In this study, we investigated changes in risk-avoidant behaviors. Adult male (n = 37) and female (n = 35) Sprague Dawley rats (Envigo, Indianapolis, IN, USA) weighing 200–400 g underwent a partial bilateral 6-hydroxydopamine (6-OHDA)-induced dopaminergic lesion to model the premotor stages of PD [35, 36]. The estrous cycle of the female rats was not considered in the analysis. Traditional fully lesioned hemi-parkinsonian 6-OHDA models show decreased rodent mobility [37], while the partially lesioned premotor model used in this study allows for the non-motor symptoms of PD to emerge prior to the presentation of the motor symptoms. Clinical DBS is not typically performed at this stage of neural degeneration in PD. However, the selected lesion severity is more consistent with the time when the non-motor symptoms of PD start to manifest [38]. Furthermore, this model reduces confounds from mobility restrictions ensuing from full lesions [35, 36] during assessment of risk-avoidant behaviors. The dopaminergic lesion level was confirmed using tyrosine hydroxylase (TH) immunohistologic analysis (see Immunohistologic Analysis section below). Rats were housed in a social environment prior to surgery in accordance with the Institutional Animal Care and Use Committee (IACUC) standards. Rats had ad libitum access to food and water and were kept on a 12-hour light/dark cycle. All behavioral tests were performed during the light portion of the light/dark cycle. All rats were housed individually following surgery.

Rats were divided into four experimental groups: lesioned undergoing chronic stimulation (n = 10), lesioned undergoing acute stimulation (n = 18), lesioned undergoing sham stimulation (n = 12), and vehicle controls (n = 10; no lesion and no DBS electrode). An additional 22 rats were used for anxiogenic caffeine control experiments, with 12 rats used for the elevated zero maze test, which received 100 mg of caffeine, and 10 rats used for the open field test, which received 50 mg of caffeine.

Anesthesia was induced using 3% isoflurane (NDC 66794-019-10, Piramal Critical Care, Inc, Bethlehem, PA, USA) and maintained with 1%–2% isoflurane throughout the procedure. Rats received 1 mg/kg buprenorphine (NDC 79926-058-17, Wedgewood Connect, San Jose, CA, USA) subcutaneously for analgesia prior to surgery. Burr holes were drilled above the striatum (Medial Lateral, ML $\pm$ 2.7, Anterior Posterior, AP +1.2, Dorsal Ventral, DV –4.2) [39] and STN (ML 2.5, AP –3.6, DV –7.8) [39] regions. To induce the dopaminergic lesion, 5 µg 6-OHDA (Sigma-Aldrich, Milwaukee, WI, USA) was first dissolved in a 2.5 µL saline solution containing 0.02% ascorbic acid (Sigma-Aldrich) and bilaterally injected into the striatum at a rate of 500 nL/min using a Hamilton syringe (Hamilton Company, Reno, NV, USA) and micro syringe pump (UMP3 Ultramicropump, World Precision Instruments, Sarasota, FL, USA). Rats in the non-lesioned group (vehicle controls) received a saline solution containing 0.02% ascorbic acid. All rats received a subcutaneous injection of 25 mg/kg desipramine (Sigma-Aldrich) 15–30 minutes prior to surgery to prevent noradrenergic neuron degeneration. A bipolar platinum-iridium electrode (MS303/8C, Plastics One, Roanoke, VA, USA) was implanted unilaterally in all groups into the right STN. The electrode was secured in place using four skull screws (MD-1310, BASi, West Lafayette, IN, USA) covered with a Metabond (Parkell, Edgewood, NY, USA) headcap and fixed with dental cement. Rats were allowed to recover for 2 weeks prior to electrical stimulation and behavioral testing.

Rats underwent unilateral STN DBS at 130 Hz with 60-µs charge-balanced biphasic pulses delivered using a stimulator built in-house [40]. Stimulation amplitude was determined for each animal by first identifying the minimum amplitude that evoked dyskinetic activity (motor threshold) [41]. Rats were stimulated at 80% of their motor threshold, which ranged from 50 to 300 µA. Rats who did not exhibit dyskinetic activity when stimulated at 300 µA were considered non-responders and excluded from the study, as stimulating above that threshold could cause significant tissue damage [42]. All rats were tethered to the stimulator for 2 hours a day starting on day 14 and continuing for the remainder of the study (day 23). Rats in the chronic stimulation group underwent daily stimulation for 2 hours starting on day 14 and continuing for the remainder of the study (day 23) (Fig. 1A). Acute rats were only stimulated during the behavior tests on days 21 to 23. Rats were kept in their home cages during this period of chronic stimulation or tethering.

Fig. 1.

Experimental approach. (A) Experimental timeline and test groups. Surgery to induce dopaminergic 6-hydroxydopamine (6-OHDA) lesioning or vehicle injection and deep brain stimulation electrode implantation was performed on day 0. Sham and chronic stimulation (when appropriate) began on day 14 and continued until day 23. Behavioral analysis using the elevated zero maze was performed on day 21 and again on day 22. The open field test was performed on day 23. Perfusion was performed following completion of all behavioral tests. (B) Elevated zero maze setup. Rats that spent more time in the closed arms exhibited behavior associated with anxiety and fear, while rats that spent more time in the open arms were considered less anxious and more willing to explore open environments. Dashed lines indicate the areas where the closed and open arms intersect and are used to quantify exploratory behavior. (C) Open field test setup. The shaded region indicates the center of the field used to quantify exploratory behavior. Anxious rats spent less time in the central, shaded, and more exposed areas of the field. The open field test was also used to quantify total distance traveled as a measure of locomotor function.

Elevated Zero Maze: an elevated zero maze was used to assess anxiety-related behaviors (time spent in open and closed arms, number of entries into closed arms, and number of head dips) in rats. The elevated zero maze consisted of a circular, 100 cm in diameter elevated platform with 10-cm wide tracks with two enclosed quadrants (closed arms) providing a sense of safety and security and two open quadrants (open arms) exposing the rat to the open, elevated environment (Fig. 1B). Prior to their placement in the elevated zero maze, rats were tethered and stimulated (when appropriate) in their home cage for 2 hours prior to testing. Habituation to the room with the elevated zero maze took place for 10 minutes while the rats were in their home cages. Following habituation, rats were placed in the elevated zero maze and allowed to move freely while tethered to the electrical stimulator, which was suspended above the maze. The test was conducted over 2 days to investigate the impact of a second day of testing. The first day of the elevated zero maze test (day 21) was used to capture animal behavior in a new environment, while the second day (day 22) was used to capture animal behavior in a non-novel environment [30, 43]. Rats were placed in the maze at one of the four intersections (Fig. 1B, dashed lines) between the open and closed arms but facing the closed arms. The specific intersection selected was varied for each animal. On day 21, rats were allowed to explore the maze for 10 minutes. On day 22, rats were allowed to explore the maze for 5 minutes. Times were determined based on the literature [43, 44]. Only the first 5 minutes of day 21 were analyzed to account for the difference in time between days 21 and 22. Stimulation was delivered to the acute and chronic experimental groups during the entire time rats were connected to the stimulator. Cameras placed directly above the maze recorded animal behavior, which was analyzed using AnyMaze (Stoelting, Chicago, IL, USA). Time spent in the open and closed arms as well as total distance traveled were also measured using AnyMaze. Entries into the closed arm of the maze, defined as 75% of the animals entering the closed arm, were determined by video analysis. Head dips, defined as the times the animal’s head dropped below the bottom of the maze while in the elevated open arms, were determined by video analysis. Open Field Test: risk-avoidance and mobility were characterized using an open field test setup 60 cm in length, 60 cm in width, and 44 cm in height (Fig. 1C). A high-definition video camera and stimulator were placed directly above the field to allow for simultaneous stimulation and video recording while the rats moved freely in the field. Time spent in the center of the field (defined as 15 cm from the wall of the field) and the total distance traveled were measured using EthoVision (Noldus, Wageningen, the Netherlands). On the day of testing, rats were tethered and connected to the stimulator for 2 hours in their home cage, and this was followed by 10 minutes of habituation and 5 minutes of behavioral testing inside the open field setup. Stimulation was delivered to the acute and chronic experimental groups during the entire time rats were connected to the stimulator.

Two additional experimental groups of rats received an intraperitoneal (IP) injection of either caffeine or saline to validate the ability to detect anxiogenic behaviors using the elevated zero maze and open field test. These rats received no lesion. Caffeine (Sigma Aldrich) was administered at 100 mg/kg for the elevated zero maze and 50 mg/kg for the open field test [44, 45]. Rats were introduced to each test 30 minutes following administration of caffeine or saline. Testing was conducted following the same approach used for the stimulation experiments described previously.

Once all experiments were completed, rats were perfused by exposing the heart and major blood vessels under pentobarbital (Vortech Pharmaceuticals, Ltd., Dearborn, MI, USA)-anesthesia (100 mg/kg). Saline solution followed by 4% paraformaldehyde (PFA) solution (Sigma-Aldrich) were used to flush the circulatory system and replace the blood with the fixative solution prior to brain extraction. Circulatory flush was performed using an infusion pump (KD Scientific, Holliston, MA, USA) at a rate of 60 mL/min. The brain was fixed overnight in 4% PFA and then submerged in a 25% glycerol sinking solution for approximately 5 days to prepare tissue for sectioning and preserve cellular structures for further analysis. Extracted brains were frozen in dry ice and cut into 40-µm sections using a microtome (Leica, Nussloch, Germany). TH was used to determine the percentage of dopamine degeneration in the substantia nigra following the 6-OHDA lesion, quantified by optical density (ImageJ, Bethesda, MD, USA). Brain sections were rinsed in a phosphate-buffered saline (PBS) with 0.2% Triton X (Tx), soaked in 3% hydrogen peroxide for 10 minutes, and rinsed and blocked with a 10% natural goat serum in PBS blocking buffer (Vector Laboratories, Newark, CA, USA) for 1 hour at room temperature (23 °C). Brain sections were then incubated in anti-TH antibody (Rabbit) (Millipore Sigma, Darmstadt, Germany) diluted in the blocking buffer overnight at 4 °C. Following PBS-Tx rinsing, sections were incubated with Biotinylated goat anti rabbit Ig (Vector Laboratories) diluted in block (10% natural goat serum) for 1 hour, then rinsed in PBS-Tx and incubated for 1 hour in Avidin-biotin blocking reagent (VECTASTAIN ABC-HRP Kit, Vector Laboratory, Newark, CA, USA). Sections were rinsed in phosphate buffer and incubated in Vector DAB substrate (Vector Laboratory) for 5 minutes until sections had completed staining. Sections were mounted onto glass slides for imaging and brightfield images were collected using an AxioScan microscope (ZEISS, Oberkochen, Germany). Optical density was calculated by first identifying a region of interest around the substantia nigra and measuring the pixel intensity within the region. Next, background levels were subtracted from the pixel intensity values. Finally, optical density was calculated using ImageJ [46]. Electrode location was confirmed using cresyl violet (Sigma-Aldrich) staining.

Data normality was assessed for all data sets using a Shapiro-Wilk test. Nonparametric statistical tests were chosen as the most appropriate approach due to the small sample size and non-normal distribution of the data. All statistical analyses were conducted using GraphPad Prism version 10.0.0 (GraphPad, Boston, MA, USA). For data sets involving more than two unpaired cohorts with a nonparametric distribution, the Kruskal-Wallis test was used to determine if there was a significant difference between the groups. If the Kruskal-Wallis test determined there were no significant differences between the groups (i.e., p $>$ 0.05), then the Kruskal-Wallis statistic H and the p value of the Kruskal-Wallis test were calculated for the experiment. If the Kruskal-Wallis test identified a significant difference between the groups (i.e., p 0.05), then Dunn’s multiple comparison test was used as a post-hoc test to determine which groups were significantly different from each other and the p value for the Dunn’s comparison was shown. The Mann-Whitney test was used to compare nonparametric data between unpaired cohorts, and the results were reported with the Mann-Whitney U statistic and the p value. The Wilcoxon matched pairs signed rank test was applied for paired cohorts, and the results shown with the sum of signed ranks (W) and the p value. Statistical significance was determined at p 0.05. The median and interquartile range are presented on each graph. Outliers were identified using the ROUT method with a Q value of 1% (GraphPad, Boston, MA, USA) and removed from the analysis.

We quantified the percentage of time spent in the open arms of the elevated zero maze for all experimental groups. The chronically stimulated group had a median time spent in the open arms of 22.57%, while the acutely stimulated group had a median time spent of 51.83% (Fig. 2A). No statistically significant difference was observed between any of the experimental groups (H = 5.62, p = 0.132). Additionally, there were no significant differences in the total distance traveled (Fig. 2B, H = 3.89, p = 0.273) or closed arm entries (Fig. 2D, H = 3.80, p = 0.284) between the experimental groups. However, the acutely stimulated group showed significantly more head dips (Fig. 2C, p = 0.017) compared with the vehicle group.

Fig. 2.

Risk-avoidant behavior in the elevated zero maze at day 21. Risk-avoidant behavior for the first 5 minutes that rats spent in the elevated zero maze for each of the experimental groups: chronically stimulated (n = 10), acutely stimulated (n = 10), lesioned but not stimulated (n = 9), and implanted but not-stimulated vehicle (not-lesioned) controls (n = 11). Subthalamic nucleus deep brain stimulation did not evoke significant changes in (A) the time spent in the open arms (H = 5.62, p = 0.132), (B) total distance traveled (H = 3.891, p = 0.273), and (D) entries into the closed arms (H = 3.80, p = 0.284) compared with rats who were not stimulated. However, a significant change was observed in (C) head dipping behavior between stimulated and non-stimulated groups p 0.05 (H = 9.30, p = 0.026). Specifically, acute stimulation resulted in an increase in the number of head dips compared with the vehicle control group (p = 0.017). Data are presented using box and whisker plots. The median is denoted by the line inside the box, while the box edges represent the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. A p value of 0.05 was considered significant using the Kruskal-Wallis test * p 0.05).

We conducted an additional behavioral assessment in the elevated zero maze on the subsequent day, and results were consistent with those observed in the initial exposure to the maze during the previous day. No statistically significant differences (H = 2.58, p = 0.461) in the percentage of time spent in the open arms were observed between experimental groups (Fig. 3A). However, a slight increase in the percentage of time spent in the open arms was observed in the acute group. No significant differences (H = 2.36, p = 0.501) were observed in the total distance traveled between experimental groups (Fig. 3B). Secondary measurements, including the number of head dips and closed arm entries did not result in statistically significant differences (H = 2.15, p = 0.542 and H = 2.73, p = 0.435, respectively) between the experimental groups (Fig. 3C,D).

Fig. 3.

Risk-avoidant behavior in the elevated zero maze at day 22. Risk-avoidant behavior for the 5 minutes that rats spent in the elevated zero maze for each of the experimental groups: chronically stimulated (n = 7), acutely stimulated (n = 11), lesioned but not stimulated (n = 10), and implanted but not-stimulated vehicle (not-lesioned) controls (n = 11). Subthalamic nucleus deep brain stimulation did not evoke significant changes in (A) percentage of time spent in open arms (H = 2.58, p = 0.461), (B) total distanced traveled (H = 2.36, p = 0.501), (C) number of head dips (H = 2.15, p = 0.542), or (D) number of entries into the closed arms (H = 2.73, p = 0.435) compared with rats that were not stimulated. Data are presented using box and whisker plots. The median is denoted by the line inside the box, while the box edges represent the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. A p value of 0.05 was considered significant using the Kruskal-Wallis test.

Notably, significant decreases in the median percentage of time spent in the open arms were observed for rats in the acute stimulation group (W = –55.0, p = 0.002) from 51.83% to 18.00%, the no stimulation group (W = –43.0, p = 0.008) from 26.13% to 7.20%, and the vehicle group (W = –60.0, p = 0.005) from 34.37% to 9.40% (Fig. 4B–D, respectively) after 1 day of prior exposure to the maze. In the open field test, there were no significant differences in the percentage of time spent in the center of the open field (Fig. 5A, H = 2.38, p = 0.498) or total distance traveled (Fig. 5B, H = 1.47, p = 0.689) between the experimental groups. The results for both the elevated zero maze and the open field test remained consistent when accounting for the sex of the animal (Supplementary Figs. 1–6).

Fig. 4.

Percentage time spent in the open arms during second exposure to the elevated zero maze. Rats spent significantly less time (** p 0.01) in the open arms during the second exposure to the elevated zero maze (day 22), suggesting sensitization to the elevated zero maze after a single exposure in all groups except the chronically stimulated group. (A) Chronically stimulated rats (W = –14.0, p = 0.297) (n = 10 on day 21, n = 8 on day 22). (B) Acutely stimulated rats (W = –55.0, p = 0.002) (n = 10 on day 21, n = 11 on day 22). (C) Lesioned rats that were not stimulated (W = –43.0, p = 0.008) (n = 9 on day 2, n = 10 on day 22). (D) Non-stimulated vehicle controls (W = –60.0, p = 0.005) (n = 11 on day 21, n = 11 on day 22). Data are presented using box and whisker plots. The median is denoted by the line inside the box, and the box edges signify the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. A p value of 0.05 was considered significant using the Mann-Whitney test.

Fig. 5.

Risk-avoidant behavior during the first 5 minutes in the open field test. Risk-avoidant behavior during the open field test for each of the experimental groups: chronically stimulated (n = 11), acutely stimulated (n = 16), lesioned but not stimulated (n = 12), and non-stimulated vehicle controls (n = 10). Stimulation did not evoke significant changes in (A) percentage of time spent in the center of the field (H = 2.38, p = 0.498), or (B) total distance traveled (p = 1.47, H = 0.689) compared with rats that did not receive subthalamic nucleus deep brain stimulation. Data are presented using box and whisker plots, the median is denoted by the line inside the box, and the box edges signify the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. Significant changes in risk-avoidant behaviors compared with rats who were not stimulated. A p value of 0.05 was considered significant using the Kruskal-Wallis test.

A separate cohort of rats was evaluated in the elevated zero maze as positive controls to a known anxiogenic agent, caffeine. Rats in the elevated zero maze that received a caffeine injection spent a median time of 30.47% in the open arms, which was significantly less (U = 0, p = 0.029) than those who received a saline injection, which spent a median time of 49.80% in the open arms (Fig. 6A). This behavior remained consistent even after 1 day of prior exposure to the maze (Fig. 6B). On the second day in the maze, the group treated with caffeine spent a median time of 13.13% in the open arms, which was significantly less (U = 2, p = 0.017) than the median time of 43.32% in the open arms spent by the group treated with saline. Rats evaluated in the open field test that received caffeine spent a median time of 0.31% in the center of the open field (Fig. 6C), in contrast to a median time of 4.66% spent in the center of the open field observed in the cohort who received saline (U = 1, p = 0.016).

Fig. 6.

Validation of behavioral paradigm using an anxiogenic control. Percentage of time spent in the open arms of the elevated zero maze (A) during the first exposure (day 21) for the saline group (n = 4) and caffeine group (n = 5). (B) during second exposure (day 22) for the saline (n = 6) and caffeine (n = 5) groups. (C) Percentage of time spent in the center of the open field test for both the saline (n = 5) and caffeine (n = 5) groups. Rats that received caffeine (anxiogenic control) spent significantly less time (U = 0, p = 0.029) in the open arms compared with rats that received saline. This behavior remained consistent during the second exposure to the elevated zero maze (U = 2, p = 0.017). An increase in risk- avoidant behaviors was also observed in the open field test (U = 1, p = 0.016). Caffeine administration evoked anxiogenic behaviors in both the elevated zero maze and open field test. Data are presented using box and whisker plots. The median is denoted by the line inside the box, and the box edges signify the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. A p value of 0.05 was considered significant using the Mann-Whitney test (* p 0.05).

We validated the accurate placement of electrodes within the STN using cresyl violet staining (n = 72) (Fig. 7A, Ref. [39]). Rats with electrodes placed outside of the target STN region (n = 7) were excluded from the study. We confirmed and quantified the extent of the dopaminergic lesion in the substantia nigra pars compacta through TH staining (Fig. 7B,C). Immunohistologic analysis showed a significant (p 0.035) decrease in optical density across all lesioned groups when compared with the vehicle control group (Fig. 7D,E). Furthermore, no statistically significant differences (p $\mathrm{>}$ 0.999 for all comparisons) were observed between the groups that received the lesion (Fig. 7D,E).

Fig. 7.

Histological analysis. Cresyl violet staining was used to confirm electrode location within the subthalamic nucleus (STN) and appropriate lesion level. (A) Typical brain slice staining showing electrode location denoted by the ensuing tissue lesion. The inset shows a magnified view of the electrode location (marked by the black dashed line) and the STN (marked by the purple dashed line) (ML 2.5, AP –3.6, DV –7.8) [39]. The slice shown was collected 3.72 mm posterior to the bregma [39]. Typical example of brain slices showing immunohistology of TH+ cells in the substantia nigra pars compacta in the (B) vehicle and (C) partial 6-OHDA-lesioned cohorts, respectively. Both images shown are located 5.88 mm posterior to the bregma [39]. (D) Optical density analysis of TH+ cells in each experimental group in the (D) left and (E) right hemispheres used to confirm the dopaminergic lesions. For the left hemisphere, analysis showed a decrease in the number of TH+ cells across all lesioned groups compared with the vehicle group (* p 0.05, *** p 0.001). Chronic compared with vehicle (p = 0.035), acute compared with vehicle (p = 0.0005), and no stimulation compared with vehicle (p = 0.0004). No statistically significant differences (chronic compared with acute p $>$ 0.999, acute compared with no stim p $>$ 0.999, chronic compared with no stim p $>$ 0.999) were observed between the groups that received the lesion. For the right hemisphere, analysis showed a notable decrease in the number of TH+ cells across all lesioned groups when compared with the vehicle control group (** p 0.01). Chronic compared with vehicle (p = 0.0025), acute compared with vehicle (p = 0.0018), and no stimulation compared with vehicle (p = 0.0014). No statistically significant differences (chronic compared with acute p $>$ 0.999, acute compared with no stim p $>$ 0.999, chronic compared with no stim p $>$ 0.999) were observed between the lesioned groups. Data are presented using box and whisker plots. The median is denoted by the line inside the box, and the box edges signify the lower (25%) and upper (75%) quartiles. Squares represent female rats and circles represent male rats. A p value of 0.05 was considered significant using the Kruskal-Wallis test followed by Dunn’s multiple comparisons test. ML, Medial Lateral; AP, Anterior Posterior; DV, Dorsal Ventral.