Introduction

Astrocytes perform numerous functions: they provide structural support for nerve cells, modulate the environment around neurons, regulate the production of synapses, maintain the blood–brain barrier, and release a range of neuronal growth factors [1, 2]. Since glia play a variety of roles in the central nervous system (CNS), a stimulus that affects their morphology and function has widespread consequences including changes in neurotransmission, metabolism of neurotransmitters, synaptic plasticity, and propagation of action potentials [2]. Morphine administration is established as a stimulus that affects glial activity. For example, morphine administration induces region-specific upregulation of glial fibrillary acidic protein (GFAP), a protein constituent found specifically in astrocytes, in specific brain areas [3]. Interestingly, evidence has demonstrated that tolerance and dependence are associated with opioid-induced increases in glial cell activity, resulting in neuroadaptations that may directly contribute to the reinforcing effects of these agonists [4].

Activation of astrocytes induces the synthesis and release of substances capable of modulating the surrounding cells (including neurons) like neurotransmitters, neurotrophic factors, cytokines, chemokines, and extracellular matrix factors [5–7]. Morphine not only increases glial activation but also the consequent production of immune factors (chemokines and cytokines) [8]. In addition, some of these molecules exhibit a profile similar to that of neuromodulators or behave like neurotrophic factors [9, 10]. These data raise the possibility that astrocytes contribute to the synaptic plasticity during the development of drug addiction. Moreover, cytokines, in addition to dopamine (DA) and neurotrophins, have been proposed as key mediators of addiction-related neuronal plasticity [11].

Pleiotrophin (PTN) is a secreted cell signaling cytokine that acts as growth factor and as a neuromodulator with multiple neuronal functions [12]. PTN shares high homology with another peptide, denominated Midkine (MK) [13]. PTN and MK have been involved in neurodegenerative disorders and in response to chronic drug consumption. PTN is upregulated in the cortex and caudate putamen after injection of a cannabinol [14], and in NAc after acute administration of amphetamine [15]. Likewise, increased mRNA and protein levels were found in the prefrontral cortex of alcoholics and smokers [16]. In addition, PTN is also highly upregulated in substantia nigra of patients with Parkinson disease [17] and treatment with l-Dopa increases PTN levels in the striatum [18]. Given that these cytokines exert effects that are similar to those of neurotrophins, these findings support the hypothesis that these two cytokines are upregulated in order to induce neurotrophic or neuroprotective effects during drug consumption [19]. Different membrane receptors could function as a multi-molecular complex coordinated to transduce the PTN-MK signal into the cell by different signaling pathways [12]. So, PTN and MK bind common receptors, including receptor protein tyrosine phosphatase β/ζ (RPTPβ/ζ) [20], which is abundantly expressed in the CNS. The interaction of MK or PTN with RPTPβ/ζ establishes a “ligand-dependent inactivation” of RPTPβ/ζ, presumably a consequence of RPTPβ/ζ dimerization [21]. As a consequence, PTN or MK signaling through RPTPβ/ζ leads to activation of ERK and phosphatidylinositol 3-kinase (PI3K)-Akt [22, 23], important axes inducing morphological changes and modulating addictive behaviors.

To date, despite the fact that morphine causes a full addiction syndrome and persistent restructuring in the mesolimbic reward system, the role of PTN and MK in the context of morphine administration has only been assessed in morphine-induced antinociceptive effects [24, 25]. The action of many addictive substances converges on the mesolimbic dopaminergic reward pathway, inducing increased firing of dopaminergic neurons in the ventral tegmental area (VTA) of the midbrain [26, 27]. Importantly, in preliminary work from our group [28], RT-PCR analysis detected that PTN and MK in the VTA were altered under different paradigms of morphine administration. These preliminary results prompted us to study if these changes were also observed at the protein level. Given the important implications of DA neurotransmission in addiction disorders and the complexity of opiate-induced neuroadaptive responses in the brain reward dopaminergic system, the present study was focused on identifying whether the expression of PTN, MK, RPTPβ/ζ, and their intracellular signaling pathways (Akt and ERK) are altered in association with acute and chronic morphine exposure as well as with morphine withdrawal in the VTA. Then, we assessed the possible activation of astrocytes, which could lead to the release of astrocyte-related soluble factors. Finally, we also aimed to identify those cell subpopulations that produced and secreted PTN and/or MK and those that expressed RPTPβ/ζ in response to morphine administration or morphine withdrawal.

Materials and Methods

Subjects

Male Wistar rats (n = 65; Harlan, Barcelona, Spain) initially weighing 220–240 g were housed (2–3/cage) on arrival in a room with controlled temperature (22 ± 2 °C) and humidity (50 ± 10 %), with free access to water and food (Harlan Teklad standard rodent chow; Harlan Interfauna Ibérica, Barcelona, Spain). Animals were adapted to a standard 12 h light–dark cycle (lights on 08:00 h to 20:00 h) for 7 days before the beginning of the experiments. All surgical and experimental procedures were performed in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC) and were approved by the local Committees for animal research (REGA ES300305440012).

Drug Treatment and Experimental Procedure

Following habituation, rats were implanted subcutaneously (s.c.) with placebo pellets (lactose) for 6 days. Another set of rats were made dependent on morphine by implantation (s.c.) of two 75 mg morphine pellets under light ether anesthesia. This procedure has been shown to produce consistent plasma morphine concentrations beginning a few hours after the implantation of the pellets and a full withdrawal syndrome after acute injection of opiate antagonists [29]. On day 7, rats were injected intraperitoneally (i.p.) with either morphine HCl (20 mg/kg; in a volume of 1 ml/kg body weight), naloxone (1 mg/kg; 1 ml/kg body weight), or an equivalent volume of 0.9 % saline and sacrificed 2 h later. There were five experimental groups: chronic placebo + acute saline, chronic placebo + acute morphine, chronic morphine + acute saline, chronic placebo + acute naloxone, and chronic morphine + acute naloxone. The weight gain of the rats was checked during chronic treatment to ensure that the morphine was liberated correctly from the pellets because it is known that chronic morphine treatment induces a decrease in body weight gain due to lower caloric intake [30]. In addition, the animals were observed for opioid withdrawal behaviors for 30 min before and after naloxone injection.

Electrophoresis and Western Blotting

Animals were killed by rapid decapitation. The brains were removed, placed (with its ventral surface facing up) on a plaque over crushed ice, and tissue samples of the VTA were dissected out as described in [31]. Brain regions were placed in individual wells, frozen immediately on dry ice, and stored at −80 °C until assaying. Samples were placed in homogenization buffer, homogenized, and sonicated for 30 s prior to centrifugation at 10,000×g for 10 min at 4 °C. Samples containing equal quantities of total proteins (20–40 mg, depending on the protein of interest) were separated by 6, 10, or 12 % SDS-PAGE (depending on the molecular weight of the protein of interest) and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Membranes were blocked in TBS containing 0.15 % Tween-20 (TBS-T), 1 % BSA for 90 min at room temperature (RT), and incubated overnight at 4 °C with the primary antibody diluted in 1 % BSA in TBS-T. The following primary antibodies were used: goat polyclonal anti-PTN (1:1000; AF-252-PB, R&D Systems, Minneapolis, MN, USA); rabbit polyclonal anti-MK (1:500; sc-20715, Santa Cruz Biotechnology, Santa Cruz, CA, USA); mouse monoclonal anti-RPTPβ/ζ (1:750; 610180, BD Transduction Laboratories, Erembodegem, Belgium); mouse monoclonal anti-phospho-ERK 1/2 (p-ERK 1/2; 1:1000; sc-7383; Santa Cruz Biotechnology); mouse monoclonal anti-ERK 1/2 (1:1000; sc-135900; Santa Cruz Biotechnology); rabbit monoclonal anti-phospho-Akt (p-Akt; 1:2000; #4060, Cell Signaling Technology Inc., Danvers, MA, USA); rabbit polyclonal anti-Akt (1:1000; #9272, Cell Signaling Technology Inc.). Blots were washed and incubated for 90 min at RT in TBS-T with 1 % BSA with appropriate horseradish peroxidase (HRP) conjugated secondary antibodies: anti-rabbit (1:5000; sc-2004, Santa Cruz Biotechnology) anti-mouse (1:5000; sc-2005, Santa Cruz Biotechnology), or anti-goat (1:5000; sc-2350, Santa Cruz Biotechnology). After washing, immunoreactivity was detected with an enhanced chemiluminescent/chemifluorescent western blot detection system (ECL Plus; GE Healthcare, UK) and visualized by a Typhoon 9410 variable mode Imager (GE Healthcare). We used GAPDH or α-Tubulin as our loading control. Before reprobing, blots were stripped by incubation with stripping buffer (glycine 25 mM and SDS 1 %, pH 2) for 1 h at 37 °C. Blots were subsequently reblocked and probed with rabbit polyclonal anti-GAPDH (1:5000; #2118, Cell Signaling Technology Inc.) or rabbit polyclonal anti-α-Tubulin (1:2500; #2144, Cell Signaling Technology Inc.). For Akt or ERK experiments, the same membrane was processed in the following order: incubation with anti-p-Akt or anti-p-ERK 1/2 antibody, stripping, incubation with anti-Akt or anti-ERK antibody, and stripping and incubation with anti-GAPDH antibody. The ratios of PTN/GAPDH, MK/GAPDH, RPTPβ/ζ/α-Tubulin, p-Akt/t-Akt, t-Akt/GAPDH, p-ERK 1/t-ERK 1, t-ERK 1/GAPDH, p-ERK 1/GAPDH, p-ERK 2/t-ERK 2, t-ERK 2/GAPDH, and p-ERK 2/GAPDH were plotted and analyzed. Protein levels were corrected for individual levels.

Brain Perfusion and Sectioning

Another set of rats were deeply anesthetized with an overdose of pentobarbital (100 mg/kg i.p.) and perfused transcardially with saline followed by fixative containing paraformaldehyde (4 % paraformaldehyde in 0.1 M borate buffer, pH 9.5). After removal of the perfused brains, they were post-fixed in the same fixative for 3 h and stored at 4 °C in PBS containing 30 % sucrose until coronal sections (30 μm thickness) were cut rostrocaudally on a freezing microtome (Leica, Nussloch, Germany). The atlas of Paxinos and Watson (2007) [32] was used to identify the VTA. The sections were cryoprotected and stored at −20 °C until use.

GFAP Immunohistochemistry

Sections of the VTA were used for immunohistochemistry to detect astrocytes. Brain sections were rinsed in PBS and an antigen retrieval procedure was applied by treating sections with citrate buffer (10 mM citric acid in 0.05 % Tween 20, pH 6.0) at 60 °C for 20 min. Endogenous peroxidase activity was inhibited with 0.3 % H2O2 and non-specific binding sites were blocked in 3 % normal goat serum (Sigma, USA)/0.3 % Triton X-100 in PBS for 1 h at RT. The sections were incubated for 72 h (4 °C, constant shaking) with primary antibody/mouse monoclonal anti-GFAP (1:400; sc-33673, Santa Cruz Biotechnology). The sections were then incubated for 2 h at RT in biotinylated secondary antibody/anti-mouse (1:500; BA-2000, Vector Laboratories, Burlingame, CA, USA). Tissue was processed by the avidin-biotin immunoperoxidase method (1:250; Vectastain ABC Elite Kit, Vector Laboratories) and immunopositive cells were visualized by addition of the chromogen 3,3′-diaminobenzidine tetrahydrochloride (DAB; Sigma, USA) and 0.005 % H2O2. Sections were mounted on chrome-alum gelatin-coated slides and dehydrated in graded ethanol series and xylene before being coverslipped.

Quantification GFAP-Positive Cells

Neuroanatomical sites were identified using the Paxinos and Watson (1997) atlas. Photomicrographs were captured by means of Leica microscope (DM 4000B; Leica) connected to a video camera (DFC290; Leica). GFAP-positive cell nuclei were counted using a computer-assisted image analysis system (QWIN; Leica). Positive cells were counted at ×20 magnification. A square field (325 μm) was superimposed upon captured image to use as reference area. The number of astrocytes was counted bilaterally in four to five sections from each animal, and averaged to obtain a single value for each rat. The whole histological quantification was performed blindly. Total counts for different brain regions are expressed as mean ± SEM.

GFAP Densitometric Analysis

The same conventional light microscopy described above was used for optical density (OD) study of the nuclei and processes. Photomicrographs were captured at ×10 magnification (image size, 936 × 702 μm). To avoid observer bias, all sections were quantified by a blinded investigator. The intensity of the different brain areas was evaluated through densitometric analysis using ImageJ 1.43 software (NIH ImageJ, Bethesda, MD, USA). The region of interest was outlined in each image, and a mean optical density value was obtained. Background was determined from unstained areas in the tissue section. The corrected densitometric signal was calculated by subtracting the background from each densitometric determination. Four to five sections from each animal were evaluated. A mean value for different regions of each animal was then obtained. The area outlined in each image was also calculated to assure that there were no differences between the regions of interest analyzed in different groups.

Immunofluorescence

Triple immunofluorescence study was carried out to detect GFAP, PTN, and MK. Sections were treated with citrate buffer (60 °C for 20 min). Non-specific Fc binding sites were blocked with 2 % normal horse serum/0.3 % Triton X-100 in PBS for 1 h at RT, and the sections were incubated for 72 h (4 °C, constant shaking) with primary antibodies/mouse monoclonal anti-GFAP (1:400; sc-33673, Santa Cruz Biotechnology), goat polyclonal anti-PTN (1:400; AF-252-PB, R&D Systems), and rabbit polyclonal anti-MK (1:250; sc-20715, Santa Cruz Biotechnology). Secondary antibodies were applied sequentially for 4 h: Alexa Fluor 488 anti-rabbit IgG (1:1000; A-21206, Invitrogen, Eugene, OR, USA), Alexa Fluor 594 anti-goat IgG (1:1000; A-11058, Invitrogen), and Alexa Fluor 405 anti-mouse IgG (1:1000; A-31553, Invitrogen). After washing, the sections were mounted in ProLong® Gold antifade reagent (Invitrogen).

Double-fluorescent labeling was processed as previously explained. Mesencephalic sections were incubated (72 h, 4 °C) with mouse monoclonal anti-RPTPβ/ζ (1:50; 610180, BD Transduction Laboratories) and goat polyclonal anti-tyrosine hydroxylase (TH; 1:4000; ab101853, Abcam, Cambridge, UK) to identify midbrain dopaminergic neurons. Appropriate secondary antibodies were used: Alexa Fluor 488 anti-goat IgG (1:1000; A-11055, Invitrogen) and Alexa Fluor 594 anti-mouse IgG (1:1000; A-21203, Invitrogen). Sections were incubated in DAPI (1:100,000) for 1 min, and mounted in ProLong® Gold antifade reagent (Invitrogen).

Confocal Analysis

The brain sections were examined using a Leica DMIRE2 confocal microscope and Leica Confocal Software (Leica Microsystems). Images were captured from low magnification to high magnification (×20 to ×63 oil objective). Confocal images were obtained using 405-nm excitation for Alexa Fluor 405 or DAPI, 488-nm excitation for Alexa Fluor 488, and 543-nm excitation for Alexa Fluor 594. Emitted light was detected in the range of 450 nm for DAPI, 515–530 nm for Alexa Fluor 488, and 605 nm for Alexa Fluor 594. Every channel was captured separately to avoid spectral crosstalking. A series of optical sections were performed determining an upper and lower threshold using the Z/Y position for Spatial Image Series setting. The optical series covered 20 μm of thickness through the tissue. Images were deconvolved using Huygens Essential 3.6 by Scientifica Volume Imaging (SVI, Hilversum, The Netherlands).

Materials

Morphine HCl and morphine base were supplied from Alcaliber Laboratories (Madrid, Spain) in cooperation with the Área de Estupefacientes y Psicotropos, Agencia Española del Medicamento y de Productos Sanitarios (Madrid, Spain). Naloxone HCl was purchased from Sigma-Aldrich (Sigma Chemical Co, St Louis, MO, USA). Morphine HCl and naloxone HCl doses are expressed as the weight of the salt. Protease inhibitors were purchased from Boehringer Mannheim (Mannheim, Germany); phosphatase inhibitor Cocktail Set was purchased from Calbiochem (Darmstadt, Germany). Morphine HCl and naloxone were prepared fresh each day by reconstitution in sterile saline (0.9 % NaCl; ERN Laboratories, Barcelona, Spain).

Statistical Analysis

Data are presented as mean ± standard error of the mean (SEM). Data were analyzed using one-way or two-way analysis of variance (ANOVA) followed by a post hoc Newman–Keuls test. Correlations between changes in protein expression were assessed using Pearson correlation. Differences with a p <0.05 were considered significant. Statistical analyses were performed with GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA).

Results

PTN and MK Were Upregulated in the VTA After Morphine Administration and During Precipitated Morphine Withdrawal

PTN and MK have been found to be upregulated in different brain areas after amphetamine, alcohol, or delta-9-tetrahydrocannabinol administration (see “Introduction” section); however, their possible alteration after morphine administration or during morphine withdrawal in the mesolimbic pathway has not been determined. This experiment addressed two questions regarding PTN and MK expression. The first was whether acute or chronic morphine treatment was associated with changes in PTN and/or MK in VTA. ANOVA showed a significant effect after acute morphine for PTN (F(2,20) = 8.656; p = 0.0023) and MK (F(2,19) = 6.949; p = 0.0062) in the VTA. As shown in Fig. 1a and c, post hoc comparisons showed that acute morphine administration significantly elevated PTN (p < 0.01) and MK (p < 0.05) expression in the VTA. This elevation was not seen during chronic morphine administration (p < 0.01) compared with acute morphine injection. Two-way ANOVA for PTN expression showed a significant effect of pretreatment (F(1,22) = 7.23; p = 0.0134). Post hoc test revealed that PTN levels in the VTA were increased after naloxone precipitated morphine withdrawal (p < 0.05), as shown in Fig. 1b. Two-way ANOVA for MK showed a significant effect of acute naloxone injection (F(1,23) = 5.885; p = 0.0235) and an interaction between pretreatment and acute treatment (F(1,23) = 4.516; p = 0.0445). Post hoc test revealed that MK levels in the VTA were increased after naloxone precipitated morphine withdrawal (p < 0.05), as shown in Fig. 1d.

Fig. 1
figure 1

PTN, MK, and RPTPβ/ζ protein expression are altered by acute and chronic morphine administration and during morphine withdrawal in the VTA. Over a 7-day period, control (pla) and morphine (mor)-dependent rats received saline (sal), morphine (mor; 20 mg/kg i.p.), or naloxone (nx; 1 mg/kg s.c.) on day 7 and were sacrificed 2 h later. Semi-quantitative analysis and representative immunoblots of PTN (a, b), MK (c, d), and RPTPβ/ζ (e, f) protein in VTA isolated from rats receiving the above treatments. Each bar corresponds to mean ± SEM. Values are expressed as percentage of controls. * p < 0.05, ** p < 0.01 vs. pla + sal; ++ p < 0.01, +++ p < 0.001 vs. pla + mor; # p < 0.05 vs. pla + nx; &p < 0.05 vs. mor + sal

Full size image

As our first goal was to establish whether acute morphine administration, chronic morphine, or morphine withdrawal affects PTN and/or MK VTN expression, it was important to know if VTA expresses RPTPβ/ζ, one of the main receptors for PTN/MK, as well as whether morphine, morphine dependence, and/or morphine withdrawal alter its expression. So, we next used semi-quantitative Western blot to examine whether morphine and or morphine withdrawal affected RPTPβ/ζ expression protein levels. ANOVA showed a significant effect after acute morphine (F(2,17) = 17.21; p < 0.0001). As shown in Fig. 1e, post hoc comparisons showed that acute morphine administration significantly elevated RPTPβ/ζ (p < 0.01) expression. However, there was a decrease in its expression during morphine dependence compared with the control group receiving placebo plus saline (p < 0.05) and with acute morphine-treated rats (p < 0.001). Two-way ANOVA for RPTPβ/ζ expression showed a significant effect of chronic pretreatment (F(1,24) = 4.29; p = 0.0492) and an interaction between pretreatment and acute treatment (F(1,24) = 5.498; p = 0.0276). Post hoc test revealed that RPTPβ/ζ levels in the VTA were decreased in morphine-dependent rats (p < 0.05), as compared with the control group (Fig. 1f). No changes were observed in morphine-withdrawn rats.

The body weight loss after saline or naloxone injection to placebo-pelleted and morphine-dependent rats was also recorded as a sign of opiate withdrawal (Fig. 2a). In agreement with our previous results [31], two-way ANOVA revealed that morphine pretreatment (F(1,46) = 157.4, p < 0.0001), naloxone injection (F(1,46) = 167.5, p < 0.0001), and the interaction between pretreatment and acute treatment (F(1,46) = 103.1, p < 0.0001) had a significant effect on body weight loss. Post hoc analysis showed that naloxone injection to morphine-dependent animals significantly increased (p < 0.001) body weight loss when compared with the placebo-pelleted group also receiving naloxone and with morphine-treated rats given saline. We next compared the expression of cytokine levels with the severity of the withdrawal syndrome (measured as body weight loss). We did not detect any correlation between MK expression and the withdrawal index (Fig. 2c). In contrast, we observed that PTN expression was significantly correlated with body weight loss after naloxone injection to morphine-dependent rats (Fig. 2b). The data presented here might suggest that PTN is a genetic factor that plays a role in somatic morphine-withdrawal syndrome.

Fig. 2
figure 2

a Effects of naloxone-induced morphine withdrawal on body weight loss. Rats were implanted with morphine or placebo pellets for 7 days. Weight loss was checked immediately before and 2 h after naloxone injection. Data shown are means ± SEM. Post hoc test revealed that there was an increase in body weight loss during morphine withdrawal. ### p < 0.001 vs. pla + nx; &&& p < 0.001 vs. mor + sal. b, c Correlation analysis between PTN or MK and body weight loss. The percent increase in PTN levels was positively correlated with the severity of the withdrawal syndrome

Full size image

Astrocytes Were Activated by Morphine and Morphine Withdrawal in the VTA

Astrocytes undergo a process of proliferation, morphological changes, and enhancement of GFAP expression, termed the activation of astrocytes or astrogliosis [33]. In this work, we assessed the activation of astrocytes in the VTA by both measuring the number of astrocytes (as the number of GFAP-positive cells) in the different groups and studying the levels of GFAP immunoreactivity (GFAP-IR) by densitometric analysis. ANOVA showed a significant effect of morphine administration for GFAP-positive cells (F(2,12) = 5.909; p = 0.0202) and GFAP-IR (F(2,14) = 8.632; p = 0.0048) in the VTA. Post hoc test revealed that acute morphine elevated the number of GFAP-positive cells and GFAP-IR (p < 0.05 and p < 0.01, respectively) (Fig. 3g, i). In addition, chronic morphine administration also increased GFAP-IR (p < 0.05). Two-way ANOVA for GFAP revealed the main effects for chronic pretreatment (F(1,14) = 22.16; p = 0.0003), and significant interaction between acute and chronic treatment (F(1,14) = 10.530; p = 0.0059). We found that precipitated morphine withdrawal significantly increased the number of GFAP-positive cells in the VTA compared with its control group (p < 0.001) and with chronic morphine-treated rats receiving saline (p < 0.01; Fig. 3h). Two-way ANOVA for GFAP-IR in the VTA revealed the main effect for chronic pretreatment (F(1,16) = 17.08; p = 0.0008). Post hoc test showed that both chronic morphine and morphine withdrawal elevated (p < 0.05) GFAP-IR (Fig. 3j).

Fig. 3
figure 3

GFAP expression is enhanced by acute and chronic morphine administration and maintained during morphine withdrawal in the VTA, while astrocyte proliferation only occurs in acute morphine-injected and morphine-withdrawn rats. Over a 7-day period, control (pla) and morphine (mor)-dependent rats received saline (sal), morphine (mor; 20 mg/kg i.p.), or naloxone (nx; 1 mg/kg s.c.) on day 7 and were sacrificed 2 h later. The analyzed region within the VTA is schematically illustrated in (a) (modified from Paxinos and Watson 2007). A rectangle indicates the size of the photomicrographs. b–f Representative photomicrographs showing immunohistochemical detection of GFAP+ nuclei and fibers in midbrain coronal sections (scale bar = 200 μm). b′–f′ High magnifications from each photomicrograph (scale bar = 100 μm). g, h Quantitative analysis of astrocytes in the VTA. i, j Mean optical density measurement of GFAP immunoreactivity in the VTA from rats receiving the treatments mentioned above. k, l Reference area used in the densitometric analysis did not differ between groups. SN substantia nigra. Each bar corresponds to mean ± SEM. * p < 0.05, ** p < 0.01 vs. pla + sal; # p < 0.05, ### p < 0.001 vs. pla + nx; &&p < 0.01 vs. mor + sal

Full size image

PTN but not MK was Overexpressed in Astrocytes During Acute Morphine Administration and Morphine Withdrawal in the VTA

Activation of astrocytes produces the accumulation of intermediate-filament GFAP; consequently, they can be recognized by their stellate morphology. Triple immunofluorescence study revealed that acute morphine (Fig. 4b–b′′′′) or morphine withdrawal (Fig. 4c–c′′′′) mediated the activation of astrocytes that expressed high levels of PTN protein, but not MK protein in the VTA. Cells that produced MK are supposed to be neurons due to their size and morphology and the fact that MK and RPTPβ/ζ colocalized (data not shown).

Fig. 4
figure 4

PTN but not MK is overexpressed in astrocytes during acute morphine administration and morphine withdrawal in the VTA. a–c Stack of confocal images from the midbrain areas immunostained for GFAP (blue), PTN (red), and MK (green) in control rats, rats treated with acute morphine, or morphine-dependent animals injected with naloxone. Scale bar = a–c, 50 μm

Full size image

RPTPβ/ζ was Expressed in Midbrain Dopaminergic Neurons

RPTPβ/ζ is a transmembrane tyrosine phosphatase, the extracellular domain of which carries chondroitin sulfate chains. The anti-RPTPβ/ζ antibody used recognizes only the intracellular portion of the RPTPβ/ζ to avoid non-specific labeling of other proteoglycans and especially of phosphacan, a secreted peptide which has the same molecular structure as the extracellular portion of the RPTPβ/ζ.

In order to determine where this receptor is located, we immunolabeled RPTPβ/ζ in combination with specific cell markers on the mesencephalon of morphine-treated rats. At low magnification, we observed that following morphine injection, there was a region-specific increase of RPTPβ/ζ protein in the VTA, but weak or absent staining was detected in the interpeduncular nucleus (IP) or red nucleus (RN) (Fig. 5a–a′′′). At high magnification, RPTPβ/ζ immunoreactivity was widely distributed throughout the VTA, and double-labeling with anti-TH (the rate-limiting enzyme of DA synthesis) showed that RPTPβ/ζ is located on cell bodies and proximal processes of almost all dopaminergic neurons in the VTA (Fig. 5c–c′′′). There were some RPTPβ/ζ+/TH− cells that might represent GABAergic interneurons in the VTA.

Fig. 5
figure 5

RPTPβ/ζ is expressed in midbrain dopaminergic neurons. a RPTPβ/ζ (red) protein in the VTA is region-specific since strong staining in the VTA contrasts with weak or absent staining in the interpeduncular nucleus (IP) or red nucleus (RN). b, c Stack of confocal images from the midbrain areas in control rats and animals treated with acute morphine. Double-labeling with anti-TH (green) showed that, after a morphine challenge, RPTPβ/ζ is located on cell bodies and proximal processes of almost all dopaminergic neurons in the VTA and some non-dopaminergic cells. DAPI (blue) was used as a counterstaining in both nuclei. Scale bar = a, 200 μm; b–c, 50 μm

Full size image

ERK Pathway but not Akt was Activated by Morphine and by Precipitated Morphine Withdrawal in the VTA

In each experiment, the specific signal of p-Akt or p-ERK proteins was normalized to the corresponding Akt or ERK signals, respectively, and then to the level of GAPDH measured in the same preparation. Previously, it has been described that PTN or MK signaling through RPTPβ/ζ leads to activation of ERK and Akt pathways [22, 23]. ANOVA for p-Akt showed a significant effect after acute or chronic morphine administration (F(2,18) = 5.265; p = 0.0175) failed to detect significant changes in t-Akt expression in the VTA (F(2,18) = 2.845; p = ns) (Fig. 6a, c). Post hoc test revealed that p-Akt levels in the VTA were decreased in morphine-dependent rats (p < 0.05). Two-way ANOVA revealed that morphine withdrawal had a significant effect on p-Akt (F(1,21) = 5.190; p = 0.0332). Post hoc test revealed that p-Akt levels in the VTA were significantly (p < 0.05) elevated in morphine-withdrawn rats compared with the morphine-dependent group receiving saline instead naloxone (Fig. 6b). Two-way ANOVA revealed that morphine pretreatment (F(1,21) = 1.83; p = 0.0539), acute naloxone injection (F(1,21) = 0.04; p = 0.8352), or the interaction between pretreatment and acute treatment (F(1,21) = 0.13; p = 0.7172) had no significant effects on t-Akt (Fig. 6d).

Fig. 6
figure 6

p-Akt levels but not t-Akt levels are diminished during chronic morphine administration in the VTA. Over a 7-day period, control (pla) and morphine (mor)-dependent rats received saline (sal), morphine (mor; 20 mg/kg i.p.), or naloxone (nx; 1 mg/kg s.c.) on day 7 and were sacrificed 2 h later. Semi-quantitative analysis and representative immunoblots of p-Akt/t-Akt ratio (a, b) and t-Akt levels (c, d) in VTA isolated from rats receiving the above treatments. Each bar corresponds to mean ± SEM. Values are expressed as percentage of controls. * p < 0.05 vs. pla + sal; + p < 0.05 vs. pla + mor; &p < 0.05 vs. mor + sal

Full size image

Because t-ERK showed changes after the different treatments, p-ERKs 1/2 were quantified by Western blot using both t-ERK and GAPDH as loading controls (Fig. 7). ANOVA showed a significant effect after acute or chronic morphine for p-ERK 1/t-ERK 1 (F(2,19) = 11.65; p = 0.0007) and for p-ERK 2/t-ERK 2 (F(2,19) = 15.26; p = 0.0002). As shown in Fig. 7a and g, post hoc comparisons showed that both acute and chronic morphine administration significantly elevated p-ERK 1/t-ERK 1 (p = 0.001; p = 0.01) and p-ERK 2/t-ERK 2 (p = 0.001; p = 0.001) expression in the VTA. Two-way ANOVA for p-ERK 1/t-ERK 1 and p-ERK 2/t-ERK 2 expression showed a significant effect for morphine pretreatment (p-ERK 1—F(1,24) = 21.15, p = 0.0001; p-ERK 2—F(1,24) = 41, p < 0.0001) and acute naloxone injection (p-ERK 1—F(1,24) = 4.62, p = 0.0419; p-ERK 2—F(1,24) = 8.85, p = 0.0066). Post hoc comparisons showed that morphine withdrawal significantly (p < 0.001) elevated p-ERK 1 and p-ERK 2 levels compared with their respective controls and with morphine-dependent rats (p < 0.05) (Fig. 7b, h). In addition, p-ERK 2 levels were also increased (p < 0.01) in morphine-dependent rats compared with its control (Fig. 7h). Similar results were seen by using the p-ERK 1/GAPDH and p-ERK 2/GAPDH. Thus, ANOVA showed a significant effect after acute morphine for p-ERK 1/GAPDH (F(2,19) = 17.87; p < 0.0001) and after acute and chronic morphine for p-ERK 2/GAPDH (F(2,19) = 14.91; p = 0.0002). As shown in Fig. 7e and k, post hoc comparisons showed that acute morphine administration injection significantly elevated p-ERK 1/GAPDH and p-ERK 2/GAPDH (p < 0.001) as well as and p-ERK 2/GAPDH after chronic morphine treatment (p < 0.01) expression in the VTA (Fig. 7k). Two-way ANOVA showed a significant effect of chronic pretreatment for p-ERK 1/GAPDH (F(1,24) = 12.41; p = 0.0017) and p-ERK 2/GAPDH (F(1,24) = 21.31; p = 0.0001) and acute naloxone injection (p-ERK 1/GAPDH—F(1,24) = 6.14, p = 0.0207; p-ERK 2/GAPDH—F(1,24) = 6.81, p = 0.0154). As shown in Fig. 7, there was an increase (p < 0.01) of p-ERK 1 and p-ERK 2 after naloxone administration to morphine-dependent rats. p-ERK 2 was also elevated during morphine dependence (Fig. 7l).

Fig. 7
figure 7

p-ERK/t-ERK ratio, t-ERK levels, and p-ERK absolute levels are altered by acute and chronic morphine administration and during morphine withdrawal in the VTA. Over a 7-day period, control (pla) and morphine (mor)-dependent rats received saline (sal), morphine (mor; 20 mg/kg i.p.), or naloxone (nx; 1 mg/kg s.c.) on day 7 and were sacrificed 2 h later. Semi-quantitative analysis and representative immunoblots of p-ERK 1/t-ERK 1 ratio (a, b), t-ERK 1 levels (c, d), p-ERK 1 absolute levels (e, f), p-ERK 2/t-ERK 2 ratio (g, h), t-ERK 2 levels (i, j), and p-ERK 2 absolute levels (k, l) in VTA isolated from rats receiving the above treatments. Each bar corresponds to mean ± SEM. Values are expressed as percentage of controls. * p < 0.05, ** p < 0.01, *** p < 0.001 vs. pla + sal; + p < 0.05, ++ p < 0.01 vs. pla + mor; ## p < 0.01, ### p < 0.001 vs. pla + nx; &p < 0.05 vs. mor + sal

Full size image

We next used semi-quantitative Western blot to examine the effect of acute morphine administration, chronic morphine, and morphine withdrawal on t-ERK 1 (ratio t-ERK 1/GAPDH) and t-ERK 2 (ratio t-ERK 2/GAPDH) protein levels (Fig. 7c–j). ANOVA showed significant effects after acute morphine for t-ERK 1 (F(2,19) = 6.160; p = 0.0097) and for t-ERK 2 (F(2,19) = 4.459; p = 0.0277). As shown in Fig. 7c and i, post hoc comparisons showed an increase of t-ERK 1 (p < 0.01) and t-ERK 2 (p < 0.05) after acute morphine administration. This effect was suppressed in chronic morphine-treated rats. Two-way ANOVA for VTA t-ERK1 and t-ERK2 levels failed to detect any significant effects of chronic morphine pretreatment or morphine withdrawal (Fig. 7d, j).

Relationship Between PTN and RPTPβ/ζ Expression and Between RPTPβ/ζ and ERK Phosphorylated Levels in the Different Experimental Groups

We next compared the expression of PTN and MK with the induction of RPTPβ/ζ protein levels by Pearson correlation. There were no significant correlations in the different experimental groups between MK expression and RPTPβ/ζ protein levels in the VTA (data not shown). In contrast, we observed that during morphine dependence and withdrawal, the expression of PTN was significantly positively correlated with RPTPβ/ζ levels (Fig. 8a–c).

Fig. 8
figure 8

Correlation analysis between PTN and RPTPβ/ζ expression and between RPTPβ/ζ and ERK phosphorylated levels in the different experimental groups. a–c The percent increase in PTN levels was positively correlated with RPTPβ/ζ protein during morphine dependence and morphine withdrawal. d–g During morphine dependence and withdrawal, RPTPβ/ζ expression tended to be positively correlated with p-ERK 1/2 levels. h Schematic illustration showing the possible mechanism through which PTN/RPTPβ/ζ signaling may influence ERK phosphorylation during morphine dependence and morphine withdrawal. # p < 0.05

Full size image

Then we compared the expression of RPTPβ/ζ with the induction of p-ERK protein levels in control rats that were injected an acute dose of morphine. However, RPTPβ/ζ expression in acute morphine-injected rats was not correlated with changes on p-ERK or t-ERK levels. On the other hand, during morphine dependence and withdrawal, expression of RPTPβ/ζ tended to be positively correlated with p-ERK. There was a significant positive correlation between RPTPβ/ζ and both p-ERK 1 and p-ERK 2 protein levels during chronic morphine administration (Fig. 8d, f). As depicted in Fig. 8e and g, a similar trend for positive correlation between RPTPβ/ζ and p-ERK 1/p-ERK 2 was found after naloxone-induced morphine withdrawal.

Discussion

The data presented in this study show that different paradigms of morphine treatment promote a diverse profile of activation of cytokines in the brain. While we observed that a single dose of morphine and morphine withdrawal increased the protein levels of PTN and MK in the VTA, chronic morphine administration had the opposite effect since protein levels returned to basal levels or even decreased. Our results are supported by previous findings showing that acute morphine or withdrawal but not chronic morphine promotes the expression of pro-inflammatory cytokines [34]. The present study also showed that PTN up-regulation following morphine administration or during morphine withdrawal is restricted to astrocytes. In agreement with this data, PTN mRNA levels were up-regulated in reactive astrocytes in a model of cryo-injured mouse brain [35]. Besides, in primary cultures of mesencephalon, GFAP-positive astrocytes express PTN mRNA and protein [36], and recently, it has been described that PTN is overexpressed only in astrocytes following an adenoviral vector injection [37]. In contrast, we observed that MK is produced and secreted by non-astrocytic cells following a morphine challenge. Hippocampal pyramidal neurons in adult mouse brain expressed MK, but kainic acid injection induced cell death of pyramidal neurons and enhanced expression of MK by astrocytes [38]. A possible explanation is that, depending on the nature of the insult or the damage/cell death it may produce, MK can be overexpressed by neurons or astrocytes. Moreover, there is evidence that nearly every cell of the CNS, including neurons, is capable of contributing and modifying the central immune signaling [39], for example, by secreting cytokines. For instance, morphine stimulates chemokine CCL2 production by neurons [40]. Cytokines have also been implicated in withdrawal-related behavior. For example, it was showed that corticotrophin releasing factor (CRF) and cytokines work together to worsen ethanol withdrawal phenotypes [41]. Our results indicate that the increase in PTN expression is related with a less severe withdrawal syndrome, characterized by reduced body weight loss during morphine withdrawal, suggesting that the modulation of PTN signaling pathway might counterbalance the somatic signs of withdrawal. In concordance, it has been described that an augmented withdrawal syndrome score in PTN−/− mice, demonstrating that the lack of endogenous PTN enhances the severity of the physical opiate withdrawal [42]. Although the role of PTN and MK in conditioned place aversion (CPA) remains uncharacterized, future studies should address this issue.

We also found that the expression of PTN− and MK− target receptor (RPTPβ/ζ) was regulated in the same way that these cytokines were regulated by morphine administration. Although the role and the functions of RPTPs are barely characterized, it has been suggested that each of these enzymes has different and specific functions in neurons [43]. Regarding the RPTPβ/ζ signal, we found a staining pattern in neurons. Our results are in line with previous studies where this protein was found to be located in neurons but not in astrocytes [43, 44]. Moreover, TH-positive neurons in primary cultures of mesencephalon have been described to express RPTPβ/ζ [36]. Although it has been reported that adult control rats do not express RPTPβ/ζ in the VTA [45], we detected (by WB and IF) that systemic morphine injection induced RPTPβ/ζ in VTA-dopaminergic neurons in a region-specific manner. Thus, it suggests that RPTPβ/ζ is tightly up-regulated in the VTA after determinate challenges. This expression pattern supports our hypothesis of an interaction between glial and neuronal function during morphine administration and withdrawal.

Astrocytes can display both hypertrophy and proliferation upon treatment with drugs of abuse. Chronic morphine exposure, achieved by surgically implanted morphine pellets and previously verified to produce opioid tolerance and dependence [46], increased GFAP immunoreactivity in the VTA [3]. This effect was not observed when morphine tolerance and dependence was blocked by concomitant administration of the mu-opioid antagonist, naltrexone [3]. Accordingly, in the present study, acute and chronic morphine caused a robust activation of astrocytes, as evidenced by an increase in the level of GFAP-IR in the VTA, while morphine withdrawal maintained the increase in GFAP-IR levels. Regarding proliferation, repeated methamphetamine administration induces proliferation of astrocytes in rodent striatum and SN [47, 48]. Moreover, morphine and opioid signaling have been shown to promote proliferation of astroglia in the postnatal brain [49]. We observed a rapid astrocyte proliferation in the VTA, the brain area where the rewarding properties of morphine are believed to be firstly mediated (via binding to mu-opioid receptors located on inhibitory neurons) [50]. A question that arises is whether glial activation (measured as GFAP-IR) observed during chronic morphine is maintained throughout a longer time or, on the contrary, it slowly decreases. Data supporting this second hypothesis is the evidence that, after 2 months, methamphetamine induced a long-lasting astrocytic activation and behavioral sensitization, while morphine produced a reversible activation of astrocytes and a reversibility of behavioral sensitization [51].

PTN and MK have been involved in the regulation of the survival and function of dopaminergic neurons [52, 53]. In addition, many reports attribute neuroprotective roles against drug-induced neurotoxicity to these two cytokines. For instance, amphetamine-induced loss of dopaminergic fibers in the striatum was enhanced in PTN knockout (PTN−/−) mice and even caused dopaminergic cell loss in the substantia nigra (SN) of PTN−/− mice [47]. It is important to note that moderate glial activation is thought to be neuroprotective, while exacerbated activation can be deleterious for the brain [54–56]. Accordingly, amphetamine-induced astrocytosis in the nigrostriatal pathway was increased in PTN−/− mice and in MK−/− mice compared to wild-type mice [47, 48, 57]. As exposed above, morphine withdrawal may promote cytokines and other inflammatory responses that have the potential of exacerbating neuronal damage [34]. The increase in pro-inflammatory cytokines may explain why withdrawal causes caspase-3-dependent apoptosis [58]. So, the enhanced expression of PTN and MK that we observed during acute morphine injection and morphine withdrawal can be regarded as a physiological response to prevent the possible neurotoxic effects of morphine, an excessive astrogliosis or pro-inflammatory cytokine release. Moreover, it has been demonstrated that PTN increases the expression of nuclear receptor related-1 (Nurr1, a transcription factor which is critical for the generation, survival, and maintenance of DA neurons) in neural stem cells to promote production of dopaminergic neurons [52]. Previously, we have observed an increase of Nurr1 after acute morphine administration and withdrawal [31], which parallels that of PTN. So, given the present results, it is tempting to speculate that neuroprotective effects of PTN during acute morphine injection and morphine withdrawal may be, in part, mediated through the induction of Nurr1.

PTN/MK signaling may function through a multi-receptor complex, combining different receptors and most probably other adaptor proteins; then, PTN/MK action over previously mentioned receptors could in turn signal through different signal pathways [12]. It has been documented that the interaction of RPTPβ/ζ with PTN blocks the intrinsic tyrosine phosphatase activity of RPTPβ/ζ [19], leading to an increase in the phosphorylation levels in RPTPβ/ζ target proteins. Interestingly, these target proteins include the PI3K-Akt and Ras-ERK signaling, which have been demonstrated to play a crucial role in opiate addiction (for review see [59]). We detected a direct relationship between PTN and RPTPβ/ζ in morphine-dependent and morphine-withdrawn animals. We propose that PTN expression after chronic morphine and morphine withdrawal would lead to adaptative responses, such as the increase in RPTPβ/ζ levels, which might try to compensate the ligand-dependent inactivation. As exposed above, PTN neuroprotective effects constitute a relevant role, suggesting that PTN signaling pathways are involved in neurodegenerative disorders as well as in response to injuries and chronic drug consumption [12]. On the other hand, several studies have implicated the activation of ERK pathway in drug dependence and naloxone-precipitated withdrawal in different brain areas [60, 61], including the VTA [62, 63]. Interestingly, PTN protective effects against amphetamine-induced toxicity in PC12 cells (which express RPTPβ/ζ and synthesize and release DA) were mediated by the ERK 1/2 signaling pathway [48]. Thus, given the notion that the protection against amphetamine-induced toxicity in vivo and in vitro is mediated by the increased phosphorylation of ERK [47, 48], and our observation of a positive correlation between PTN/RPTPβ/ζ and p-ERK 1/2, we propose that this signaling pathway may be also mediating neuroprotective effects during morphine chronic administration and withdrawal.

On the other hand, a decrease of Akt signaling was detected in the VTA during chronic morphine. This down-regulation is considered critical to mediate the effects of chronic morphine on soma size and electrical excitability [64, 65]. PTN and MK are also known to be neurotrophic factors for DA neurons [36]. Initial up-regulation of PTN promoted regeneration of axons, but subsequent decline in the expression of PTN has been proposed to be responsible for failure to regenerate denervated nerves [66], highlighting the importance of sustained levels of these growth factors to accomplish their trophic function. Nevertheless, we did not find any correlation between PTN, MK, or RPTPβ/ζ and p-Akt levels, so it is reasonable to think that these cytokines do not participate in the decreased dendrite complexity and spine density and cell body size that chronic opiates produce in the mesolimbic system [59].

In summary, given that PTN, MK, and RPTPβ/ζ levels increase after acute morphine injection, return to basal levels during chronic opioid treatment, and are up-regulated again during morphine withdrawal, we hypothesize that signaling through these cytokines plays a role in mediating, at least in part, neuroprotective and behavioral adaptations that are observed during opiate addiction.