Olanzapine depot formulation in rat: a step forward in modelling antipsychotic-induced metabolic adverse effects Free

Abstract

Rats are used as animal models in the study of antipsychotic-induced metabolic adverse effects, with oral drug administration yielding hyperphagia, weight gain and, in some cases, lipogenic effects. However, the rapid half-life of these drugs in rats, in combination with development of drug tolerance after a few weeks of treatment, has limited the validity of the model. In order to prevent fluctuating drug serum concentrations seen with daily repeated administrations, we injected female rats with a single intramuscular dose of long-acting olanzapine formulation. The olanzapine depot injection yielded plasma olanzapine concentrations in the range of those achieved in patients, and induced changes in metabolic parameters similar to those previously observed with oral administration, including increased food intake, weight gain and elevated plasma triglycerides. Moreover, the sensitivity to olanzapine was maintained beyond the 2–3 wk of weight gain observed with oral administration. In a separate olanzapine depot experiment, we aimed to clarify the role of hypothalamic AMP-activated protein kinase (AMPK) in olanzapine-induced weight gain, which has been subject to debate. Adenovirus-mediated inhibition of AMPK was performed in the arcuate (ARC) or the ventromedial hypothalamic (VMH) nuclei in female rats, with subsequent injection of olanzapine depot solution. Inhibition of AMPK in the ARC, but not in the VMH, attenuated the weight-inducing effect of olanzapine, suggesting an important role for ARC-specific AMPK activation in mediating the orexigenic potential of olanzapine. Taken together, olanzapine depot formulation provides an improved mode of drug administration, preventing fluctuating plasma concentrations, reducing handling stress and opening up possibilities to perform complex mechanistic studies.

Introduction

Several atypical antipsychotic drugs, in particular olanzapine and clozapine, are associated with metabolic adverse effects such as obesity, dyslipidemia, and derangements in glucose metabolism, which contribute to the increased risk for cardiovascular disorders observed in patients suffering from schizophrenia (American Diabetes Association, 2004; Daumit et al., 2008; McGrath et al., 2008). Rodents are commonly used to model antipsychotic-induced metabolic adverse effects, with hyperphagia and weight gain consistently replicated after subchronic olanzapine administration to female rats, either orally or via injections (Albaugh et al., 2006; Baptista et al., 1993; Choi et al., 2007; Goudie et al., 2002; Minet-Ringuet et al., 2006; Skrede et al., 2012a). Still, certain metabolic disturbances, such as antipsychotic-induced dyslipidemia, have not been convincingly modelled in rat (Boyda et al., 2010; Coccurello and Moles, 2010; Davoodi et al., 2009; Fell et al., 2007), although in a few recent experiments, olanzapine exposure led to increased serum triglyceride levels both in female (Skrede et al., 2012b) and in male (McNamara et al., 2011) rats. Discrepant findings in human and rat are likely related to species-specific aspects of energy and drug metabolism, in particular the short half-life (t_1/2) of olanzapine in rats – 2.5–3 h, compared to the average 30 h in humans (Aravagiri et al., 1999; Callaghan et al., 1999; Mattiuz et al., 1997).

Despite shortcomings, the rodent model is essential in the ongoing efforts to improve our understanding of the molecular mechanisms underlying antipsychotic-induced metabolic adverse effects. Data from clinical and preclinical studies have demonstrated stimulation of appetite as the main cause of antipsychotic-induced weight gain, presumably mediated through antagonistic effects on serotonin 5HT2C and/or histamine H1 receptors in the hypothalamus (Reynolds and Kirk, 2010). Other mechanisms of action, such as the up-regulation of fatty acid and cholesterol biosynthesis genes in peripheral tissues, have also been linked to antipsychotic-induced metabolic disturbances (Ferno et al., 2005, 2009; Jassim et al., 2012; Minet-Ringuet et al., 2007; Vik-Mo et al., 2008). Interestingly, alterations in lipid metabolism may also be of relevance at the level of the central nervous system, more specifically in the hypothalamus. Here, modulation of AMP-activated protein kinase (AMPK), a key regulator of fatty acid biosynthesis, has been demonstrated to be relevant for the regulation of appetite (Minokoshi et al., 2004) and whole body metabolism (Lage et al., 2008; Lopez et al., 2008, 2010; Minokoshi et al., 2004). The hypothalamus is organized in anatomically discrete neuronal clusters known as nuclei, with the arcuate nucleus (ARC) considered the ‘master hypothalamic centre’ for feeding control (Morton et al., 2006; Lopez et al., 2007) and the ventromedial nucleus (VMH) demonstrated to be important in the control of energy expenditure (Cannon and Nedergaard, 2004; Lopez et al., 2010; Whittle et al., 2012). Thus, different hypothalamic nuclei may be involved in the various metabolic disturbances induced by antipsychotics. Recent studies in mice demonstrated that olanzapine and clozapine induced activation (phosphorylation) of hypothalamic AMPK, suggested to cause both increased appetite (Kim et al., 2007) and hepatic insulin resistance (Martins et al., 2010). However, none of the studies examined whether these effects were nucleus-specific.

Administration of antipsychotics to rats is commonly achieved by means of s.c. or i.p. injections (Cooper et al., 2005; Fell et al., 2005), or by oral administration through gavage (Arjona et al., 2004), chow (Minet-Ringuet et al., 2006) or drinking water (Cooper et al., 2007; Perez-Costas et al., 2008). As the number of daily administrations through injection or gavage is, for practical reasons, limited, and food/water intake is unevenly distributed throughout the light/dark phases, all these approaches result in marked variation in serum concentrations during a 24 h period (Perez-Costas et al., 2008). In order to circumvent the problem of fluctuating drug serum concentrations, osmotic minipumps have been employed (Choi et al., 2007; Mann et al., 2013; Shobo et al., 2011). However, results are conflicting and minipumps have been characterized as a non-optimal mode of long-term administration of antipsychotics to rats (Remington et al., 2011; van der Zwaal et al., 2008).

In the present study, we injected female rats with a single intramuscular dose of long-acting olanzapine formulation (ZypAdhera^®/Zyprexa Relprevv^®), yielding stable olanzapine plasma levels significantly higher than those achieved through oral administration. Depot injection induced orexigenic effects as well as changes in lipid metabolism that were either comparable to or more pronounced than previously observed after oral administration. Furthermore, we were able to mechanistically confirm that activation of AMPK in ARC is involved in olanzapine-induced weight gain. Administration by means of depot formulation thus represents a promising method in the ongoing efforts to increase the validity of the rat model for antipsychotic-induced adverse effects.

Methods and materials

Animals

All experiments were approved by and carried out in accordance with the guidelines of the Norwegian Committee for Experiments on Animals (Forsøksdyrutvalget, FDU) through standardized applications to the animal facility at Haukeland University Hospital, with ID 20092167, or the University of Santiago de Compostela Institutional Bioethics Committee, the Xunta de Galicia (Local Government) and the Ministry of Science and Innovation, with ID15005AE/10/FUN01/FIS02/MALP1. Rats were kept under standard conditions with an artificial 12/12 h light/dark cycle (lights on: 08:00 hours) and constant 48% humidity. Animals were housed individually and allowed access to tap water and free (ad libitum) access to standard laboratory chow during the experimental period, except for pair-fed groups, in which rats received an amount of food corresponding to that consumed by control rats during the previous 24 h. Tissue samples were harvested as previously described (Skrede et al., 2012b).

Treatment schemes

Female Sprague–Dawley rats received an intramuscular injection of commercially available olanzapine pamoate depot formulation (ZypAdhera^®, Eli Lilly, USA) or commercially available vehicle solution (injection volume 160 µl/250 g BW).

Patients treated with long-acting olanzapine injections receive olanzapine doses approximately 15 times as high as patients treated with oral formulations (Kane et al., 2010). We and others have previously found hyperphagia and weight in female rats treated orally with 5–10 mg/kg olanzapine daily (Goudie et al., 2002; Kalinichev et al., 2005; Skrede et al., 2012b). With an average dose of 7.5 mg/kg.d⁻¹, this yielded a starting dose of 112 mg/kg (7.5 mg/kg × 15) in a pilot experiment, where intramuscular injections were administered to N = 6 rats. Marked hyperphagia and weight gain were observed in the female rats, which was also evident when halving the dose (56 mg/kg; N = 6) (Fig. 1a). At 2 wk, half of the rats receiving each respective dose were sacrificed, whereas the other half received a second injection of olanzapine depot solution (100 mg/kg; N = 6). The subsequent experiments in this study were all carried out within the range of the doses examined in the pilot experiment. In order to avoid diverging food intake in the control and treatment groups due to olanzapine-induced sedation, all rats were fasted during the first 12 h after injection (day 0).

In addition to the pilot experiment, we performed four separate experiments using depot olanzapine injections in female rats. First, we examined the subchronic effect of an olanzapine depot injection (100 mg/kg; N = 15 in the vehicle and olanzapine ad libitum-fed groups, N = 10 in the olanzapine pair-fed group) on metabolic parameters, in order to compare with previously demonstrated effects after oral injection. Rats and chow were weighed daily for 7 d after injection, and rats were sacrificed on day 8. All rats were subject to an overnight fast prior to sacrifice. In the second experiment, investigating changes in olanzapine serum concentrations over time as well as intra-day variation after depot injection, rats (N = 10) were injected with 100 mg/kg olanzapine, and venous blood was sampled from non-fasted rats 6, 12 and 24 h after injection in one cohort of rats, and 78, 84 and 96 h after injection in a second cohort of rats. Plasma olanzapine concentration in all rats was also measured in trunk blood upon sacrifice on day 14. Finally, in two separate mechanistic experiments, rats (N = 10) were stereotaxically injected with adenoviruses carrying either GFP (green fluorescent protein) or dominant negative forms of AMPK (DN-AMPK, for both AMPKα1 and AMPKα2) to suppress AMPK activity in specific hypothalamic nuclei (Lopez et al., 2008, 2010). Rats were placed in a stereotaxic frame (David Kopf Instruments, USA) under ketamine–xylazine anesthesia. Adenoviruses vectors (Viraquest, USA) carrying either GFP (1 × 1012 pfu/ml, control adenoviruses) or AMPKα dominant negative (AMPKα-DN; 1 × 1012 pfu/ml) was delivered at a rate of 200 nl/min for 5 min (1 µl per injection site). The pipette was left in place for 5–10 min and then removed. For the VMH-specific experiment, stereotaxic injections were performed using the coordinates −2.4/−3.2 posterior to bregma and ± 0.6 mm lateral to bregma (depth: 10.2 mm), as previously described (Lopez et al., 2008). For the ARC-specific experiment, stereotaxic injections were performed using the coordinates 2.8 mm posterior and ± 0.3 mm lateral to bregma (depth: 10.2 mm), as previously described (Martins et al., 2012). Directly after stereotaxic delivery into the VMH, rats received an intramuscular injection of olanzapine long-acting injection (100 mg/kg) in accordance with the two foregoing experiments. However, in the ARC-specific experiment, carried out subsequent to the VMH-specific experiment, the olanzapine dose (70 mg/kg) was lowered to reduce sedation and the observed ‘dip’ in food intake. Non-fasted rats were sacrificed on day 9 after injection in both the mechanistic experiments.

RNA extraction, cDNA synthesis and real-time PCR

RNA extraction and cDNA synthesis and RT-PCR were performed as previously described (Skrede et al., 2012b). Relative gene expression levels were determined using the comparative ΔC_t method, using the well-established β-actin (Actb) and ribosomal protein, large, (Rplp0) as endogenous controls.

Lipid and olanzapine measurements

Triglycerides, phospholipids and cholesterol were measured as previously described (Skrede et al., 2012b). Plasma levels of olanzapine were determined by means of liquid chromatography/mass spectrometry (LC-MS), as previously described (Skrede et al., 2012b), or using high performance liquid chromatography (1290 Infinity Binary LC, Agilent Technologies, USA) coupled to an Agilent 6490A triple quadrupole mass spectrometer using positive electrospray ionization (Agilent Technologies, USA) at the Section of Clinical Pharmacology, Laboratory of Clinical Biochemistry, Haukeland University Hospital, Norway.

Western blotting

Western blots were run as previously described (Lopez et al., 2008; Ferno et al., 2011). Primary antibodies used were: ACC, pACCα-Ser⁷⁹, AMPKα1 and AMPKα2 (Upstate, USA), pAMPKα-Thr¹⁷² (Cell signaling Technology, USA), SREBP1 (Santa Cruz Biotechnology, USA), FAS (BD Biosciences, USA) and β-actin (Abcam, UK). Fold change calculations are based on an average of at least N = 6 independent samples.

Statistical analysis

Food intake and weight gain were analysed using two-way ANOVA repeated measures, and Dunnett's post-hoc-test was used to identify significant differences relative to vehicle. For all other comparisons, we used two-sided Student's t-test. p values ⩽0.05 were considered statistically significant. PASW Statistics Version 18 (PASW statistics, IBM, USA) was used for all calculations.

Results

Intramuscular olanzapine injection induces elevation in food intake, weight gain and fat pad mass

The effect of a single intramuscular injection of long-acting olanzapine (100 mg/kg) on food intake and weight gain was measured daily for 7 d in female rats with ad libitum access to food and in chow-restricted (pair-fed) rats, relative to vehicle-treated rats. Pair-fed rats consumed the entire amount of chow offered. A two-way ANOVA analysis, including treatment (three treatment groups) and time (7 d) as factors, revealed significant effect of olanzapine treatment on daily food intake (F(2, 37) = 40.57, p < 0.0001) (Fig. 1b), cumulative food intake (F(2, 37) = 15.88, p < 0.0001) (Fig. 1c) and cumulative weight gain (F(2, 37) = 11.27, p < 0.001) (Fig. 1d). Dunnett's post-hoc tests demonstrated significant increase in daily food intake (p < 0.001), cumulative food intake (p < 0.01) and weight gain (p < 0.01) in olanzapine ad libitum-fed rats relative to control. In pair-fed rats, there were no significant alterations in weight gain (p = 0.2) despite significantly reduced daily food intake (p < 0.01) and cumulative food intake (p < 0.05) relative to vehicle-treated rats, presumably caused by the observed initial sedation (Fig. 1). In ad libitum-fed, olanzapine-treated rats, weight gain was reflected in increased white adipose tissue (WAT) mass, with bilateral retroperitoneal fat pad mass significantly increased (Table 1: vehicle 6.9 ± 0.41 g/kg; olz ad lib 8.2 ± 0.29 g/kg, p < 0.05). In contrast, no significant changes were observed in mesenteric and parametrial WAT mass (Table 1). Of note, olanzapine also induced a small, but significant increase in liver mass in ad libitum fed animals (Table 1), whereas there were no significant changes in pair-fed rats for either WAT or liver mass (Table 1).

The orexigenic effects of the first dose of depot injection declined after about two weeks (Fig. 1a). However, a second dose of depot injection invigorated food intake and weight gain (Fig. 1a), suggesting that the drug tolerance observed during oral treatment is less pronounced after depot injection.

Depot injection yields high, stable plasma olanzapine concentrations

On day 8 after a single dose of long-acting olanzapine formulation (100 mg/kg), the average olanzapine plasma concentration across all rats was 86.4 ± 6.7 nm, with no significant difference between ad libitum-fed (81.1 ± 7.1 nm) and pair-fed (94.3 ± 13.2 nm) rats (Fig. 2a).

In order to measure whether olanzapine depot injection also prevents fluctuations in serum concentrations throughout the day, blood samples were taken at three time points during day 1 (6, 12 and 24 h after injection) or during day 4 (78, 84 and 96 h after injection), in separate cohorts. As expected, the average plasma concentration was significantly higher on day 1 (515.2 ± 41.0 nm) than on day 4 (258.3 ± 36.0, p < 0.001), while variation in plasma concentrations throughout the day was negligible during both days (Fig. 2b). Thus, the slow drug release following depot injection yields high, stable olanzapine plasma concentrations not observed with other widely used methods of drug administration (Perez-Costas et al., 2008). Two weeks after the 100 mg/kg depot injection, the average olanzapine concentration across all rats from these cohorts was 62.6 ± 6.3 nm.

Long-acting depot injection affects triglyceride and cholesterol levels both in plasma and in liver

Exposure to 100 mg/kg long-acting olanzapine yielded significantly elevated plasma triglyceride levels (Fig. 3a) in olanzapine ad libitum fed rats on day 8 after injection (vehicle 0.58 ± 0.02; olanzapine ad libitum 0.85 ± 0.06 mmol/l, p < 0.001). A trend toward increased plasma triglyceride levels was also observed in olanzapine-treated pair-fed rats (vehicle 0.58 ± 0.02 mmol/l; olanzapine pair-fed 0.69 ± 0.05 mmol/l, p = 0.06), despite significant weight loss in this treatment group. For plasma cholesterol (Fig. 3b), we found no significant changes in any of the treatment groups after depot injection, whereas plasma phospholipids (Fig. 3c) were increased in ad libitum fed rats (vehicle 2.17 ± 0.09 mmol/l; olanzapinr ad lib 2.45 ± 0.06 mmol/l, p < 0.05), but not in pair-fed rats. With respect to plasma free fatty acids (Fig. 3d), the effect of olanzapine was different, with significantly reduced levels in pair-fed rats (vehicle 0.25 ± 0.01 mmol/l; olanzapine pair-fed 0.15 ± 0.01 mmol/l, p < 0.0001) but no significant changes in ad libitum-fed rats.

Contrasting the elevation in plasma triglyceride and cholesterol levels, olanzapine depot injection reduced hepatic triglyceride levels (Fig. 3e) both in the olanzapine ad libitum (vehicle 22.6 ± 1.6 µmol/g; olanzapine ad libitum 17.0 ± 1.1 mmol/l, p < 0.01) and in the olanzapine pair-fed treatment groups (vehicle 22.6 ± 1.6 µmol/g; olanzapine pair-fed 15.4 ± 3.1 mmol/l, p < 0.05). Similarly, liver cholesterol levels (Fig. 3f) were reduced in the ad libitum group (vehicle 7.41 ± 0.13 µmol/g; olanzapine ad libitum 6.68 ± 0.10 mmol/l, p < 0.01), with a trend toward reduction in the pair-fed group (vehicle 7.41 ± 0.13 µmol/g; olanzapine pair-fed 6.87 ± 0.28 mmol/l, p = 0.06). Hepatic phospholipids were unaltered by olanzapine treatment (Fig. 3g).

Olanzapine activates SREBP target genes in liver and visceral adipose tissue

Olanzapine depot injection induced lipogenic transcriptional activation both in the liver and in parametrial WAT, including up-regulation (fold change 2–5 relative to vehicle) of acetyl-CoA carboxylase 1 (Acc1), fatty acid synthase (Fasn), stearoyl-CoA desaturase (Scd1) and sterol regulatory element binding transcription factor 1 (Srebf1c) (Fig. 4a,c). Transcriptional changes in WAT were comparable with previous observations made after oral treatment, while depot injection also induced transcriptional activation in the liver, which was not obtained using oral administration. Elevated hepatic protein levels of the lipogenic proteins FAS and mature SREBP1 were observed both after depot injection (Fig. 4b) and after oral administration (Skrede et al., 2012b).

Due to the finding of reduced plasma free fatty acids in the olanzapine pair-fed group, we examined expression levels of two major lipases, hormone-sensitive lipase (Hsl/Lipe) and adipose triglyceride lipase (Atgl) in WAT. Only minor changes were detected, but Hsl transcription was significantly increased in olanzapine pair-fed rats, but not in the ad libitum-fed group (Fig. 4c).

The effect of olanzapine long-acting formulation on hypothalamic pAMPK

Investigating whether long-acting olanzapine injection (100 mg/kg) resulted in activation of hypothalamic AMPK, we rather surprisingly found that despite elevated food intake and weight gain, pAMPK and phosphorylated ACC (pACC) levels were not elevated in protein lysates from whole hypothalamus 8 d after drug exposure (Fig. 5). We also measured related proteins, but with the exception of a minor decrease in AMPKa1 levels, neither of the proteins pAMPK, AMPKa2, pACC or ACC were affected (Fig. 5). Despite the apparent lack of pAMPK and pACC elevation, it should be noted that levels of these proteins were not decreased as observed after oral administration (Ferno et al., 2011).

Olanzapine-induced orexigenic effects are attenuated following ARC-specific DN-AMPK injection

In order to further elucidate the role of hypothalamic AMPK in antipsychotic-induced weight gain, we applied a mechanistic approach. Adenoviruses carrying the dominant negative form of AMPK (DN-AMPKα; carrying both AMPKα1 and AMPKα2 isoforms) or GFP (control) were stereotaxically delivered bilaterally into either the VHM, demonstrated to be significant for metabolic functions other than appetite (Lopez et al., 2008, 2010; Martinez de Morentin et al., 2012; Whittle et al., 2012) or ARC, the most relevant nucleus for appetite control (Lopez et al., 2007), immediately followed by an intramuscular injection of long-acting olanzapine solution or vehicle. In the VMH-specific experiment, we used the same olanzapine dose (100 mg/kg) as in the first experiment, shown to induce the desired phenotype of hyperphagia, weight gain and elevated serum lipids (Figs 1 and 3). Due to initial sedation, as expected with this high dose of olanzapine, there was a clear decrease in daily food intake during the first days following olanzapine depot injection, evident both in GFP- and DN-AMPK rats (Ahnaou et al., 2003) (Fig. 6a). However, this initial decline was followed by a marked, statistically significant olanzapine-induced elevation in daily food intake relative to control, with no significant differences between GFP-injected and DN-AMPK-injected rats. A trend of elevated cumulative food intake (Fig. 6b) and cumulative weight gain (Fig. 6c) was also observed after olanzapine treatment relative to vehicle-treated rats, although cumulative values did not reach statistical significance. Importantly, there were no significant differences between olanzapine-injected GPF and olanzapine-injected DN-AMPK rats for any of the calculated measures, demonstrating that reduced AMPK activity in VMH does not markedly affect olanzapine hyperphagia and weight gain.

In the ARC-specific experiment, a slightly lower dose of olanzapine (70 mg/kg) was used in order to minimize the initial ‘dip’ in food intake. Although this change in protocol may pose a challenge with regard to direct comparison of results from the two mechanistic experiments, the dose is well within the range that was initially demonstrated to yield hyperphagia and weight gain (Fig. 1). Indeed, in GFP-injected olanzapine-treated rats, the initial plunge in food intake was much less pronounced in the ARC experiment as compared to the VMH experiment, and the early decline was followed by significant elevation of daily food intake (Fig. 6d), cumulative food intake (Fig. 6e), as well as a trend toward cumulative weight gain (Fig. 6f). In contrast, in the DN-AMPK olanzapine treated rats, orexigenic effects were markedly reduced, and a statistically significant difference was evident between GFP-injected olanzapine treated rats and DN-AMPK-injected olanzapine treated rats, both with regards to cumulative food intake (Fig. 6e) and cumulative weight gain (Fig. 6f). Of note, the loss of orexigenic effect in olanzapine-treated DN-AMPK rats was most pronounced for cumulative weight gain, whereas reduction in daily and cumulative food intake was more moderate. Furthermore, daily food intake in olanzapine-exposed DN-AMPK-treated rats was significantly higher than in vehicle-exposed GFP rats, indicating that the olanzapine-stimulated increase in food intake was not entirely lost in the DN-AMPK rats.

Of note, when investigating ARC-enriched hypothalamic protein extracts, we found that olanzapine markedly increased the levels of pAMPK (Fig. 7a) and pACC (Fig. 7b) in non-fasted rats. In keeping with the loss of orexigenic effects in the olanzapine-treated DN-AMPK rats, the elevation of pACC was blunted in this group (Fig. 7b), demonstrating that DN-AMPK hampered the ability of pAMPK to phosphorylate its target. The olanzapine-induced elevation of pAMPK itself was not blunted in DN-AMPK injected rats, demonstrating that DN-AMPK blocks downstream effects without affecting AMPK phosphorylation per se. Unphosphorylated ACC levels were not significantly changed by olanzapine exposure in either GFP or DN-AMPK rats (Fig. 7b). Unphosphorylated AMPK (AMPKα1 and AMPKα2) were not measured, since the antibodies used do not distinguish between the DN-AMPK isoform and endogenous AMPK.

Discussion

The rat is commonly used in animal studies modelling metabolic adverse effects of antipsychotic drugs. Features recognized in clinical practice, such as hyperphagia, weight gain, glucose deregulation and, to a certain extent, elevation in serum lipids, have all been reproduced in female, but not male, rats (Boyda et al., 2010). However, the short half-life of antipsychotic drugs in rats represents a major challenge and, unlike in patients, dysmetabolic effects of antipsychotics in the rat are often transient (Aravagiri et al., 1999; Goudie et al., 2002; Skrede et al., 2012a). Continuous administration of moderate olanzapine doses by means of osmotic minipumps has resulted in higher and more stable serum concentrations than oral administration or repeated injections (Choi et al., 2007; Shobo et al., 2011), in some studies associated with hyperphagia and weight gain in female rats (Choi et al., 2007). However, the use of minipumps has been subject to challenges regarding drug solubility and degradation, as well as to complications such as mechanical failure, and this mode of administration has been considered non-optimal (Remington et al., 2011; van der Zwaal et al., 2008). Therefore, to achieve stable plasma concentrations via an alternative mode of administration, we investigated the effect of a single intramuscular injection of long-acting olanzapine solution to female rats. This approach yielded hyperphagia, weight gain and increased plasma triglycerides, as well as plasma olanzapine levels (53–153 nm) comparable to those achieved by use of minipumps (96–122 nm), and within the reference range in patients (16–246 nm) (Robertson and McMullin, 2000). Despite a much higher average olanzapine plasma concentration compared with those obtained through oral dosing (∼0.5 nm), effects on metabolic parameters and weight gain were comparable between the two modes of administration (Perez-Costas et al., 2008; Skrede et al., 2012b).

In ad libitum-fed rats receiving olanzapine depot injections, the effect on hyperphagia, weight gain, plasma lipids and transcriptional changes in WAT did not appear to be more potent than previously observed with oral administration (Ferno et al., 2011; Skrede et al., 2012b). It should be noted that depot injection at the doses used in our experiments induced initial decline in food intake, probably caused by sedation, and that sedation during the treatment period may have blunted hyperphagia. This is in agreement with the fact that hyperphagia and weight gain in the female rat have been demonstrated during exposure to low, oral doses (0.5–2 mg/kg) of olanzapine, while higher doses may in fact fail to induce significant weight gain (Kalinichev et al., 2005; Weston-Green et al., 2010). A dysmetabolic phenotype was still present in our experiments, and together with the fact that weight gain reoccurred after a second depot injection, underscores the advantage of slow drug release in order to maintain drug sensitivity in the rodent model. Future studies should include dose-response experiments in order to identify the lowest efficient dose of olanzapine depot solution in rats with regard to metabolic adverse effects.

As in our previous study, in olanzapine-treated pair-fed rats we found potent, weight-independent lipogenic effects in peripheral metabolic tissues, as well as increased plasma triglycerides (Skrede et al., 2012b). In fact, the most pronounced transcriptional effects on SREBP target genes were observed in pair-fed rats both in the liver and in parametrial WAT, indicating that these effects are caused directly by the drug and are not secondary effects of olanzapine-induced weight gain. Furthermore, olanzapine depot injection significantly reduced cholesterol and triglyceride levels in the liver, both in ad libitum-fed and in pair-fed rats, which is puzzling considering the hepatic transcriptional activation of key genes in lipid biosynthesis. However, in line with earlier findings, recent in vitro studies demonstrated that antipsychotics may in fact inhibit cholesterol biosynthesis at the enzymatic level, and the observed concomitant activation of SREBP-controlled gene expression was explained as a feedback response (Canfran-Duque et al., 2013; Kristiana et al., 2010; Sanchez-Wandelmer et al., 2009; Skrede et al., 2013; Summerly and Yardley, 1965). Our current finding of reduced hepatic cholesterol levels and increased plasma triglyceride levels are in accordance with this hypothesis, but the reduced hepatic triglyceride levels remain unexplained, and further studies are required to elucidate this phenomenon. However, our results confirm the notion that olanzapine induces several weight-independent effects on lipid metabolism in the rat, the molecular basis of which are beginning to be unravelled.

Hypothalamic AMPK has been demonstrated to be activated by antipsychotic drugs in the acute setting, but a relevant phenotype of hyperphagia or weight gain was not reported in these studies (Kim et al., 2007; Martins et al., 2010). In the subchronic setting, displaying a relevant orexigenic phenotype, discrepant findings have been reported regarding hypothalamic AMPK activation (Ferno et al., 2011; Sejima et al., 2011). To properly elucidate the relevance of hypothalamic AMPK in olanzapine-induced hyperphagia in female rats, we employed a well-established knockdown approach (Puntel et al., 2010), focusing on the role of AMPK in discrete hypothalamic nuclei, namely the ARC and the VMH. This nucleus-specific approach is of relevance because recent evidence has showed that modulation of AMPK in the ARC is related to feeding, whereas the VMH regulates energy expenditure (Lopez et al., 2007, 2010; Varela et al., 2012; Whittle et al., 2012). In keeping with this idea, functional inhibition of AMPK by means of DN-AMPK adenoviruses in the ARC reduced the orexigenic effects of olanzapine, an effect not found when the same adenoviruses where injected into the VMH. The effects of DN-AMPK in the hypothalamus was confirmed at the protein level, with Western blots demonstrating reduced levels of pACC, the main target of AMPK, in ARC-enriched protein extracts. Focusing on ARC-specific extracts instead of whole hypothalamus is probably of major importance, and may explain why we did not observe any pAMPK activation in the initial experiment in this study as well as in our previous study on the topic (Ferno et al., 2011).

The experimental set-ups in the present studies were not identical, since the dose in the ARC-specific experiment (70 mg/kg) was lower than in the VMH-specific experiment (100 mg/kg) in order to reduce sedation. Nevertheless, both doses were well within the range of the doses yielding marked hyperphagia and weight gain. It should be noted that in the ARC-specific experiment, the reduction in cumulative weight gain in olanzapine-treated DN-AMPK rats as compared to their GFP counterparts was more pronounced than the reduction observed for daily and cumulative food intake, underlining the possibility that ARC-specific AMPK may play a role in antipsychotic-induced changes in metabolism beyond control of food intake. Nevertheless, this evidence reinforces the notion that AMPK is relevant for olanzapine-induced weight gain and that hypothalamic nuclei-specific effects must be taken into consideration when investigating various aspects of antipsychotic-induced effects on metabolism.

In summary, intramuscular administration of olanzapine depot injection represents a significant step forward in the exploration of metabolic adverse effects of this drug in rats. Depot injection resulted in clinically relevant plasma concentrations, robust hyperphagia and weight gain mediated by modulation of AMPK in the ARC as well as lipogenic stimulation in liver and WAT. This mode of administration markedly reduces animal stress and workload, and the use of depot injections will facilitate complex experimental set-ups, allowing a more targeted approach to the major challenges still facing the field.

Acknowledgments

The research leading to these results has received funding from the Research Council of Norway (including the FUGE grant Nos 151904 and 183327), ExtraStiftelsen Helse og Rehabilitering, Helse Vest RHF, Dr Einar Martens’ Fund, the European Community's Seventh Framework Programme (FP7/2007–2013) under grant agreement no 281854 - the ObERStress project (ML), 245009, Xunta de Galicia (ML:10PXIB208164PR), Fondo Investigationes Sanitarias (ML: PI12/01814). CIBER de Fisiopatología de la Obesidad y Nutrición is an initiative of ISCIII. The authors wish to thank Marianne S. Nævdal and Kjell Ove Fossan for excellent technical assistance.

Conflict of interest

None.

References

Author notes