A total of 52 male Sprague Dawley (SD) rats (5 weeks, 150–200 g) were purchased from Dahuo Animal Company (Chengdu, China) and housed in a specific pathogen free (SPF) grade animal laboratory at the Southwest Medical University. Groups of six rats were housed in a translucent box (35 cm $\times$ 30 cm $\times$ 16 cm), with tail markers numbered from one to six to facilitate behavioural observation and video recording. Room temperature was controlled (22 °C), with a 12-hour light cycle (8:00–20:00). During the first 3 days of the study; subjects were held and handled for 5 min per day and then placed on the open behaviour field for 30 min. Subsequently, each rat (1–52) was given an ear tag to record the initial weight, activity level, and age. All experiments with live animals were performed in strict accordance with animal testing standards.

The 52 SD rats were divided randomly into QNP model groups (n = 46) and one saline control group (NS group, n = 6). A subcutaneous neck injection of QNP (Q102, Sigma-Aldrich, St. Louis, MO, USA; 0.5 mg/kg dissolved in saline, NaCl 0.9%, at a volume of 1 mL/kg) was given twice a week for a total duration of 5 weeks to the 46 QNP group rats. Ten minutes after each injection, the rats were observed and recorded to determine whether they continued typical foraging behaviour (exploring the ground or an object with the nose for at least 5 sec), grooming, and other compulsive behaviours. A subcutaneous neck injection of saline (1 mL/kg; NaCl 0.9%) was also similarly given twice a week to each of the six rats in the NS group.

To verify the accuracy of coordinates of each surgical target, much early work was required. Target coordinates were obtained from the anatomical map of rats (Paxinos & Watson, 1997): NAC (2 mm anterior to bregma, 1.5 mm lateral to the midline, and 8.5 mm ventral to skull); AC (1.5 mm posterior to bregma, 3 mm lateral to the midline, and 7.1 mm ventral to skull); MD (3 mm posterior to bregma, 0.7 mm lateral to the midline, and 5.5 mm ventral to skull).

After the model setup, 46 QNP rats were randomly divided into five groups: 6-OHDA anterior limb of internal capsule (AC) stereotactic injection group (QNP+AC group, n = 10), 6-OHDA mediodorsal thalamic nucleus (MD) stereotactic injection group (QNP+MD group, n = 10), 6-OHDA nucleus accumbens (NAC) stereotactic injection group (QNP+NAC group, n = 10), saline (NS) stereotactic injection group (QNP+NS-S group, n = 10), and non-surgical group (QNP+Non-S group, n = 6). In the NS group, rats simultaneously received a 6-OHDA stereotactic injection (NS+6-OHDA-S group, n = 6: AC2, MD2, NAC2).

Rats were anaesthetized for surgery with 2% sodium pentobarbital (P3761, Sigma-Aldrich, St. Louis, MO, USA; 60 mg/kg) intraperitoneal injection, fixed on the animal stereotaxic, and the tooth and ear rods were adjusted to fit appropriately. After confirmation of no active bleeding, a 5-µL trace syringe was slowly inserted intracranially to reach targets. A semi-automatic constant rate trace injection pump injected 12.5 µg 6-OHDA (H116-5 mg, Sigma, Lot# 066M4602V, Pcode:1002800500, USA, 1 µL/5 µg) within 2 min into the targets, making the drugs fully diffuse. The same procedure was applied to the other side of the injection group (normal saline injection in the same way). The operation was completed, and a 38 °C constant temperature water bath blanket was used for rewarming. When fully awake after anesthesia, the rats were put back into their original environment.

First, it was verified whether the behavior of the rats tested in the groups differed significantly from that of rats tested separately. Twenty-four rats with the same conditions were randomly divided into four groups, with six in each group. After the injection of QNP, each rat in group A was evaluated separately, whereas after injection of QNP, all three rats in group B were evaluated together; after the injection of normal saline, each rat in group C was evaluated separately; after the injection of normal saline, all three rats in group D were evaluated together. The results showed no significant differences in behaviour between groups A and B or C and D (Table 1). Therefore, in formal experiments, we took the approach of placing three rats simultaneously on an open platform for behavioral assessments (a schematic of behavioral assessment is shown in Fig. 1). The Noldus EthoVision XT (Noldus Information Technology BV., Beijing, China) behavior observation, record, and analysis system recorded and analyzed the behavior of the rats in detail. After the evaluation of each group, any rat feces on the open platform were removed, and the venue was cleaned and disinfected to avoid interference with subsequent observations.

Fig. 1.

Behavioral assessment and modeling schematic. The open site platform size was 160 cm $\times$ 120 cm, divided into 12 squares of 30 cm $\times$ 30 cm. Four transparent glass boxes (10 cm $\times$ 10 cm $\times$ 8 cm) had one side open toward the center of the platform, allowing the rats to insert their head and front paws into the glass box. The four boxes were labelled A, B, C, and D, of which two were fixed in the corner and the other two were near the center. An overhead video camera was used to observe the body movements of the rats and their position in the open field.

Rodents characteristically explore new things and seek shelter. Szechtman et al. [13] observed that rats visited a box on the open platform the most and stayed the longest in a certain period after QNP injection, so it was defined as “home”. The rats’ behavior was mainly analyzed as the behavioural indicators related to “home”. The home base consisted of four transparent glass boxes (10 cm $\times$ 10 cm $\times$ 8 cm) that had one side open toward the centre of the platform to allow rats to insert their head and front paws into the box.

Number of home base visits (NOH): the box with the most visits was identified as the home base [13]. The total number of home base visits within 30 min was recorded. Most parts of a rat’s head and at least one forepaw went into the glass box for at least 1 s per “home” visit, and if they either visited the same “home” within a 3-s interval or there was a clear tendency to move away from home but then returned, this was counted as a single visit.

Average time between each home base visit (ATBO) (s): the average time interval between each trip to and from the home base within 30 min [13].

Total distance travelled (TDM) (number of squares): the total number of squares entered by each rat during the 30 min observation period was counted [13].

The rats were decapitated under deep anaesthesia (2% sodium pentobarbital 60 mg/kg), before which QNP was subcutaneously injected for the last time. Frozen specimen sections in the coronal plane at a 5-µm thickness were obtained from six brains of each group. There were four consecutive coronal sections for each region. Verification of placements was performed using the atlas of Paxinos and Watson. Only injections placed correctly in the target area were included in the statistical analysis of results. The sections were soaked in 0.1% Triton X-100 for 30 min, then blocked by 1% bovine serum albumin (BSA) for 1 h, followed by incubation with rabbit anti-tyrosine hydroxylase (TH) (ab6211, Abcam, Cambridge, MA, USA; 1:1000) for 48 h at 4 °C. The sections were then washed three times with PBS (15 min per wash) and incubated with Alexa Fluor™ 647 Donkey anti-Rabbit IgG (ab150075, Abcam, Cambridge, MA, USA; 1:500) overnight at 4 °C. Images were collected as soon as possible under an orthotopic two-photon laser confocal microscope (NIKON, Tokyo, Japan). Finally, the results were analyzed using the NIS-Elements BR image analysis system (NIKON, Tokyo, Japan).

SPSS version 19.0 (IBM, Armonk, NY, USA) was used to process data; measurement data were expressed as mean $\pm$ standard deviation (M $\pm$ SD). Considering the small sample size and the non-normal data distributions, a non-parametric test was adopted. The Kruskal-Wallis H and Mann-Whitney U tests were used to compare the behavioural scores between groups. Friedman’s variance analysis was used to detect the variation trend of each group at different time points (5 weeks before surgery, 3 weeks, and 4 weeks post-surgery). The Wilcoxon rank sum test was used to examine the difference between each postoperative and preoperative time points. The Mann-Whitney U test was used to compare the differences between groups at the same time point. The Mann-Whitney U test was used to compare the fluorescence values of different brain regions in the three groups with those in the non-surgical model group, 4 weeks after surgical intervention. All data graphs were made using GraphPad Prism 5 software (GraphPad Software, Boston, MA, USA).

After 5 weeks of QNP injections, the rats showed significant compulsive behaviour in the open field, and the number of “home” visits (NOH) in each group was nearly 5–7 times that of the NS+6-OHDA-S group. The results of the Mann-Whitney U analysis indicated that all the QNP groups were significantly different to the NS+6-OHDA-S group 2.6 (1.7) and the results were as follows: QNP+AC group 16.6 (5.2), Z = –3.261, p = 0.001; QNP+MD group 14.3 (5.8), Z = –3.259, p = 0.001; QNP+NAC group 14.2 (4.9), Z = –3.271, p = 0.001; QNP+NS-S group 13.9 (5.1), Z = –3.261, p = 0.001; QNP+Non-S group 13.3 (4.6), Z = –2.887, p = 0.004. The grade of NOH results at the 5-week time-point were as follows: 10 rats extremely severe (more than 20), 14 rats severe (16–20), 14 rats moderately severe (11–15), four rats mild severe (6–10), and four rats invalid (0–5). The same results were obtained for ATBO and TDM, with all p-values 0.05, which was assumed to indicate statistical significance (Table 2).

After the rats received a QNP injection for 3 weeks post-surgery, open-field NOH was reduced to different degrees in the three target groups (QNP+AC, QNP+MD, and QNP+NAC groups) (Fig. 2a). The Wilcoxon signed rank test results indicated differences between pre-surgical 16.6 (5.2) and 3-week post-surgery 12.4 (4.6), Z = –1.901, p = 0.057 in the QNP+AC group; pre-surgical 14.3 (5.8) and 3-week post-surgery 8.8 (3.9), Z = –2.316, p = 0.021 in the QNP+MD group; and pre-surgical 14.2 (4.9) and 3-week post-surgery 11.3 (4.3), Z = –1.480, p = 0.139 in the QNP+NAC group. The QNP+MD group showed the most remarkable improvements (p 0.05). Four weeks after the rats received a QNP injection, the NOH further decreased to different degrees in the QNP+AC and QNP+NAC groups in the open field, while the QNP+MD group presented a certain degree of aggravation. In addition, the ATBO and TDM of rats in the QNP+AC, QNP+MD, and QNP+NAC groups also improved to some extent after surgery (Fig. 2b,c). The behavior of rats in the QNP+NS-S, QNP+Non-S, and NS+6-OHDA-S groups were not significantly different before or after surgery (Fig. 2a–c).

Fig. 2.

Behavioral changes of rats in each group before surgery, and 3 and 4 weeks after surgery. (a) Statistical analysis of NOH, (b) ATBO, and (c) TDM at three time points in each group. Compulsive behaviors of rats in the QNP+AC, QNP+MD and QNP+NAC groups were significantly improved after surgery, especially in the QNP+AC group. *Significant statistical difference compared with the QNP+Non-S group, p 0.05; “ns” indicates no statistical difference. NOH, number of home base visits; ATBO, average time between each home base visit; TDM, total distance travelled; AC, anterior limb of internal capsule; QNP, quinpirole; MD, mediodorsal thalamic nucleus; NAC, nucleus accumbens; NS-S, saline stereotactic injection group; Non-S, non-surgical group; NS, saline; 6-OHDA-S, 6-hydroxydopamine stereotactic injection group.

Immunofluorescence image acquisition at 4 weeks post-surgery showed TH staining in the corresponding target regions (internal capsule (IC), MD, NAC and ventral tegmental area (VTA)) was significantly reduced in the three 6-OHDA-injected groups (QNP+AC, QNP+MD, and QNP+NAC) compared with the QNP+Non-S group (Fig. 3a,b). TH fluorescence was widely seen in all regions of the QNP+Non-S group, among which the IC region is mainly composed of DA nerve fibers, while the NAC, MD, and VTA regions are primarily composed of DA neurons. TH fluorescence analysis in each region of the QNP+AC group indicated that the fluorescence of DA nerve fibers across the IC region was significantly reduced. DA neurons in the VTA region were decreased, and a comparison between the QNP+AC group 79.3 (27.6) and QNP+Non-S group 184.1 (23.6) in the VTA region, Z = –2.446, p = 0.017, was found to be statistically significant. On the contrary, the NAC and MD regions showed no significant change. TH fluorescence analysis of each region in the QNP+MD group indicated that the fluorescence of DA neurons in the MD region was significantly decreased. While DA nerve fibers across the IC region were also partially reduced, the central areas were concentrated in the middle segment’s medial part. DA neurons in the IC region in the QNP+MD group 143 (58.3) were compared with the QNP+Non-S group 238 (72.6), Z = –1.473, p = 0.037, and the difference was statistically significant. TH fluorescence analysis of all regions in the QNP+NAC group indicated that, compared with the QNP+Non-S group, the numbers of DA neurons in the NAC and VTA regions were decreased. The number of DA nerve fibers across the IC region was also partially reduced. However, the MD region showed no significant change in DA neurons.

Fig. 3.

Immunofluorescence analysis of dopamine neurons and fibers. (a) Representative images of TH (red) immunofluorescence staining (the scale bar represents 100 µm). IC, Internal capsule; GP, Globus pallidus; chp, Choroid plexus; MD, Mediodorsal thalamic nucleus; D3V, Dorsal 3rd ventricle; AcbC, Core of nucleus accumbens; AcbSh, Shell of nucleus accumbens; NAC, nucleus accumbens (including AcbC and AcbSh); aca, Anterior union; VTA, Ventral tegmental area; SNR, Reticular part of substantia nigra. (b) Immunofluorescence expression analysis of dopamine neurons and fibers in rats of each group. *Significant statistical difference compared with the QNP+Non-S group, p 0.05; “ns” indicates no statistical difference.