Transduction of brain dopamine neurons by adenoviral vectors is modulated by CAR expression: rationale for tropism modified vectors in PD gene therapy

Abstract

Background: Gene-based therapy is a new paradigm for the treatment of Parkinson disease (PD) and offers considerable promise for precise targeting and flexibility to impact multiple pathobiological processes for which small molecule agents are not available. Some success has been achieved utilizing adeno-associated virus for this approach, but it is likely that the characteristics of this vector system will ultimately create barriers to progress in clinical therapy. Adenovirus (Ad) vector overcomes limitations in payload size and targeting. The cellular tropism of Ad serotype 5 (Ad5)-based vectors is regulated by the Ad attachment protein binding to its primary cellular receptor, the coxsackie and adenovirus receptor (CAR). Many clinically relevant tissues are refractory to Ad5 infection due to negligible CAR levels but can be targeted by tropism-modified, CAR-independent forms of Ad. Our objective was to evaluate the role of CAR protein in transduction of dopamine (DA) neurons in vivo.

Methodology/principal findings: Ad5 was delivered to the substantia nigra (SN) in wild type (wt) and CAR transgenic animals. Cellular tropism was assessed by immunohistochemistry (IHC) in the SN and striatal terminals. CAR expression was assessed by western blot and IHC. We found in wt animals, Ad5 results in robust transgene expression in astrocytes and other non-neuronal cells but poor infection of DA neurons. In contrast, in transgenic animals, Ad5 infects SNc neurons resulting in expression of transduced protein in their striatal terminals. Western blot showed low CAR expression in the ventral midbrain of wt animals compared to transgenic animals. Interestingly, hCAR protein localizes with markers of post-synaptic structures, suggesting synapses are the point of entry into dopaminergic neurons in transgenic animals.

Conclusions/significance: These findings demonstrate that CAR deficiency limits infection of wild type DA neurons by Ad5 and provide a rationale for the development of tropism-modified, CAR-independent Ad-vectors for use in gene therapy of human PD.

Figure 1. Ad5 drives high levels of transgene production.

Sections through the core of the injection volume in the SN stained for the GFP transgene, demonstrate the robust levels of expression in this system. This activity made unbiased stereologic counting unfeasible. Panel A is SN of a wt animal. Panel B is SN of an hCAR transgenic animal.

Figure 2. Assessment of Ad5 transduction in the SN in vivo.

Confocal immunohistochemical assessment of 1e8 vp of Ad5-GFP delivered to the mouse SN (red) shows broad transgene expression in cells of glial morphology (green). The absence of double-labeled cells indicates Ad5 poorly infects neurons of the SNc (section through periphery of the injection site). Stereotactic coordinates: AP = 3.1, ML = −1.2, DV = −4.6. Red  =  TH–positive neurons, Green  =  GFP. Bar  = 50 µm.

Figure 3. hCAR transgenic animals exhibit transduction of dopamine neurons by Ad5 vectors.

Delivery of Ad5 to the SN in the hCAR mouse shows a broadened transduction profile including TH+ neurons by confocal microscopy. Transgene expression extends into the STR terminal fields. (A) Low power image of the hCAR SN shows a broadened transduction profile of Ad that includes TH-positive neurons, in addition to astrocytes. Magnification of the region outlined in A shows many TH-positive cells (red, B) staining positive for GFP (green, C) overlaid in D (see closed arrowheads). GFP-positive cells with astrocyte morphology can also be seen in this region (open arrowheads). (E) Visualization in the STR shows many fibers labeling with GFP, indicating their transduction by Ad5 delivered in the SNc. Red  =  TH-positive neurons, Green  =  GFP, Yellow  =  double positive. Scale bar A: 500 µm, B–D: 100 µm, E: 50 µm.

Figure 4. Quantitation of transgene in the striatal terminal fields.

Densitometric analysis of transgene expression in the STR shows the CAR-dependence for transduction of SNc dopaminergic neurons by Ad delivered to the ventral midbrain. (A) Representative flattened confocal z-stacks through a 35 µm section of the dorsolateral STR in three wt and hCAR animals. GFP-filled axons projecting from the SNc show the level of neuronal transduction that is augmented by the addition of hCAR transgene. (B) Densitometric quantification of individual confocal z-planes shows a significant increase in Ad5 transduction capacity in hCAR mice. (C) Drawing of a sagittal section through the mouse brain shows the location of vector delivery in the SNc and the location of transgene expression analysis in the STR, approximately 4.25 mm distal from the site of injection. Scale bar in A = 150 µm, Green  =  GFP. Box plot in B depicts sample minimum, 1^st quartile, median, 3^rd quartile and sample maximum. Statistical significance assigned using two-tailed Student's t-test (* indicates P<0.0001). LV  =  left ventricle, MFB  =  medial forebrain bundle.

Figure 5. CAR expression in wild type and hCAR transgenic brain.

Regional analysis of CAR protein shows variability in expression levels, which are elevated in hCAR transgenic animals. (A) A representative immunoblot of CAR protein levels in the ventral midbrain (vMB) and striatum (STR) shows large differences in expression. The truncated transgenic hCAR protein is seen running with a lower apparent molecular mass. Both mCAR and hCAR are visualized with the pan-CAR antibody H300. Transgenic hCAR is expressed in both the vMB and STR at a similar level to that of STR mCAR. (B) Quantitation of total CAR in the ventral midbrain by combining endogenous mCAR and transgenic hCAR shows five-fold more CAR protein available in this region in hCAR animals. Total CAR protein level of wt animals is considered baseline. Statistical significance assigned using two-tailed Student's t-test (* indicates P<0.02).

Figure 6. Histological analysis of hCAR localization shows expression in close apposition to SN neurons.

Although we were unable to visualize mCAR by immunohistochemistry (data not shown), confocal analysis of hCAR in the substantia nigra of transgenic animals (right panels) showed robust expression throughout the ventral midbrain (D). hCAR was readily seen within the SNc (D, E), and in close apposition to DA-positive SN neurons (F). (A–C) Left panels show lack of cross-reactivity of antibody RmcB for mCAR. Scale bar A, D: 250 µm, B, E: 50 µm, C, F: 10 µm. Red  =  TH-positive neurons, Green  =  hCAR.

Figure 7. Transgenic hCAR localizes to post-synaptic structures.

Assessing the expression pattern of exogenous hCAR in the SN and STR of transgenic animals, we found co-localization with the post-synaptic density marker PSD95 (A–H, closed arrowheads indicate areas of co-localization). SN =  substantia nigra, STR  =  striatum. Bars in A and E = 50 µm, bars in B and F = 5 µm.