Early-life risperidone alters locomotor responses to apomorphine and quinpirole in adulthood

Abstract

An escalating trend of antipsychotic drug use in children with ADHD, disruptive behavior disorder, or mood disorders has raised concerns about the impact of these drugs on brain development. Since antipsychotics chiefly target dopamine receptors, it is important to assay the function of these receptors after early-life antipsychotic administration. Using rats as a model, we examined the effects of early-life risperidone, the most prescribed antipsychotic drug in children, on locomotor responses to the dopamine D₁/D₂ receptor agonist, apomorphine, and the D₂/D₃ receptor agonist, quinpirole. Female and male Long-Evans rats received daily subcutaneous injections of risperidone (1.0 and 3.0 mg/kg) or vehicle from postnatal day 14–42. Locomotor responses to one of three doses (0.03, 0.1, and 0.3 mg/kg) of apomorphine or quinpirole were tested once a week for four weeks beginning on postnatal day 76 and 147 for each respective drug. The locomotor activity elicited by the two lower doses of apomorphine was significantly greater in adult rats, especially females, administered risperidone early in life. Adult rats administered risperidone early in life also showed more locomotor activity after the low dose of quinpirole. Overall, female rats were more sensitive to the locomotor effects of each agonist. In a separate group of rats administered risperidone early in life, autoradiography of forebrain D₂ receptors at postnatal day 62 revealed a modest increase in D₂ receptor density in the medial caudate. These results provide evidence that early-life risperidone administration can produce long-lasting changes in dopamine receptor function and density.

1. Introduction

The use of antipsychotic drugs in the treatment of pediatric psychiatric disorders has grown substantially over the last 30 years [1–5]. Among these drugs, the second-generation, atypical antipsychotic drug, risperidone, is the most widely used in children [4, 6, 7]. Risperidone has been approved for use in children with autism between the ages of 5–13, yet is most prescribed off-label to children, some younger than 5 years of age, for attention-deficit hyperactivity disorder (ADHD), disruptive behavior disorder, and mood disorders [3, 4, 6, 7, 8]. A major concern regarding the treatment of children with antipsychotic drugs like risperidone is that such treatment might alter brain development and establish a state of atypical neural and behavioral tone that persists into adulthood.

Using rats as an animal model of human development, we have shown that early-life risperidone administration leads to neural and behavioral alterations in adulthood. Rats that receive risperidone, at ages analogous to early childhood through early adolescence in humans, exhibit elevated locomotor activity [9, 10] and decrements in working memory [11] during adulthood. Adult rats administered risperidone early in life are also more sensitive to the rewarding and locomotor effects of the dopamine/norepinephrine agonist, D-amphetamine [12, 13]. These findings suggest that chronic, early-life risperidone administration changes forebrain dopamine function, which is likely related to the high affinity for and antagonism of dopamine D₂ receptors associated with nearly all antipsychotic drugs, including risperidone [14–16].

Previous work has assayed dopamine function in adult rats that were administered antipsychotics early in life. Exposure to the typical antipsychotic and potent D₂ antagonist, haloperidol, during early postnatal life, leads to increases in striatal D₂ receptors [17–19] in adulthood. Later administration (from postnatal days 21–42) of the typical antipsychotic, fluphenazine is associated with more D₂ receptors in the nucleus accumbens, dorsal striatum, and hippocampus [20]. At a functional level, daily injections of haloperidol during the first 20 days of life leads to heightened locomotor responses to the D₁/D₂ receptor agonist, apomorphine, and reduced haloperidol-induced dopamine turnover during adulthood [18, 23]. Conversely, early-life administration of the prototypical, atypical antipsychotic drug, clozapine, which antagonizes both D₂ and 5HT_2a receptors [14, 24], has more limited effects on dopamine receptors and their function in adulthood. Daily clozapine injections over postnatal days 21–42 increase D₄ receptors in the nucleus accumbens and dorsal striatum, but do not substantially modify forebrain D₂ density [20]. Moreover, clozapine administration during the first 20 days of life does not affect behavioral sensitivity to apomorphine in adulthood [25].

Emulating these previous studies, others have considered the effects of early-life risperidone on forebrain dopamine receptor expression. Risperidone administration from postnatal days 21–42 elevates D₁, D₂, and D₄ receptor densities throughout much of the forebrain when examined 24 hours after the last injection [26]. Oral administration of a lower dose of risperidone across roughly the same postnatal period produces more variable and less consistent effects on forebrain dopamine receptor density when assayed two days to two months after treatment cessation [27, 28]. These reports suggest that early-life risperidone initiates a long-term change in the expression of dopamine receptors, particularly those of the D₂ family, that falls somewhere between the effects of typical versus atypical drugs; however, no studies have directly assayed dopamine receptor function in adult rats administered risperidone early in life.

To this end, we assessed locomotor and stereotypy responses to the dopamine D₁/D₂ agonist, apomorphine [21, 22], and the D₂/D₃ agonist, quinpirole [21] in adult rats administered risperidone early in life. Risperidone was administered to young female and male rats daily from postnatal days 14–42. These ages were intended to represent the period when antipsychotic drugs are administered to children [4, 7] spanning from early childhood (see [29] for review) through early adolescence (see [30, 31] for reviews). Female and male rats were studied even though, clinically, boys are two-three times more likely to receive antipsychotics than girls [4, 7]. Nonetheless, since a significant number of girls receive antipsychotic drugs, empirical attention to potential sex differences in outcomes after early-life antipsychotics is merited. This emphasis is further warranted by the significant sex differences observed in the development of forebrain dopamine systems [31, 69, 70] targeted by antipsychotic drugs and by reports of sex-dependent behavioral and neural consequences of early-life antipsychotic drug administration [27, 28, 68].

After habituating the rats to the testing environment for four weeks, locomotor and stereotypy responses to three different doses of apomorphine were recorded once a week for four weeks beginning on postnatal day 76. Five weeks after the last apomorphine injection, locomotor activity and stereotypy were recorded once a week after injection of three different doses of quinpirole. Apomorphine and quinpirole were selected for study because of the well-characterized behavioral effects of each drug elicited by its action at D₁/D₂ and D₂/D₃ receptors respectively [21, 22, 55, 71, 72], and because many previous studies of the long-term behavioral effects of early-life antipsychotics have focused on apomorphine [18, 23, 25]. In a separate group of rats, we measured the effect of early-life risperidone administration on the density of forebrain D₂ receptors during adulthood, since this receptor possesses affinity for risperidone, as well as apormorphine and quinpirole. Relative receptor density was measured in the medial and lateral striatum and nucleus accumbens since these regions are among the primary forebrain targets of antipsychotic drugs like risperidone [73]. Specific subregions of the medial and lateral striatum (anterior and posterior, dorsal and ventral) were assessed, as performed in other studies [80], because distinct behavioral responses to dopaminergic manipulations can vary as a function of subregion [77], and some subregions make precise contributions to specific behaviors in rodents and humans [78, 79].

2. Methods

2.1. Animals and housing

Fifty-nine Long-Evans rats (29 females, 30 males) were used in the behavioral experiments. They were derived from six litters (Envigo, Indianapolis, IN) that arrived at the Department of Psychological Science animal facility at Northern Kentucky University on postnatal day 7.

The relative density of striatal dopamine D₂ receptors was measured in a separate group of rats described in a previously published study of dopamine transporter density [13]. A total of 18 Long-Evans rats (ten females and eight males) were used for this experiment. They were derived from six litters and dams (Envigo, Indianapolis, IN) that arrived on postnatal day 7.

From each litter, 1–2 rats were used per sex per drug treatment. Litters were culled on postnatal day 8 to ten pups: five females and five males. Rats were weaned on postnatal day 21, at which time they were housed two per cage with continuous access to food and water. Within each cage, the two rats belonged to different risperidone dose groups. The lights in the housing room were on between 6:30 and 18:30. All experimental procedures were carried out according to the Current Guide for the Care and Use of Laboratory Animals (USPHS) under a protocol approved by the Northern Kentucky University Institutional Animal Care and Use Committee. A timeline of events for locomotor and receptor binding experiments is depicted in Figure 1.

Figure 1.

Timeline of treatment and testing in the cohorts of rats used in the locomotion and dopamine D₂ receptor experiments. PND = postnatal day, day of birth considered PND 0.

2.2. Drugs

The risperidone doses (1.0 and 3.0 mg/kg), subcutaneous route of administration, and once-a-day injection approach were based on our previous studies [9–13, 32] and other reports demonstrating the effects of these doses on neurotransmitter receptor levels [25, 33, 34] after developmental administration in rats. The 1.0 mg/kg dose reportedly occupies 60–80% of D₂ receptors in the rat forebrain [35] – a degree of receptor blockade associated with antipsychotic efficacy in humans [35]. The 1.0 mg/kg dose also decreases amphetamine-induced hyperactivity in rats by 50% [36]. However, because this dose of risperidone does not consistently produce drug blood levels in adult rats that are close to those reported in adult humans [35], a higher dose (3.0 mg/kg) of risperidone was also studied. This latter dose was also considered because it is some children may be maintained on antipsychotic drug doses that are equivalent to or exceed those recommended for adults [37].

Risperidone was kindly provided by the National Institute of Mental Health’s Chemical Synthesis and Drug Supply program. It was dissolved in a small volume of 10% glacial acetic acid, brought to volume with 0.9% saline, and adjusted to a pH ~6.2 with 1M sodium hydroxide. A vehicle solution, prepared in an identical manner but without risperidone, was administered to control rats. The injection volume was 2.0 ml/kg of body weight. Injections were performed with BD Lo-Dose U-100 Insulin syringes (0.5 ml volume, 28-gauge needle) (Thermo Fisher Scientific, Waltham, MA, USA).

Rats were weighed and administered risperidone or vehicle daily between 8:00 and 14:00 from postnatal day 14–42. A chronic daily injection regimen was studied since many children receive daily antipsychotic drug treatment for years [38–40]. The ages of administration were chosen since they may best approximate the time between early childhood and adolescence in humans [29–31] when children receive antipsychotic drugs [4, 7].

For the studies of behavioral responses to apomorphine and quinpirole, apomorphine (Sigma) was dissolved in saline no more than three hours prior to testing and stored in a light-resistant container. Despite this precaution, it is likely that the apomorphine oxidized at the rate of 3% an hour (Ang et al., 2016) between the time it was dissolved and injected. The apomorphine doses described below have not been corrected for this oxidation rate since some animals were tested soon after the solution was made, while others were tested 1.5–3 hours later. Since an equal number of female and male animals from each of the three risperidone groups were always tested at the same time, apormorphine oxidation should not have differentially affected any group. Quinpirole (Sigma) was dissolved in saline each day of testing.

2.3. Baseline locomotor activity

Testing for baseline locomotor activity began one week after the end of the risperidone injections on postnatal day 49. Each rat was tested twice a week (either Monday/Thursday or Tuesday/Friday) for an hour for four consecutive weeks. Following our previously published methods of testing spontaneous and drug-evoked changes in locomotor activity [9–12], testing occurred between 8:00 and 16:00 in a dark room after each rat was placed in a clear polypropylene cage measuring 25.9 cm wide × 47.6 cm long × 20.9 cm tall. The floor of the cage was covered with wood chip bedding and the cage was placed within Kinder Scientific Smart-Frame activity monitoring frames (Kinder Scientific Inc., Poway CA). The number of photobeam breaks was tabulated every five minutes. Cages were cleaned once a week throughout testing. Rats were tested in the same cage each week. Half the cages were dedicated to the female rats and the other half to the males.

2.4. Locomotor activity after apomorphine and quinpirole administration

Between postnatal days 76 and 107, locomotor and stereotypy responses to apomorphine were tested once a week for one hour for four consecutive weeks. Each rat was also tested one additional day per week after a saline injection with the intended goal of diminishing associations between the test cage and active drug injections, although the efficacy of this approach was not tested statistically. There was at least 48 hours between these two weekly tests. On the active drug days, each rat received a subcutaneous injection of saline or one of three doses of apomorphine (0.03, 0.1, or 0.3 mg/kg). Rats were placed directly in the activity chamber immediately after injection. The order of apomorphine doses (including the saline injection) was counter-balanced across the four-week period within the risperidone and vehicle groups, and the female and male groups. The number of photobeam breaks measured every five minutes after the saline injection portion of this dose regimen was used as a covariate in the analyses of covariance (ANCOVA) described below. This was intended to account for group differences in response to injection alone in interpreting any group differences in response to the different apomorphine doses.

On days when the rats received apomorphine injections, they were observed for signs of stereotypy by evaluators blind to whether the rats had previously received risperidone or vehicle, or apomorphine or saline. Every 15 minutes, each rat was observed for 20 seconds for signs of stereotypy that included rearing, grooming, sniffing/chewing, or head weaving. Each animal received a score between 0 and 3 with 0 indicative of no stereotypy, 1 indicating that stereotypy were present < five seconds, 2 indicating that stereotypy was present for > 5 but < 15 seconds, and 3 indicating that stereotypy was present > 15 seconds.

After apomorphine testing, rats were given a 35-day washout period and then locomotor/stereotypy responses to quinpirole were recorded each week between postnatal days 147 and 171 in a manner identical to that reported for behavioral responses to apomorphine. The only difference was that, on the experimental treatment days, each rat received a subcutaneous injection of saline or a one of three doses of quinpirole (0.03, 0.1, or 0.3 mg/kg). As done earlier, rats were placed directly in the activity chamber immediately after injection, and the order of quinpirole injections was counter-balanced across the four-week period within the risperidone and vehicle groups, and the female and male groups. An ANCOVA was performed as described above using locomotor responses to saline injection as a covariate in the analyses of locomotor responses to each quinpirole dose.

2.5. Dopamine D₂ receptor binding

In the experiment assessing the effects of chronic early-life risperidone on forebrain D₂ receptor binding, nine rats (5 females, 4 males) received daily injections of 3.0 mg/kg of risperidone and nine rats (5 females, 4 males) received vehicle from postnatal day 14–42. This study was limited to the 3.0 mg/kg risperidone dose only since this dose had a greater impact on locomotor responses to apomorphine and quinpirole. On postnatal day 62, brains were collected via rapid decapitation without anesthesia. Each brain was hemi-dissected, and the left hemisphere placed in powdered dry ice and stored at −80° C. Brains were sliced on a Leica CM1850 cryostat (Leica Biosystems, Deer Park, IL, USA) into a series of 16-μm-thick sagittal sections, thaw-mounted onto Superfrost Plus slides (Thermo Fisher Scientific, Waltham, MA, USA), and dried overnight in a desiccator at 4°C. Slides were stored at −80°C until use.

Autoradiography was performed based on previous methods (Tu et al., 2000) by slowly bringing the slides to room temperature and then preincubating them for 30 minutes at 20°C in a buffer containing 50 mM Tris HCl and 150 mM NaCl. The slides were then incubated for 90 minutes at 20°C in a buffer containing 50 mM Tris HCl, 150 mM NaCl, 5mM KCl, 2mM CaCl₂, 1mM MgCl₂, 0.1% ascorbic acid and 2nM [³H]raclopride (Perkin-Elmer Life Sciences, Boston, MA, USA). No blocking agent or control buffer solution lacking [³H]raclopride was used on contiguous sections. Slides were then washed six times for 1 minute with the preincubation buffer at 4°C, followed by a 10 second dip in 10% preincubation buffer at 4°C, and another 10 second dip in double distilled water at 4°C. After washing, the tissue was dried under a gentle stream of ambient air and placed in a vacuum desiccator overnight at room temperature. Tissue was then exposed to RayMax Beta High Performance Autoradiography Film (ICN Biomedicals, Aurora, OH, USA) for four weeks. All films were processed using Kodak GBX developer (Kodak, Rochester NY, USA) and images of the films were captured and converted into TIFF files.

The optical density of the TIFF files containing regions of the lateral and medial caudate and nucleus accumbens was quantified using ImageJ software (v 2.0). These regions (Figure 2) were located at roughly + 2.4 and + 3.9 mm lateral to midline, respectively (analogous to plates 82/82a and 85/85a of Paxinos and Watson [45]). Raters were blind to the treatment status and sex of each brain. Mean uncalibrated optical density was calculated for each individual region from a series of 4–6 contiguous sections from the lateral and medial forebrain areas. A region just dorsal to the caudate was used to measure non-specific binding as this region should represent the corpus callosum which lacks D₂ receptors. This value was subtracted from the optical density measures for each region of interest, and the product used for statistical analyses.

Figure 2.

Images of [³H]raclopride binding in the medial (A.) and lateral (B.) forebrain and graphical depictions of subregions sampled within each section to obtain the optical density measures. The numbers in parentheses are the distance of the slice from midline (drawings and numbers based on plates # 82/82a and 85/85a of Paxinos and Watson [45]). A – anterior caudate, P – posterior caudate, D – dorsal caudate, V - ventral caudate, N - nucleus accumbens, NSB – non-specific binding in the corpus callosum.

2.6. Statistical analyses

From postnatal day 49 – 74, baseline locomotor activity (i.e., photobeam breaks) was recorded for 60 minutes twice a week across four weeks. The totals from the two weekly sessions were averaged and compared using a three-way, repeated-measures analysis of variance (ANOVA) with risperidone (vehicle, 1.0, and 3.0 mg/kg) and sex as between-groups factors, and week as a within-groups factor. For the apomorphine and quinpirole experiments, the locomotor activity recorded for one hour after the saline injection was compared separately for each experiment between groups using a three-way, repeated measures ANOVA with risperidone (vehicle, 1.0, and 3.0 mg/kg) and sex as between-groups factors, and time across the 60-minute session as a within-group factor.

Locomotor responses to each dose of apomorphine and quinpirole were assessed using a repeated-measures analysis of covariance (ANCOVA). In each analysis, the locomotor activity recorded during each five-minute time-period after the saline injection described in the last paragraph served as a covariate for the locomotor activity recorded during the corresponding five-minute time-period following each dose of apomorphine. The same approach was used to analyze the quinpirole locomotor data. In each experiment, the effects of risperidone (vehicle, 1.0, and 3.0 mg/kg) and sex were treated as between-groups factors, and the effect of time across the 60-minute session was treated as a within-group factor. Separate ANCOVAs were performed for each apomorphine and quinpirole dose.

Stereotypy scores recorded every 15 minutes after drug injections were summed across the one-hour testing period. Scores observed in the apomorphine and quinpirole experiments were analyzed in separate three-way repeated-measures ANOVAs for each experiment with risperidone (vehicle, 1.0, and 3.0 mg/kg) and sex as between-groups factors, and apomorphine or quinpirole dose as a within-group factor.

In the D₂ receptor binding experiment, the relative average optical density within each brain region was compared using a two-way ANOVA using risperidone (vehicle, 3.0 mg/kg) and sex as between-group factors.

All model assumptions were assessed using a combination of residual plots and formal testing. When overall model interaction effects of interest were significant, relevant pairwise comparisons were performed using Fishers LSD test. Across all analyses, differences were considered statistically significant if the corresponding p value was < 0.05. Analyses were performed using Jamovi (v 2.3.18) [76] and SAS 9.4 (SAS Institute Inc.; Cary, NC, USA).

3. Results

3.1. Baseline locomotor activity

Beginning one week after the cessation of daily risperidone injections, locomotor activity was recorded for one hour twice a week for four weeks, and the total activity from the two weekly sessions was averaged into one score. When comparing these scores across the four weeks of testing, there were significant main effects of week [F(3, 159) = 3.16, p < 0.001], sex [F(1, 53) = 17.89, p < 0.001], and risperidone group [F(2, 53) = 3.20, p = 0.05] but no interactions between any of these variables. Locomotor activity decreased across the four weeks of testing, and females were more active than males during each week (Figure 3). Post-hoc analyses indicated that rats in the 3.0 mg/kg risperidone group were significantly more active than vehicle controls during weeks 1 and 3 (each p < 0.05).

Figure 3.

Effects of early-life risperidone (Risp) on locomotor activity recorded over four weeks beginning on postnatal day 49. Data for each week were averaged from the two weekly, 60-minute sessions and represent total mean (+ S.E.M.) photobeam breaks for each week. Differences (p < 0.05) between rats administered 3.0 mg/kg of risperidone and vehicle rats in weeks 1 and 3 indicated by *. Females (circles) were significantly more active than males (squares) as indicated by♀just above the X axis, and activity decreased significantly across weeks. n = 10/10 females/males each for the Risp 3.0 and Vehicle groups, n = 9/10 females/males for the Risp 1.0 group.

3.2. Locomotor responses to apomorphine

Analyses of the locomotor responses to saline that served as part of the dose-response study of apomorphine revealed statistically significant effects of time within session [F(11, 583) = 133.74, p < 0.001], sex [F(1, 53) = 17.45, p < 0.001], and risperidone group [F(2, 53) = 3.24, p = 0.05], but no interactions between any of these variables. As seen previously, locomotor activity decreased significantly across the one-hour test session (Figure 4A). Post-hoc comparisons indicated that females were significantly more active than males between 20 and 50 minutes after injection (each p < 0.05). Additionally, rats administered 3.0 mg/kg risperidone early in life were significantly more active than vehicle rats at 15 and 35 minutes after injection (each p < 0.05), with a trend (p = 0.07) towards greater activity in the former group at 40 minutes.

Figure 4.

Effects of early-life risperidone (Risp) on locomotor activity after apomorphine (Apo) in female (circles) and male (squares) rats. Data at each point represent mean (+ S.E.M.) photobeam breaks for each group after injection of A.) saline, B.) 0.03 mg/kg of apomorphine, C.) 0.1 mg/kg of apomorphine, or D.) 0.3 mg/kg of apomorphine. Group differences (p < 0.05) at single time points between vehicle rats and those previously administered 1.0 mg/kg or 3.0 mg/kg of risperidone respectively denoted by + or *. Group differences (p < 0.05) limited to comparisons of female or male rats between the three groups indicated by +♀, *♀, or *+♀, or +♂, *♂, or *+♂, respectively. In D., the shaded symbols (*♀) indicate group differences (p < 0.05) based on a trend (p = 0.07) towards a three-way interaction between risperidone group, sex, and time. Time points in which females were more active than males are indicated by ♀ near the X-axis. n = 10/10 females/males each for the Risp 3.0 and Vehicle groups, n = 9/10 females/males for the Risp 1.0 group.

When comparing group responses to each of the three apomorphine doses, it seemed important to adjust for the differences in responses to saline alone seen between the risperidone groups as well the female and male rats. Accordingly, the locomotor activity seen in each rat at each time point after saline injection was used as a covariate in a repeated-measures ANCOVA that compared the effects of sex and risperidone on locomotor responses at each corresponding time point to each apomorphine dose. Separate ANCOVAs were conducted for each of the three doses (0.03, 0.1, and 0.3 mg/kg) of apomorphine.

Following injection of the low dose (0.03 mg/kg) of apomorphine, an ANCOVA revealed a significant three-way interaction between time, sex, and risperidone group on levels of locomotor activity [F(22, 53) = 2.04, p = 0.02]. Females were significantly more active than males at each time point (each p < 0.05, Figure 4B). As suggested by the three-way interaction, the effects of risperidone varied depending on the time interval and sex of the rat. Female rats from the 3.0 mg/kg risperidone group were significantly more active than female vehicle rats at 5, 20, 25, and 50 minutes after apomorphine injection, while the female rats from the 1.0 mg/kg risperidone group were significantly more active than female vehicle rats at 30 minutes (each p < 0.05). Male rats in the 3.0 mg/kg risperidone group were significantly more active than male vehicle rats at 15 minutes post-apomorphine injection (p < 0.05).

A significant three-way, time × sex × risperidone group interaction also emerged from the ANCOVA conducted on the locomotor responses to the middle dose (0.1 mg/kg) of apomorphine [F(22, 53) = 2.13, p = 0.01]. Except for the first five minutes of testing, female rats were significantly more active than male rats at every time point (each p < 0.05; Figure 4C). Female rats in the 3.0 mg/kg risperidone group were more active than female vehicle controls at 20, 25, 30, 35, 40 and 45 minutes after apomorphine injection, and the females in the 1.0 mg/kg risperidone group were more active than vehicle controls at 40 minutes (each p < 0.05). Male rats in the 3.0 mg/kg risperidone group were more active than male vehicle controls at 25 minutes after apomorphine injection, and the males in the 1.0 mg/kg risperidone group were more active than male vehicle controls at 30 minutes (each p < 0.05).

The ANCOVA conducted on the locomotor responses to the high (0.3 mg/kg) apomorphine dose indicated a significant time × sex interaction [F(11, 53) = 6.28, p < 0.001]. Female rats were more significantly active than male rats between 25 and 60 minutes post-apomorphine injection (each p < 0.05). Additionally, there was a trend towards a significant three-way interaction between time × sex × risperidone group [F(22, 53) = 1.66, p = 0.07]. Pairwise comparisons of female and male risperidone groups at each time point suggested that this trend was based on significantly greater activity in the 3.0 mg/kg risperidone group of female rats relative to the vehicle group of female rats between 45 and 55 minutes after apomorphine injection (each p < 0.05).

3.3. Locomotor responses to quinpirole

After a five-week washout period, locomotor responses to quinpirole were tested using the same design used to assess the locomotor responses to apomorphine. Unlike the outcome in the apomorphine experiment, locomotor responses to the saline injections performed as part iof the quinpirole experiments did not differ between the risperidone and vehicle groups across the one-hour testing period (Figure 5A). There was a significant decrease in activity over time [F(11, 583) = 135.74, p < 0.01], and females were more active than males [F(1, 53) = 15.77, p < 0.001]. Post-hoc comparisons at each time point indicated that females were significantly more active than males at each time point except at 5 and 50 minutes post injection (each p < 0.05).

Figure 5.

Effects of early-life risperidone (Risp) on locomotor activity after quinpirole (Quin) in female (circles) and male (squares) rats. Data at each point represent mean (+ S.E.M.) photobeam breaks for each group after injection of A.) saline, B.) 0.03 mg/kg of quinpirole, C.) 0.1 mg/kg of quinpirole, or D.) 0.3 mg/kg of quinpirole. Group differences (p < 0.05) at single time points between vehicle rats and those previously administered 3.0 mg/kg of risperidone denoted by *, and between the latter group and rats previously administered 1.0 mg/kg of risperidone denoted by ☨. In C., shaded symbols (*) indicate group differences (p < 0.05) based on a trend (p = 0.07) towards a two-way interaction between risperidone group and time. Time points in which females were more active than males are indicated by ♀ near the X-axis. n = 10/10 females/males each for the Risp 3.0 and Vehicle groups, n = 9/10 females/males for the Risp 1.0 group.

Even though there were no significant effects of risperidone on the locomotor activity seen after saline injection, separate ANCOVAs were still conducted for the locomotor data obtained after each dose (0.03, 0.1, & 0.3 mg/kg) of quinpirole. In these analyses, the level of activity seen in each animal at each time point after saline injection was used as the covariate for the level of activity seen in each animal at each corresponding time point after the injection of each dose of quinpirole.

Main effects of time [F(11, 571) = 30.39, p < 0.001], sex [F(1, 52) = 23.95, p < 0.001], and risperidone group [F(2, 52) = 5.26, p = 0.008] emerged from the ANCOVA assessing locomotor responses to the low (0.03 mg/kg) dose of quinpirole. Activity decreased across the test session (Figure 5B). Female rats were more significantly active than males at 10, 20, 25, 35, 55, and 60 minutes (each p < 0.05). The 3.0 mg/kg risperidone group was significantly more active than the vehicle group at 20 minutes post-injection (p < 0.05).

The locomotor activity observed after the middle dose (0.1 mg/kg) of quinpirole was marked by a significant time × sex interaction [F(11, 53) = 2.30, p = 0.02]. Female rats were significantly more active than male rats at 5, 10, 35, 40, 50, 55, and 60 minutes after injection (each p < 0.05; Figure 5C). There was also a trend towards a significant time × risperidone interaction [F(22, 53) = 1.65, p = 0.07]. Post-hoc pairwise comparisons suggested that this trend was driven by greater levels of activity in the 3.0 mg/kg risperidone group compared to the vehicle group between 50–60 minutes post-injection (each p < 0.05).

The high dose (0.3 mg/kg) of quinpirole caused a bi-phasic change in activity with decreases seen at 10–15 minutes post-injection and increases seen after 20 minutes post-injection – a pattern that was most marked in the female rats as indicated by a time × sex interaction [F(11, 52) = 2.56, p = 0.01]. Female rats were significantly more active than males between 20 and 60 minutes after injection (each p < 0.05; Figure 5D). There was also a time × risperidone interaction [F(22, 52) = 1.91, p = 0.03]. Pairwise comparisons at each time point revealed that rats in the 3.0 mg/kg risperidone group were significantly more active than rats in the 1.0 mg/kg risperidone group at 20 minutes post-injection (p < 0.05).

3.4. Stereotypy after apomorphine and quinpirole

During the locomotor testing performed after the apomorphine and quinpirole injections, stereotypy was rated every 15 minutes for 20 seconds and the ratings across each of these four time points were summed into a total score for each drug dose. Stereotypy was significantly elevated after injection of the highest dose of apomorphine [dose effect: F(3, 159) = 49.63, p < 0.001], and quinpirole [dose effect: F(3, 147) = 45.26, p < 0.001] when compared to saline and the lower two doses of each respective drug (each p < 0.05; Figures 6A & 6B). In each experiment, female rats displayed significantly more stereotypy than males [F (1, 53) = 7.05 & (1, 49) = 10.90, p < 0.01 & 0.001, respectively for apomorphine and quinpirole]. There was also a significant interaction between quinpirole dose and risperidone on stereotypy [F(6, 147) = 2.44, p = 0.03]. After injection of the 0.1 mg/kg quinpirole dose, rats in the risperidone 3.0 mg/kg group displayed significantly more stereotypy than vehicle rats (p < 0.05). There was also a trend towards greater stereotypy in the risperidone 1.0 mg/kg group versus the vehicle group (p = 0.08; Figure 6B).

Figure 6.

Effects of early-life risperidone on stereotypy after apomorphine (A. Apo) or quinpirole (B. Quin). Data represent mean (+ S.E.M.) stereotypy scores summed over one hour after injections of each drug. Stereotypy was significantly greater after injection of the 0.3 mg/kg dose of apomorphine or quinpirole when compared to the effects of the lower two doses of each drug or saline as denoted by #. Females demonstrated significantly more stereotypy than males in both experiments. In B., a difference (p < 0.05) between the vehicle and 3.0 mg/kg of risperidone groups was seen after injection of the 0.1 mg/kg dose of quinpirole (denoted by *). There was a trend (p = 0.08) towards a difference between the vehicle and 1.0 mg/kg risperidone groups (denoted by the shaded +). n = 10/10 females/males each for the Risp 3.0 and Vehicle groups, n = 9/10 females/males for the Risp 1.0 group.

3.5. Dopamine D₂ receptor binding

Dopamine D₂ receptor binding was quantified in a separate cohort of young adult female and male rats administered either vehicle or 3.0 mg/kg of risperidone early in life. Tissue was collected twenty days after the cessation of risperidone administration (Figure 1) and [³H]raclopride was used to label D₂ receptors. D₂ receptor density was defined as the optical density of the region of interest minus the optical density of the corpus callosum (Figure 2). Based on this measure, D₂ receptor density was significantly greater in the 3.0 mg/kg risperidone group in the medial anterior caudate [F(1, 14) = 4.26, p = 0.05] (Table 1). Females possessed higher densities of D₂ receptors than males in most areas, including the medial anterior, dorsal, and ventral caudate, and the lateral anterior and posterior caudate [F (1, 14) = 51.69, 36.11, 18.24, 29.14, & 28.21, p < 0.0001 – 0.01 for each respective region], but there were no statistically significant interactions between sex and risperidone administration in any region.

Table 1.

Effects of early life risperidone on relative optical density of striatal [³H]raclopride binding.

	Vehicle		Risperidone 3.0 mg/kg
	Female	Male	Female	Male
Medial striatum
Dorsal caudate	37.6 + 3.8	20.1 + 1.3	43.3 + 2.4	24.3 + 3.6*
Anterior caudate	42.3 + 3.7	24.4 + 1.7	50.1 + 1.0	28.1 + 3.6 *
Posterior caudate	2.4 + 3.0	2.8 + 2.2	9.9 + 3.0	1.8 + 1.7
Ventral caudate	34.3 + 4.4	18.7 + 4.6	38.7 + 4.0	19.0 + 2.9*
Nucleus accumbens	11.8 + 6.0	5.6 + 2.7	18.8 + 5.1	5.9 + 1.6
Lateral striatum
Anterior caudate	51.1 + 2.5	27.7 + 3.6	53.5 + 5.3	33.5 + 3.6*
Posterior caudate	25.0 + 2.4	9.3 + 2.3	24.9 + 2.6	12.0 + 3.3*

Note: Data represent mean (+ S.E.M.) relative optical density. n = 5 females and 4 males per drug group. Boldface indicates significant difference between drug groups (p = 0.05). An * indicates sex difference within listed region (p < 0.05).

4. Discussion

These experiments show that chronic risperidone administration early in life significantly elevates locomotor sensitivity to the D₁/D₂ agonist, apomorphine, during adulthood. These effects were observed in rats of both sexes but more frequently and robustly in females. Chronic risperidone administration early in life also significantly increased locomotor and stereotypy responses to D₂/D₃ agonist quinpirole but in a more limited manner when compared to the effects of early life risperidone on apomorphine. In line with previous work [42, 43], the two higher doses of apormorphine and quinpirole exerted biphasic effects on locomotion – an early decrease in mobility followed by elevated activity. The initial suppressive effects of apomorphine and quinpirole have been attributed to each drug’s affinity for pre-synaptic dopamine receptors [21] (although see [44]) presumably of the D₂ subtype [45]. The later elevation in locomotor activity produced by apomorphine may reflect its ability to rapidly desensitize presynaptic dopamine receptors and partially agonize post-synaptic D₁ receptors [21, 22, 46], while the same effect produced by quinpirole may be due to its direct activation of post-synaptic D₂/D₃ receptors [47, 48] or indirect stimulation of post-synaptic D₁ receptors [49]. In the present study, it is possible that one or more of these dopaminergic mechanisms were sensitized by early-life risperidone.

The ability of early-life risperidone to enhance the locomotor stimulating effects of apomorphine and quinpirole aligns with previous studies of early-life administration of typical and atypical antipsychotics. Adult rats administered the typical antipsychotic, haloperidol, during the first three postnatal weeks display potentiated responses to the locomotor activating effects of apomorphine [17, 18]. This effect is not observed in adult rats administered the atypical antipsychotic, clozapine, during the same postnatal period [25]. From a clinical and pharmacological perspective, risperidone is often categorized as an atypical antipsychotic, but seems to acquire characteristics of a typical antipsychotic (e.g., produces motor side effects) at higher doses [50, 51]. The data here fit this latter pattern since the effects of the higher risperidone dose were more pronounced and frequent than those of the lower risperidone dose. Overall, the impact of early-life risperidone seen here could be interpreted as falling between the previously reported effects of early-life haloperidol and clozapine.

These changes in behavioral sensitivity suggest that forebrain dopamine receptors may be altered by early-life risperidone. To that end, adult rats administered risperidone early in life were found to possess a greater density of D₂ receptors in the anterior medial caudate. It should be noted that our autoradiography methods did not include control sections that were incubated in [³H]raclopride and excess unlabeled D₂ ligand, or sections simply incubated without [³H]raclopride, which raises concerns about the specificity of the [³H]raclopride binding to D₂ receptors in our study. However, we feel that these data are worth reporting given that 1) the observed binding was relatively restricted to the striatum (see Figure 2) as seen in other autoradiography studies of [³H]raclopride [81, 82], 2) a non-specific binding site was used immediately dorsal to the striatum in an area consistent with the location of the corpus callosum, and 3) our results seem consistent with previous work on early life antipsychotics. For example, Moran-Gates et al. [26] found significant elevations in forebrain D₂ receptor density after chronic administration of similar risperidone doses from postnatal days 22–41. Similar effects have been observed after early-life treatment with haloperidol [17], fluphenazine [20], and olanzapine [52, 53]. However, studies [27, 28] in which rats were administered relatively lower doses of oral risperidone or olanzapine during the same developmental period used in the present study did not find changes in forebrain D₂ receptor density when assessed two or 56 days after the cessation of antipsychotic administration. These latter data allude to a potential influence of several factors on the sensitivity of dopaminergic systems to early-life antipsychotics including the dose and route of administration, antipsychotic type (typical or atypical), and the post-administration interval or animal age. Our data fits with this latter factor since a relatively modest change in D₂ receptors was seen after a 20-day interval between the end of risperidone administration and tissue collection.

Rats that received risperidone early in life were not only more sensitive to the locomotor effects of apomorphine and quinpirole but demonstrated more spontaneous locomotor activity when tested during the first month after the end of risperidone treatment. This finding is consistent with our previous work [9–13]. Some of these group differences were still observed after saline injections during the course of the apomorphine testing. The persistence of this effect mirrors our earlier work [9] where hyperactivity related to early-life risperidone was observed up to nine months of age. Because of this phenomenon, locomotor responses to each dose of apomorphine and quinpirole were subjected to an analysis of covariance that controlled for each animal’s response to saline injections embedded within each respective experiment. This should raise confidence that our conclusion that early-life risperidone elevates sensitivity to apomorphine and quinpirole in a manner independent of group differences in baseline activity.

Early-life risperidone did not alter stereotypy induced by apomorphine. This is somewhat surprising given the significant group differences in locomotor responses to this drug. Also, Cuomo et al. [23] found that early postnatal haloperidol administration led to greater apomorphine-induced stereotypy at 60 days of age. However, our results are consistent with previous work showing that early-life risperidone, while enhancing later locomotor sensitivity to amphetamine, does not alter amphetamine-induced stereotypy [12] and that apomorphine-induced stereotypy is not affected by early postnatal administration of the atypical antipsychotic, clozapine [25]. Rats in the high dose risperidone group did exhibit higher stereotypy scores than controls after administration of the middle quinpirole dose. The interpretation of this data merits caution since this group difference reflects a decrease in stereotypy by controls as much as an increase by the risperidone group. But the outcome does fit with the higher density of D₂ receptors seen in the risperidone group and quinpirole’s more direct action at such receptors. The mixed data on stereotypy supports the idea that the impact of early-life risperidone on later behavior bears features of both typical and atypical antipsychotics.

A somewhat novel finding was that female rats were more sensitive to the locomotor activating effects of apomorphine and quinpirole, even when controlling for sex differences in baseline activity. Simpson et al. [56] reported greater sensitivity in female rats to the locomotor suppressive effects of high doses of apomorphine. They suggest that this effect may reflect an enhancement of dopaminergic synaptic activity by estrogen. Our finding of increased forebrain D₂ receptor density in female rats is consistent with this idea, but it should be emphasized that other studies of forebrain D₂ density in adult rats have not been observed such sex differences [57–59]. Our unique observation of greater D₂ density in females may stem from the timing of our brain tissue collection on postnatal day 62, an age when sex-dependent, peripubertal fluctuations in D₂ density may linger [57].

Recently, more attention has been directed towards potential sex-specific effects of early-life antipsychotic administration on later behavioral and biochemical outcomes [27, 28, 68]. Other studies have shown that early-life risperidone leads to greater D₂ receptor density in the hippocampus of adult female rats, whereas it decreases D₁ receptors in the nucleus accumbens core, prefrontal and cingulate cortex of adult male rats [27, 28]. At a behavioral level, De Santis and colleagues [68] reported that adult male rats, but not adult female rats, administered risperidone early in life are more active and less anxious. Our results also revealed sex-specific effects of early-life risperidone, namely that female rats administered risperidone early in life more consistently displayed greater levels of apomorphine-induced locomotor activity relative to controls than male rats administered risperidone early in life. While it is difficult to identify a mechanism that explains sex-dependent responses to early-life antipsychotic drugs, it is likely related to one of several sex differences in the development of various facets of forebrain dopamine function (see Andersen [69] for review]. Of particular relevance to antipsychotic antagonism of D₂ receptors is the observation that male rats experience a greater surge in striatal D₂ receptor density between puberty and adolescence when compared to females [70]. It is possible that early-life risperidone administration ultimately impacts dopaminergic sensitivity during adulthood in a sex-dependent manner by disrupting the different developmental trajectories seen in female versus male rats.

One main caveat to this study was that the behavioral assessments of quinpirole occurred after the apomorphine studies. This raises concerns about how exposure to the latter drug influenced responses to the former drug, as well as the age difference between the two periods of drug exposure. This latter concern might explain why early-life risperidone elicited far fewer effects on locomotor responses to quinpirole as opposed to apomorphine. However, there are features of our approach that may assuage this concern. First, an extensive, 40-day washout interval was used between the two periods of drug testing. Second, active drug exposure during the testing weeks was limited and infrequent. During each week of drug testing, rats were tested twice after receiving an injection with four days between each test. One of the injections was always a saline injection that was intended to diminish the development of associations between the testing cage and active drug effects. For the other injection, apomorphine or quinpirole were administered within groups in a counter-balanced manner, and at doses that elicited the intended behavioral effects but were lower than those used in previous work [25, 44, 54, 56]. By spacing drug administration and using minimally effective doses, we hoped to limit drug sensitization or tolerance within or between drugs. Indeed, there was no evidence of time-dependent changes in responses to either drug or saline across the four weeks of testing as indicated by secondary data analyses. Finally, there appeared to be a reduction in the amount of locomotor activity observed after saline test injection across quinpirole testing weeks when compared to the amount seen across the apomorphine testing weeks (compare Figure 5A to Figure 4A). This could reflect the effect of age differences or habituation to the testing conditions across time. Because we used data from the saline tests as a baseline control for the data generated after active drug injections, this approach should have controlled for the effects of age or habituation on responses to either drug.

Two other limitations deserve mention. First, all litters were shipped to our facilities during the first postnatal week. This likely served as a stressor to the dams and litters and may have influenced how risperidone affected later behavior, given the literature on the effects of early-life stress on later neural and behavioral function [see 74, 75 for reviews]. On the other hand, previous work [11, 13] from our lab has used litters raised from dams shipped during the second week of pregnancy and yielded similar effects of risperidone on baseline locomotor behavior during adulthood, suggesting that the effects reported here are not completely attributable to postnatal stress. A second limitation to this study is the lack of an adult comparison group (i.e., adult rats administered chronic risperidone and thereafter assessed for locomotor responses to apomorphine and quinpirole). Without such a group and appropriate age-matched control, we are unable to conclude that the findings reported here are specific to risperidone administration during early postnatal development. However, given the absence of research regarding the long-term effects of early-life antipsychotic treatment, we feel that the data reported here remain of interest.

By using direct agonists for dopamine receptors as well as receptor binding methods, our study revealed significant modifications of dopamine receptor function and density in adult rats administered risperidone early in life. These changes may help to explain some of our previous observations of altered behavior in these rats during adulthood such as elevated locomotor activity [9, 10], impairments in working memory [11], and enhanced sensitivity to the locomotor and reinforcing effects of amphetamine [12, 13]. They may also have implications for the use of antipsychotic drugs in children and suggest that such treatment has the potential to generate a state of dopaminergic supersensitivity that can last into adulthood. This is of particular concern because: 1) increases in the sensitivity of forebrain dopamine systems by antipsychotic drug treatment in adults have the potential to increase the risk for substance misuse and use disorders [60], and 2) antipsychotic drugs are most likely to be prescribed to children with ADHD and disruptive behavioral disorders [4, 6, 8] – populations at greater risk for later substance use disorders [64–66]. Like preclinical efforts aimed at lessening the side effects of antipsychotic drugs in adult psychiatric disorders [73], it will be important to identify the long-term consequences of antipsychotic drug use during development to better inform their use in children and minimize potential adverse effects. Further research that specifically isolates the neurobiological impact of developmental antipsychotic administration will be an important part of realizing this goal.

Highlights:

• Early-life risperidone enhances locomotor responses to apomorphine in adulthood.

• Effects of early-life risperidone on apomorphine are greater in females.

• More limited effect of early-life risperidone on locomotor responses to quinpirole.

• Locomotor activating effects of apomorphine and quinpirole are greater in females.

• Adult rats administered risperidone early in life possess more striatal D₂ receptors.