus promotes high-fat diet-induced hyperphagia and obesity. J. Endocrinol., 197:471–482.

12.

Graham

, Shutter

J.R.

, Sarmiento

, Sarosi

, Stark

K.L.

1997. Overexpression of Agrt leads to obesity in transgenic mice. Nat. Genet., 17:273–274.

13.

Grimm

, Kay

M.A.

, Kleinschmidt

J.A.

2003. Helper virus-free, optically controllable, and two-plasmid-based production of adeno-associated virus vectors of serotypes 1 to 6. Mol. Ther., 7:839–850.

14.

Haskell-Luevano

, Chen

, Li

, Chang

, Smith

M.S.

, Cameron

J.L.

, Cone

R.D.

1999. Characterization of the neuroanatomical distribution of agouti-related protein immunoreactivity in the rhesus monkey and the rat. Endocrinology, 140:1408–1415.

15.

Hauck

, Chen

, Xiao

2003. Generation and characterization of chimeric recombinant AAV vectors. Mol. Ther., 7:419–425.

16.

Hermonat

P.L.

, Muzyczka

1984. Use of adeno-associated virus as a mammalian DNA cloning vector: Transduction of neomycin resistance into mammalian tissue culture cells. Proc. Natl. Acad. Sci. U.S.A., 81:6466–6470.

17.

Howard

D.B.

, Powers

, Wang

, Harvey

B.K.

2008. Tropism and toxicity of adeno-associated viral vector serotypes 1, 2, 5, 6, 7, 8, and 9 in rat neurons and glia in vitro. Virology, 372:24–34.

18.

Kas

M.J.

, Tiesjema

, van Dijk

, Garner

K.M.

, Barsh

G.S.

, ter Brake

, Verhaagen

, Adan

R.A.

2004. Induction of brain-region-specific forms of obesity by agouti. J. Neurosci., 24:10176–10181.

19.

Klein

R.L.

, Hamby

M.E.

, Gong

, Hirko

A.C.

, Wang

, Hughes

J.A.

, King

M.A.

, Meyer

E.M.

2002. Dose and promoter effects of adeno-associated viral vector for green fluorescent protein expression in the rat brain. Exp. Neurol., 176:66–74.

20.

Klein

R.L.

, Dayton

R.D.

, Leidenheimer

N.J.

, Jansen

, Golde

T.E.

, Zweig

R.M.

2006. Efficient neuronal gene transfer with AAV8 leads to neurotoxic levels of tau or green fluorescent proteins. Mol. Ther., 13:517–527.

21.

Klein

R.L.

, Dayton

R.D.

, Tatom

J.B.

, Henderson

K.M.

, Henning

P.P.

2008. AAV8, 9, Rh10, Rh43 vector gene transfer in the rat brain: Effects of serotype, promoter and purification method. Mol. Ther., 16:89–96.

22.

Kwon

, Schaffer

D.V.

2008. Designer gene delivery vectors: Molecular engineering and evolution of adeno-associated viral vectors for enhanced gene transfer. Pharm. Res., 25:489–499.

23.

McCrimmon

R.J.

, Shaw

, Fan

, Cheng

, Ding

, Vella

M.C.

, Zhou

, McNay

E.C.

, Sherwin

R.S.

2008. Key role for AMP-activated protein kinase in the ventromedial hypothalamus in regulating counterregulatory hormone responses to acute hypoglycemia. Diabetes, 57:444–450.

24.

McFarland

N.R.

, Lee

J.S.

, Hyman

B.T.

, McLean

P.J.

2009. Comparison of transduction efficiency of recombinant AAV serotypes 1, 2, 5, and 8 in the rat nigrostriatal system. J. Neurochem., 109:838–845.

25.

Rabinowitz

J.E.

, Bowles

D.E.

, Faust

S.M.

, Ledford

J.G.

, Cunningham

S.E.

, Samulski

R.J.

2004. Cross-dressing the virion: The transcapsidation of adeno-associated virus serotypes functionally defines subgroups. J. Virol., 78:4421–4432.

26.

Reed

S.E.

, Staley

E.M.

, Mayginnes

J.P.

, Pintel

D.J.

, Tullis

G.E.

2006. Transfection of mammalian cells using linear polyethylenimine is a simple and effective means of producing recombinant adeno-associated virus vectors. J. Virol. Methods, 138:85–98.

27.

Reimsnider

, Manfredsson

F.P.

, Muzyczka

, Mandel

R.J.

2007. Time course of transgene expression after intrastriatal pseudotyped rAAV2/1, rAAV2/2, rAAV2/5, and rAAV2/8 transduction in the rat. Mol. Ther., 15:1504–1511.

28.

Renström

, Payne

, Nordström

, Brito

E.C.

, Rolandsson

, Hallmans

, Barroso

, Nordström

, Franks

. GIANT Consortium. 2009. Replication and extension of genome-wide association study results for obesity in 4923 adults from northern Sweden. Hum. Mol. Genet., 18:1489–1496.

29.

Royo

N.C.

, Vandenberghe

L.H.

, Ma

J.Y.

, Hauspurg

, Yu

, Maronski

, Johnston

, Dichter

M.A.

, Wilson

J.M.

, Watson

D.J.

2008. Specific AAV serotypes stably transduce primary hippocampal and cortical cultures with high efficiency and low toxicity. Brain Res., 1190:15–22.

30.

Schaeren-Wiemers

, Gerfin-Moser

1993. A single protocol to detect transcripts of various types and expression levels in neural tissue and cultured cells: In situ hybridization using digoxigenin-labelled cRNA probes. Histochemistry, 100:431–440.

31.

Semjonous

N.M.

, Smith

K.L.

, Parkinson

J.R.

, Gunner

D.J.

, Liu

Y.L.

, Murphy

K.G.

, Ghatei

M.A.

, Bloom

S.R.

, Small

C.J.

2009. Coordinated changes in energy intake and expenditure following hypothalamic administration of neuropeptides involved in energy balance. Int. J. Obes. (Lond.), 33:775–785.

32.

Shafrir

, Ziv

2009. A useful list of spontaneously arising animal models of obesity and diabetes. Am. J. Physiol. Endocrinol. Metab., 296:E1450–E1452.

33.

Shevtsova

, Malik

J.M.

, Michel

, Bähr

, Kügler

2005. Promoters and serotypes: Targeting of adeno-associated virus vectors for gene transfer in the rat central nervous system in vitro and in vivo. Exp. Physiol., 90:53–59.

34.

Small

C.J.

, Liu

Y.L.

, Stanley

S.A.

, Connoley

I.P.

, Kennedy

, Stock

M.J.

, Bloom

S.R.

2003. Chronic CNS administration of Agouti-related protein (Agrp) reduces energy expenditure. Int. J. Obes. Relat. Metab. Disord., 27:530–533.

35.

Tice

J.A.

, Karliner

, Walsh

, Petersen

A.J.

, Feldman

M.D.

2008. Gastric banding or bypass? A systematic review comparing the two most popular bariatric procedures. Am. J. Med., 121:885–893.

36.

Veldwijk

M.R.

, Topaly

, Laufs

, Hengge

U.R.

, Wenz

, Zeller

W.J.

, Fruehauf

2002. Development and optimization of a real-time quantitative PCR-based method for the titration of AAV-2 vector stocks. Mol. Ther., 6:272–278.

37.

Vihinen-Ranta

, Suikkanen

, Parrish

C.R.

2004. Pathways of cell infection by parvoviruses and adeno-associated viruses. J. Virol., 78:6709–6714.

38.

Willer

C.J.

, Speliotes

E.K.

, Loos

R.J.

, Li

, Lindgren

C.M.

, Heid

I.M.

, Berndt

S.I.

, Elliott

A.L.

, Jackson

A.U.

, Lamina

, Lettre

, Lim

, Lyon

H.N.

, McCarroll

S.A.

, Papadakis

, Qi

, Randall

J.C.

, Roccasecca

R.M.

, Sanna

, Scheet

, Weedon

M.N.

, Wheeler

, Zhao

J.H.

, Jacobs

L.C.

, Prokopenko

, Soranzo

, Tanaka

, Timpson

N.J.

, Almgren

, Bennett

, Bergman

R.N.

, Bingham

S.A.

, Bonnycastle

L.L.

, Brown

, Burtt

N.P.

, Chines

, Coin

, Collins

F.S.

, Connell

J.M.

, Cooper

, Smith

G.D.

, Dennison

E.M.

, Deodhar

, Elliott

, Erdos

M.R.

, Estrada

, Evans

D.M.

, Gianniny

, Gieger

, Gillson

C.J.

, Guiducci

, Hackett

, Hadley

, Hall

A.S.

, Havulinna

A.S.

, Hebebrand

, Hofman

, Isomaa

, Jacobs

K.B.

, Johnson

, Jousilahti

, Jovanovic

, Khaw

K.T.

, Kraft

, Kuokkanen

, Kuusisto

, Laitinen

, Lakatta

E.G.

, Luan

, Luben

R.N.

, Mangino

, McArdle

W.L.

, Meitinger

, Mulas

, Munroe

P.B.

, Narisu

, Ness

A.R.

, Northstone

, O'Rahilly

, Purmann

, Rees

M.G.

, Ridderstråle

, Ring

S.M.

, Rivadeneira

, Ruokonen

, Sandhu

M.S.

, Saramies

, Scott

L.J.

, Scuteri

, Silander

, Sims

M.A.

, Song

, Stephens

, Stevens

, Stringham

H.M.

, Tung

Y.C.

, Valle

T.T.

, Van Duijn

C.M.

, Vimaleswaran

K.S.

, Vollenweider

, Waeber

, Wallace

, Watanabe

R.M.

, Waterworth

D.M.

, Watkins N; Wellcome Trust Case Control

Consortium

, Witteman

J.C.

, Zeggini

, Zhai

, Zillikens

M.C.

, Altshuler

, Caulfield

M.J.

, Chanock

S.J.

, Farooqi

I.S.

, Ferrucci

, Guralnik

J.M.

, Hattersley

A.T.

, Hu

F.B.

, Jarvelin

M.R.

, Laakso

, Mooser

, Ong

K.K.

, Ouwehand

W.H.

, Salomaa

, Samani

N.J.

, Spector

T.D.

, Tuomi

, Tuomilehto

, Uda

, Uitterlinden

A.G.

, Wareham

N.J.

, Deloukas

, Frayling

T.M.

, Groop

L.C.

, Hayes

R.B.

, Hunter

D.J.

, Mohlke

K.L.

, Peltonen

, Schlessinger

, Strachan

D.P.

, Wichmann

H.E.

, McCarthy

M.I.

, Boehnke

, Barroso

, Abecasis

G.R.

, Hirschhorn

J.N.

Genetic Investigation of ANthropometric Traits Consortium. 2009. Six new loci associated with body mass index highlight a neuronal influence on body weight regulation. Nat. Genet., 41:25–34.

39.

Xiao

, Li

, Samulski

R.J.

1998. Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J. Virol., 72:2224–2232.

40.

Yang

, Scott

K.A.

, Hyun

, Tamashiro

K.L.

, Tray

, Moran

T.H.

, Bi

2009. Role of dorsomedial hypothalamic neuropeptide Y in modulating food intake and energy balance. J. Neurosci., 29:179–190.

41.

Zolotukhin

, Potter

, Zolotukhin

, Sakai

, Loiler

, Fraites

T.J.

Jr. , Chiodo

V.A.

, Phillipsberg

, Muzyczka

, Hauswirth

W.W.

, Flotte

T.R.

, Byrne

B.J.

, Snyder

R.O.

2002. Production and purification of serotype 1, 2, and 5 recombinant adeno-associated viral vectors. Methods, 28:158–167.