Ras homolog enriched in striatum, which is encoded by the Rasd2 gene, was targeted for knockdown via CRISPR/Cas9 technology by using four guide RNAs (gRNAs) designed to target exon 2 of the mouse Rasd2 gene (T1, T2, T3, and T4) (Figure 1A). These gRNAs were then subcloned into a px552 vector, which expressed the gRNAs under the U6 promoter and red fluorescent protein (RFP) under the CMV promoter (Figure 1B). The expression of RFP served as an indicator of Rasd2 gRNA expression. To assess the targeting efficiency of Rasd2 gRNAs, we transfected them along with a Cas9 plasmid into cultured mouse neuroblastoma cells (N2a). Analysis of the DNA using T7E1 digestion revealed that Rasd2 gRNAs efficiently edited the Rasd2 gene (Figure 1C). For further validation in vivo, we selected two gRNAs (T3 and T4) that exhibited higher editing efficiency in N2a cells.

To assess the targeting effect of Rasd2 gRNAs in vivo, we packaged the px552 gRNA into an AAV9 virus. Stereotaxic injection of AAV viruses expressing control gRNA or Rasd2 gRNA was performed in the striatum of Cas9 mice (Figure 2A), that express Cas9 ubiquitously under the CAG promoter (Liu et al., 2020a; Yang et al., 2020). AAV-Rasd2-gRNAs of T3 and T4, either alone or in combination, were injected into one side of the striatum of 2-month-old Cas9 mice. The contralateral side was injected with AAV-control-gRNA, which did not target any known gene (Yang et al., 2017), to rule out non-specific effects caused by viral transduction. After 28 days, the brains of the AAV-injected mice were dissected, and the distribution of RFP fluorescence signal confirmed the predominant expression of Rasd2 gRNA in the striatum (Figure 2B). Western blotting analysis revealed a significant reduction in Rhes levels on the side injected with AAV-Rasd2 gRNAs (mixture of T3 and T4) compared to the contralateral side injected with AAV-control-gRNA (Figure 2C). This finding was further validated by immunohistochemical staining against Rhes in the AAV-injected brains (Figure 2D), confirming the efficient elimination of Rhes in vivo through AAV-Rasd2-gRNAs injection.

Previous studies have indicated that genetic knockout of Rhes during the embryonic stage can lead to abnormal physiological function in mice at a certain age (Spano et al., 2004; Heikkinen et al., 2021). Considering this, we decided to knockdown Rhes at different disease stages in the HD mouse model. The HD 140Q/Cas9 (referred to as HD KI) mice, which were generated previously by crossing HD 140Q KI mice with Cas9 transgenic mice (Liu et al., 2020a; Yang et al., 2020), were used in our study (Figure 3A). These mice exhibit intranuclear mHTT aggregations in the striatum and behavioral defects starting at 7 months old (28 weeks old), similar to the HD140Q KI mice (Yang et al., 2020). AAV injection was performed at 14 weeks old and 28 weeks old, corresponding to the presymptomatic and symptomatic stages, respectively (Figure 3A). One side of the HD KI mouse striatum was injected with AAV-Rasd2 gRNA, while the contralateral striatum was injected with AAV-control gRNA. Eight weeks after injection, the injected brains were examined (Figure 3A).

Since the striatum is the most affected brain region in HD and Rhes are mainly expressed in the striatum, we focused on the effect of Rhes knockdown on the striatum. To assess the impact of Rhes knockdown on the level of mHTT in the striatum, we examined the levels of mHTT in the injected mice. Intriguingly, in the 14-weeks-old-injected group, Western blotting showed a significant increase in the amount of mHTT aggregates in the AAV-Rasd2 gRNA injected side of the striatum after 8 weeks of injection, while no mHTT aggregation was observed in the AAV-control gRNA injected side (Figures 3B, C). However, in the 28-weeks-old-injected group, a substantial amount of aggregated mHTT was found on both sides of the striatum, with no significant difference (Figures 3B, C). In addition, no significant alternations were found on the level of fragments mHTT including exon1 mHTT, which were proteolyzed from full-length mHTT and play key roles in the pathogenesis of HD (Landles et al., 2010; Yang et al., 2020) (Figures 3B, C). To further validate these findings, the injected brain slices were subjected to immunohistochemistry using mEM48 staining. The results confirmed that AAV-Rasd2 gRNA increased the amount of aggregated mHTT in the 14-weeks-old-injected group (Figure 3D). These results indicate that knockdown of Rhes accelerates the aggregation of mHTT at the presymptomatic stage of HD KI mice.

Since preferential neuronal loss in the striatum is an important feature in HD patient brains, we aimed to investigate the impact of Rhes knockdown on neuronal loss in HD KI mice. Immunohistochemistry results revealed a notable reduction in NeuN staining intensity in the striatum of the AAV-Rasd2 gRNA-injected side as compared to the contralateral AAV-control gRNA-injected side in the 14-weeks-old-injected group (Figure 4A). However, in the 28-weeks-old-injected group, the difference in NeuN staining between the AAV-Rasd2 gRNA-injected side and the contralateral AAV-control gRNA-injected side was negligible (Figure 4A). Western blotting confirmed a significant decrease in NeuN levels on the AAV-Rasd2 gRNA-injected side compared to AAV-control gRNA-injected side in the 14-weeks-old injected group (Figures 4B, C).

Next, we examined astrogliosis, another hallmark of neurodegeneration characterized by reactive astrocytes with increased GFAP staining (Yang et al., 2017). As expected, GFAP staining was significantly elevated in the striatum of HD KI mice during the symptomatic stage. However, knockdown of Rhes did not notably increase GFAP staining in the 28-weeks-old-injected group (Figure 5A). Conversely, in the 14-weeks-old injected HD KI mice, GFAP staining was significantly increased in the AAV-Rasd2 gRNA-injected side compared to the contralateral AAV-control gRNA-injected side (Figures 5B, C).

Ras homolog enriched in striatum has been previously implicated as a modulator of autophagy (Mealer et al., 2014), a vital process involved in the clearance of misfolded proteins. Our results suggested that the depletion of Rhes led to the accumulation of mHTT in the striatum of HD KI mice at 22 weeks old (8 weeks after injection), suggesting a potential disturbance in autophagy during the presymptomatic stage.

To investigate the impact of Rhes knockdown on autophagy, we first examined the levels of P62, an autophagy substrate that is used as a reporter of autophagy activity (Johansen and Lamark, 2011). Western blotting analysis of the Rasd2 gRNA-injected and control-gRNA-injected striatum showed a significant increase in P62 levels in the AAV-Rasd2 gRNA-injected side, while no significant change was observed on the contralateral AAV-control gRNA-injected side (Figures 6A, B). For further investigation, we check the status of LC3, the most widely used autophagosome marker. Western blotting results showed a decrease of LC3-II and LC3-II/LC3-I in the AAV-Rasd2 gRNA-injected side (Figures 6A, C). Increased P62 level and decreased LC3-II suggested reduced autophagy activity during the presymptomatic stage upon Rhes knockdown.

Previous reports have indicated that Rhes may bind and activate mTOR to inhibit autophagy (Subramaniam et al., 2011). Nevertheless, we did not observe changes in mTOR activity, indicated by unaltered levels of mTOR and phosphorylation of mTOR (Figures 6A, D), after Rhes deletion. Considering Rhes’s potential role in regulating autophagy through an mTOR-independent mechanism involving Beclin-1 (Kang et al., 2011; Mealer et al., 2014; Menon and Dhamija, 2018), we assessed the protein level of Beclin-1 and its phosphorylation. A decreased protein level of Beclin-1 (Figures 6A, E, F) was found in the AAV-Rasd2 gRNA-injected side, which further supports impaired autophagy during the presymptomatic stage. These results suggested that Rhes deletion may lead to decreased autophagy activity through an mTOR-independent manner.

All mice were bred and maintained in the animal facility in a 12:12 h light/dark room in accordance with the institutional guidelines of the Animal Care and Use Committee at Central South University. Full-length mutant HTT knock-in (140CAG) (HD KI) mice and Cas9 transgenic mice were obtained from Professor Xiao-Jiang Li at Jinan University. Mouse genotyping was performed by PCR using genomic DNA extracted from the mouse tails with primers, HTT, forward: 5′-ACT GCT AAG TGG CGC CGC GTA G-3′ and reverse: GAC GCA GCA GCG GCT GTG CCT G; Cas9, common forward: AAG GGA GCT GCA GTG GAG TA, WT reverse: TCG AAA ATC TGT GGG AAG TC, and transgenic reverse: TGG GCC ATT TAC CGT AAG TTA T.

AAV-Rasd2-gRNAs were generated by inserting gRNA into a modified PX552 vector (Yang et al., 2017) via the SapI restriction site. gRNA sequences are: gRNA-T1: CAA GGT GTA CAA CAT CCA CG ggg, gRNA-T2: TAA GAA CTC CTA CCG CAT GG tgg, gRNA-T3: CCA GCT GGT ACA TGT CCC CG tgg and gRNA-T4: CCA GTA CAC GCC CAC TAT CG agg, and control gRNA: ACC GGA AGA GCG ACC TCT TCT [PAM sequences are shown in lowercase; control gRNA was verified previously (Yang et al., 2017)]. All constructs were confirmed by double-direction Sanger sequencing. The plasmids were packaged into AAVs using serotype 9 capsids.

Primary antibodies used in this study include Rhes (Genetex, GTX85428), mEM48 (QED Bioscience, 70102), NeuN (Cell Signaling Technology, 24307s), GFAP (Cell Signaling Technology, 3670), vinculin (Invitrogen, V9264), P62 (Abcam, ab56416), mTOR (Cell Signaling Technology, 2983), p-mTOR (Cell Signaling Technology, 5336), LC3 (Novus, NB100-2332), Beclin-1 (Cell Signaling Technology, 3495), p-Beclin-1 (Ser15, Cell Signaling Technology, 84966). Secondary antibodies were donkey anti-rabbit, and donkey anti-mouse Alexa Fluor 488 or 594 from Jackson ImmunoResearch.

Surgical procedures were performed in accordance with the guidelines for the Care and Use of Laboratory Animals and biosafety procedures at Central South University. Stereotaxic injection of AAV into mouse brains was performed following the protocol described in our previous studies (Liu et al., 2020a). Briefly, mice were anesthetized with 1.5% isoflurane inhalation and stabilized on a stereotaxic instrument (David Kopf Instruments). Meloxicam was injected (5 mg/kg) as an analgesic. The injection site of the striatum was determined based on the distance from the bregma following coordinates (Yang et al., 2017): anterior-posterior = +0.55 mm, medial-lateral = ±2 mm, dorsal-ventral = −3.5 mm. Two microliters of AAV-gRNA virus (10¹² genome copies/ml) per side was injected using a 30-guage Hamilton microsyringe at 200 nl/min speed. The microsyringe was left in the injection site for 10 min before it was withdrawn slowly. After the surgery, mice were placed on a heated blanket for recovery and monitored for 5 consecutive days.

N2a cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% FBS, 100 μg/ml penicillin, and 100 units/ml streptomycin at 37°C with 5% CO₂. Cells were transfected with 1 μg/well of DNA in a 12-well plate using lipofectamine 2000 (Invitrogen) for 48 h.

Mice were anesthetized and perfused intracardially with 0.9% saline solution, followed by 4% paraformaldehyde (PFA). Isolated mouse brains were post-fixed in 4% PFA overnight and then transferred into 30% sucrose for 48 h at 4°C. The completely dehydrated brains were sectioned at 30 μm by freezing microtome and stored in 0.1% NaN₃ in phosphate-buffered saline.

For immunohistochemistry, brain sections were permeabilized with 0.3% Triton X-100/PBS at room temperature for 1 h and then treated with citric acid buffer at 95°C for 15 min for antigen retrieval. After being blocked with 4% normal goat serum in 0.3% Triton X-100/1X PBS for 1 h, brain sections were incubated with primary antibodies overnight at 4°C. After three rounds of 1X PBS wash, brain sections were incubated with secondary antibodies. Brain sections were developed with the Avidin-Biotin Complex Kit (Vector ABC Elite, Burlingame, CA, USA) and Sigma Fast DAB kit (Millipore Sigma).

Transfected N2a cells or AAV-injected striatum of HD mice were lysed in ice-cold RIPA buffer with freshly added protease inhibitor cocktail. Samples were sonicated and heated at 100°C for 10 min to denature the protein. Proteins were separated by SDS-PAGE electrophoresis and then transferred onto a nitrocellulose membrane. After being blocked with 5% milk in 1X PBS, the membrane was incubated with the primary antibody in 3% BSA/1X PBS at 4°C overnight. The membrane was then incubated with HRP-conjugated secondary antibodies in 5% milk/1X PBS for 1 h at room temperature and then developed with ECL Prime (GE Healthcare).

The densitometric data (provided in Supplementary material) were obtained by the software ImageJ (1.51) and bar graphs were generated by the software GraphPad Prism (9.3.1). Two-tailed Student’s t-test was used to compare two groups. There were at least three injected mice striatum per group. All quantification data were presented as mean ± SEM. A P-value of less than 0.05 was considered statistically significant.