Targeting Dyrk1A with AAVshRNA attenuates motor alterations in TgDyrk1A, a mouse model of Down syndrome

Abstract

Genetic-dissection studies carried out with Down syndrome (DS) murine models point to the critical contribution of Dyrk1A overexpression to the motor abnormalities and cognitive deficits displayed in DS individuals. In the present study we have used a murine model overexpressing Dyrk1A (TgDyrk1A mice) to evaluate whether functional CNS defects could be corrected with an inhibitory RNA against Dyrk1A, delivered by bilateral intrastriatal injections of adeno-associated virus type 2 (AAVshDyrk1A). We report that AAVshDyrk1A efficiently transduced HEK293 cells and primary neuronal cultures, triggering the specific inhibition of Dyrk1A expression. Injecting the vector into the striata of TgDyrk1A mice resulted in a restricted, long-term transduction of the striatum. This gene therapy was found to be devoid of toxicity and succeeded in normalizing Dyrk1A protein levels in TgDyrk1A mice. Importantly, the behavioral studies of the adult TgDyrk1A mice treated showed a reversal of corticostriatal-dependent phenotypes, as revealed by the attenuation of their hyperactive behavior, the restoration of motor-coordination defects, and an improvement in sensorimotor gating. Taken together, the data demonstrate that normalizing Dyrk1A gene expression in the striatum of adult TgDyrk1A mice, by means of AAVshRNA, clearly reverses motor impairment. Furthermore, these results identify Dyrk1A as a potential target for therapy in DS.

Figure 1

Generation and In Vitro Characterization of AAV-shRNA Targeting Dyrk1A (A) Diagram of the recombinant AAV2/2 viral vectors containing the hairpin sequences under the control of the mouse U6 promoter and the firefly luciferase reporter gene under the control of the CMV promoter AAVshDyrk1A and AAVscDyrk1A. (B and C) Determination of luciferase activity in HEK293 cells (B) and primary cultures of CGNs (C) infected at 50, 100, and 500 MOI with the AAVshDyrk1A or AAVscDyrk1A virus. Data are expressed in relative light units (RLU or LU/μg). Values are expressed as mean ± SEM of three independent experiments. (D and E) Western blot showing Dyrk1A levels in control mock infected cultures (C) or after infection with 50, 100, and 500 MOI of AAVshDyrk1A and AAVscDyrk1A in HEK293 cells (D) or 500 MOI of AAVshDyrk1A in CGNs (E). Actin was used as an internal control. Bars show densitometric analysis of the Dyrk1A band normalized against the actin band.

Figure 2

Follow-Up of Luciferase Expression after Intrastriatal Injection of AAVshDyrk1A of TgDyrk1A Mice (A) In vivo bioluminescence detection of luciferase activity in live animals and ex vivo sliced brains (slice width of 1 mm). Units are expressed as total photon counts. A viral dose of 1 × 10⁹ vp in a 3 μl total volume was injected per hemisphere. Red dot indicates stereotaxic injection coordinates. Animals were sacrificed 2 weeks after injection. (B) Quantification of the luciferase activity present in the extracts of different brain areas 20 days after intrastriatal injection of AAVshDyrk1A (STR, striatum; CO, cortex; CB, cerebellum). Data are expressed in RLU. Values are expressed as mean ± SEM (n = 4 mice). (C) Time course of luciferase expression after intrastriatal administration of AAVshDyrk1A. Animals were sacrificed at different days after intervention. Data are given in RLU. Values are expressed as mean ± SEM (n = 4 mice). The asterisk indicates significance with a p value < 0.05.

Figure 3

Kinetics of Dyrk1A Inhibition in the Striatum of TgDyrk1A Mice after AAVshDyrk1A Administration Western-blot analysis of Dyrk1A protein levels. (A) shows representative images of Dyrk1A expression from striata ipsilaterally injected with AAVshDyrk1A and contralaterally injected with AAVscDyrk1A at 1, 2, 3, and 4 months after injection. Bars show densitometric analysis of the Dyrk1A band, normalized against the actin band and represented as the ratio of shDyrk1A versus scDyrk1A signal. Data are expressed as the mean ± SEM (n = 4 mice). (B) shows a representative western blot of Dyrk1A expression from TgDyrk1A striata injected with AAVshDyrk1A or with AAVscDyrk1A and from wild-type mice injected with saline. Bars show densitometric analysis of Dyrk1A signal from striata normalized to Dyrk1A hippocampal levels. Data are expressed as the mean ± SEM (n = 4 mice). Actin was used as an internal control.

Figure 4

Analysis of AAVshDyrk1A Striatal Toxicity Animals were stereotaxically injected with saline in the striata ipsilateral hemisphere and with AAVshDyrk1A in the contralateral hemisphere. Immunohistochemical staining of DARPP-32 and luciferase was performed on coronal sections of brains harvested 2 weeks after injection, with an anti-DARPP32 monoclonal and anti-luciferase polyclonal antibody, respectively. Representative photomicrographs of DARPP-32-positive medium spiny neurons (upper panel) and luciferase-positive neurons (lower panel) are shown. The scale bar represents 1 mm for the upper panel and 200 μm for the lower panel (100 μm for higher magnification).

Figure 5

Behavioral Analysis of Motor Tasks in TgDyrk1A Treated with Intrastriatal Injections of AAVshDyrk1A Animals were randomly assigned to four treatment groups: wild-type (saline) and TgDyrk1A (saline, AAVshDyrk1A, and AAVscDyrk1A). (A) Actimetry test. The results are given as total distance traveled by the animals in 24 hr expressed in cm. Preinjection results confirmed the hyperactive phenotype of TgDyrk1A mice compared to the control group (F_1,52 = 4.898, ^∗p = 0.031). Two months after the intervention, a decrease in the hyperactivity was observed in TgDyrk1A mice injected with AAVshDyrk1A (TG sh) with respect to saline (TG sal), (F_1,23 = 4.286, ^∗p = 0.05). This reduction became more evident when repeating the test 4 months after treatment (TG sal versus TG sh: F_1,24 = 5.225, ^∗p = 0.032; TG sc versus TG sh: F_1,23 = 9.05, ^∗∗p = 0.006), (n = 14 mice/group). (B) Treadmill test. The results are expressed as the total number of shocks received. Nontreated mice confirmed the presence of motor-coordination alterations in TgDyrk1A mice (F_1,22 = 11.937, ^∗∗p = 0.002). One month after the intervention, the AAVshDyrk1A TgDyrk1A group received a significantly less number of shocks compared to the groups of TgDyrk1A injected with either saline or AAVscDyrk1A, reaching levels similar to those of control groups (TG sal versus TG sh: F_1,6 = 8.669, ^∗p = 0.032; TG sc versus TG sh: F_1,8 = 9.844, ^∗p = 0.016). This effect was maintained 4 months after treatment in AAV-shDyrk1A TgDyrk1A mice (TG sal versus TG sh: F_1,10 = 5.575, ^∗p = 0.043; TG sc versus TG sh: F_1,10 = 6.094, ^∗p = 0.033), (n = 6 mice/group). (C–E) Prepulse inhibition (PPI) test. Results are expressed as the percentage of inhibition, that refers to the decrease in the amplitude of the startle response to a given acoustic pulse (120 dB) when it is preceded by a prepulse (74, 78, 82, 86, and 90 dB). As shown in (C), results before the intervention indicated deficits in the PPI in TgDyrk1A mice, as shown by a tendency to decreased PPI at higher prepulse intensities. As shown in (D), 4 months after the injection, AAVshDyrk1A TgDyrk1A showed a statistically significant increase in PPI at 74 dB and 86 dB prepulse intensities (TG sc versus TG sh; F_1,14 = 5.55, ^∗p = 0.035; F_1,14 = 5.234, ^∗p = 0.04). As shown in (E), comparisons of the AAVshDyrk1A-injected group at the indicated time points showed significant differences (TG sh Preinjection versus TG sh, 2 months: F_1,13 = 7.449, ^∗p = 0.018; Preinjection versus TG sh, 4 months: F_1,13 = 5.207, ^∗p = 0.042), (n = 7 mice/group). Data are expressed as means ± SEM.