|Home | About | Journals | Submit | Contact Us | Français|
This study examined the mechanism by which exposure to lipopolysaccharide (LPS) alters mu-opioid receptor (MOR) expression in immune and neuronal cells using an in vitro conditioned medium model system. We found that LPS stimulated the intracellular accumulation of reactive oxygen species (ROS) and MOR expression in macrophage-like TPA-HL-60 cells. Conditioned medium from the LPS-stimulated TPA-HL-60 cells increased MOR expression in SH-SY5Y cells, a neuronal cell model, through actions mediated by TNF-α and GM-CSF. These data suggest that the endotoxin, LPS, modulates MOR expression in nervous and immune cells via ROS signaling, and demonstrates the crosstalk that exists within the neuroimmune axis.
Lipopolysaccharide (LPS), an outer-membrane component of gram-negative bacteria, is a well characterized endotoxin that activates the immune system, and, in particular, induces inflammation (Dunn, 1991; Lopez-Bojorquez et al., 2004; Parrillo et al., 1990; Raetz and Whitfield, 2002; Rumpa, 2010). In some cases, endotoxemia progresses to severe sepsis, resulting in multiple organ dysfunction, septic shock, and death (Lopez-Bojorquez et al., 2004; Qu et al., 2009). Morbidity associated with severe sepsis is high. There are one million deaths from sepsis worldwide, and approximately 25-30% of the cases are due to gram-negative bacterial infection (Rumpa, 2010).
The host-mediated response to endotoxemia involves the secretion of inflammatory cytokines and mediators as well as the activation of the coagulation and complement cascades (Andreasen et al., 2008; Dunn, 1991; Lopez-Bojorquez et al., 2004). The increased levels of circulating inflammatory cytokines resulting from LPS endotoxemia exacerbate systemic inflammation. Our previous studies showed that the levels of the pro-inflammatory cytokines, TNF-α, IL-1β, and IL-6, are elevated in both the serum and brain of rats treated systemically with LPS (Chen et al., 2005; Ocasio et al., 2004). Other studies have reported an increase in the secretion of IL-1β and TNF-α from macrophages following LPS treatment (Evans et al., 1991; Hsu and Wen, 2002). Our previous findings also indicated that LPS couples the immune and nervous systems via actions mediated by pro-inflammatory cytokines on the hypothalamic-pituitary-adrenal (HPA) axis and that such cross-talk is necessary in order to maintain homeostasis in response to infection (Chang et al., 1998).
Inflammatory cytokines can modulate the expression of the mu-opioid receptor (MOR) in both neuronal and immune cells. In 1998, we reported that co-treatment with IL-1α and IL-1β increases MOR expression in microvascular endothelial cells (Vidal et al., 1998). IL-6 increases MOR expression and MOR binding in SH-SY5Y neuroblastoma cells (Borner et al., 2004), and TNF-α increases MOR expression in human T lymphocytes, Raji B cells, U937 monocytes, primary human polymorphonuclear leukocytes, and mature dendritic cells (Kraus et al., 2003).
The activation of the opioidergic pathway via the MOR leads to suppression of the immune response (Gaveriaux-Ruff et al., 1998; Wang et al., 2002). Chronic administration of morphine, a MOR agonist, desensitizes the pro-inflammatory cytokine-mediated effects on the HPA axis and deregulates the immune response in rats (Chang et al., 2001; Chang et al., 1995; Chang et al., 1996; Chen et al., 2005). In addition, deregulation of immune responses by exogenous opioids leads to many of the complications associated with LPS-induced endotoxic shock (Chang et al., 2001; Chang et al., 1998; Chen et al., 2005).
Reactive oxygen species (ROS) are highly reactive molecules produced during cellular respiration (Bast and Goris, 1989; Bayir, 2005; McCord and Fridovich, 1978). Both intracellular and extracellular ROS are maintained at non-lethal levels by superoxide dismutases, catalases, and a thiol-reducing buffer consisting of glutathione and thioredoxion (Gamaley and Klyubin, 1999; Nakamura et al., 1997). However, disease and stress can alter a cell’s ability to effectively regulate ROS. Elevated levels of ROS can damage proteins, DNA, RNA, and cell membranes by hydroxyl radical attack, and induce apoptosis (Bast and Goris, 1989; Machlin and Bendich, 1987; McCord and Fridovich, 1978). Exposure to LPS increases the production of ROS in murine macrophages (Hsu and Wen, 2002; Kim et al., 2004), and the accumulation of ROS is a promoting factor in the development of sepsis in rats (Bayir, 2005). Thus, ROS appears to play a key role in the LPS-induced inflammatory response and the subsequent incidence of sepsis.
In this study, an in vitro conditioned medium (CM) model was used to investigate the mechanism by which LPS exposure alters MOR expression in immune and neuronal cells. Specifically, we examined the effects LPS-induced ROS accumulation on MOR expression in TPA-differentiated HL-60 (TPA-HL-60) macrophage-like cells (Kowalski and Denhardt, 1989; Rovera et al., 1979). We also assayed the conditioned medium from the LPS-treated TPA-HL-60 cells for TNF-α, GM-CSF, IL-1β, IL-8, IL-10, IL-12p70, IL-2, IL-6, and INFγ to determine if LPS-induced ROS had a modulating effect on the cytokine secretion profile. We then evaluated MOR expression in SH-SY5Y neuroblastoma cells (Ciccarone et al., 1989) cultured in CM from the LPS-treated TPA-HL-60 cells. We have previously used undifferentiated SH-SY5Y cells to study MOR expression (Yu et al., 2003; Zadina et al., 1993; Zadina et al., 1994), and others have reported the use of SH-SY5Y as a neuronal cell model since it presents several neuronal markers, including tyrosine and dopamine-β-hydroxylase activity, uptake of norepinephrine, and core Mr 68,000 and Mr 150,000 neurofilament proteins (Ciccarone et al., 1989; Gao et al., 2001; Ruffels et al., 2004; Wu et al., 2007). Our findings indicate that ROS plays a key role in LPS-induced modulation of MOR expression in both neuronal and immune cells.
Human HL-60 promyelocytic leukemic cells (ATCC, Manassas, VA) were grown in RPMI 1640 medium supplemented with 20% FBS, 100 U penicillin, and 100 μg/mL streptomycin (Gibco, Invitrogen Corp., Grand Island, NY). Experimental 12-well plates (BD Biosciences, VWR, West Chester, PA) were seeded with HL-60 cells at 5×105 cells/mL in 1 mL/well. In this study, HL-60 cells were differentiated with 12-O-tetradecanoylphorbol-13-acetate (TPA) into macrophage-like cells (TPA-HL-60) over a period of 48 h (Sigma-Aldrich, St. Louis, MO). Stock solutions of TPA were dissolved in 100% ethanol to a concentration of 16 μM and diluted 1000-fold in medium. Previous studies have shown greater than 95% of HL-60 cells differentiate to macrophage-like cells after a 48-h treatment with 16 nM TPA (Kowalski and Denhardt, 1989; Rovera et al., 1979). After TPA differentiation of HL-60 cells, the medium containing TPA was removed and the TPA-HL-60 cells were rinsed 3 times with 1 mL of phosphate-buffered saline solution (PBS)[Invitrogen, Grand Island, NY]. Human SH-SY5Y neuroblastoma cells, a gift from R. Ross (Fordham University, New York, NY), were grown in MEM + F12 medium supplemented with 10% FBS, 100 U penicillin, and 100 μg/mL streptomycin (Gibco, Invitrogen Corp., Grand Island, NY). SH-SY5Y cells were seeded at 1×105 cells/mL in 1 mL/well. TPA-HL-60 and SH-SY5Y cells were maintained in a 5% humidified incubator at 37° C.
A stock solution of lipopolysaccharide from E. coli 055:B5 (LPS, Sigma-Aldrich, St. Louis, MO) was prepared in 0.9% saline. TPA-HL-60 cells were treated with medium containing LPS at a final concentration of 0 to 0.500 mg/mL. For ROS scavenging experiments, TPA-HL-60 cells were pre-treated with medium containing 100 μM vitamin E (Sigma-Aldrich, St. Louis, MO) for 3 h, and then treated with medium containing LPS. Unless noted otherwise, in CM experiments, TPA-HL-60 supernatants were pooled, filter sterilized (0.2 μm filters, Pall Life Sciences, Ann Arbor, MI), and immediately overlaid onto SH-SY5Y cells. SH-SY5Y cells used in CM experiments were grown for a period of 72 h prior to TPA-HL-60-CM treatment.
Intracellular ROS levels in TPA-HL-60 cells cultivated in 12-well plates were determined by confocal laser scanning microscopy (CLSM). After LPS treatment, the medium was replaced with fresh medium containing 20 μM dihydrorhodamine123 (DHR123) [Sigma-Aldrich, St. Louis, MO], and incubated for 30 min (Henderson and Chappell, 1993). DHR123 is an indicator of hydrogen peroxide, hypochlorous acid, and peroxynitrite anion (Crow, 1997; McBride et al., 1999; Radi et al., 2001). All treatment groups examined in CLSM experiments were loaded with DHR123 in medium. After incubation with DHR123, the medium containing DHR123 was removed and replaced with fresh medium alone. The loading and removal of all treatment groups with DHR123 prior to CLSM controls for any potential differences in ROS levels due to loading or leaking of DHR123. The medium was then replaced with fresh medium alone, and ROS levels were measured with a FluoView FV1000 CLSM (Olympus, Center Valley, PA) at 200× magnification. Laser transmissivity was set to 20%; the cells were excited at 488 nm, and fluorescence emission was detected at 520 nm. Changes in intracellular ROS levels were calculated as the percent control from the mean fluorescence intensity of 10 randomly selected cell clusters from 3 views from 3 wells per treatment as follows:
where ROI refers to the region of interest and avg. CHS1 refers to the average fluorescence of the pixilated ROI.
Total RNA was extracted and isolated using TRIzol or TRIreagent in accordance with the manufacturer’s protocol (Ambion, Invitrogen Corp., Grand Island, NY). cDNA was prepared from 1 μg of total RNA in 20 μl containing 1X first strand buffer, 10 mM DTT, 0.25 mM dNTPs, 0.015 μg/mL random primers, and 15 U/μl M-MLV (Invitrogen Corp., Grand Island, NY). Reverse transcription reactions were incubated in a GeneAmp 2400 Thermocycler (Eppendorf, Westbury, NY) for 1 h at 37° C, followed by 10 min at 67° C.
The relative changes in SH-SY5Y MOR expression were determined using the 2−ΔΔCt method (Livak and Schmittgen, 2001). Real-time PCR was performed using 1 μl of cDNA as the template in 20 μl containing 1X Universal PCR Master Mix, 0.4 μM probe, and 0.4 μM of both sense and antisense primers in a 7900 HT Fast Real Time PCR System (Applied Biosystems Inc., Foster City, CA) using the following cycling parameters: 2 min at 50° C, 10 min at 95° C, 40 cycles for 15 s at 95° C, and 1 min at 60° C. MOR cDNA was amplified using a TaqMan probe: 5′ /56-FAM/CTT-GCG-CCT-CAA-GAG-TGT-CCG-CA/3BHQ_1/-3′; sense primer 5-TAC-CGT-GTG-CTA-TGG-ACT-GAT-3; and antisense primer 5-ATG-ATG-ACG-TAA-ATG-TGA-ATG-3. GAPDH cDNA was used as an internal control and amplified using the TaqMan probe: 5′-56FAM/CCC-CAC-TGC-CAA-CGT-GTC-AGT-G/3BHQ-3′; sense primer 5′-GGA-AGC-TCA-CTG-GCA-TGG-C-3′; and antisense primer 5′-TAG-ACG-GCA-GGT-CAG-GTC-CA-3′. Probes and primers used in this study were synthesized by Integrated DNA Technologies, Coralville, IA.
Inflammatory cytokines secreted by TPA-HL-60 cells were measured using a 96-well human inflammatory cytokine tissue culture kit with slight modification to the manufacturer’s procedure (Meso Scale Discovery, Gaithersberg, MD). The kit selected for this study was used to assay for TNF-α, GM-CSF, IL-1β, IL-8, IL-10, IL-12p70, IL-2, IL-6, and INFγ. After treatment, supernatants were centrifuged at 12,000 x g for 1 min in order to remove cellular debris. Culture filtrates were then stored at −80° C until the assays were performed. Assay wells were initially blocked with 1% (w/v) milk for 1 h at room temperature (RT) on a plate shaker at a speed setting of 5 (Lab-line Instruments, Inc. Melrose Park, IL). Measurement of electrochemiluminescent signal intensity was determined on the SECTOR 2400 instrument (Meso Scale Discovery, Gaithersberg, MD). Calibrator solutions were diluted in RPMI 1640 medium supplemented with 20% FBS, 100 U penicillin, and 100 μg/mL streptomycin in a concentration range of 10,000 to 2.4 pg/mL. Background signal (medium alone) was subtracted, and a linear regression model was used to fit the data.
TPA-HL-60 cells were incubated with and without LPS (0.5 mg/mL) for 24 h. After LPS treatment, the CM was centrifuged for 90 min at 1,635 x g using a Sorvall 6000D centrifuge (Sorvall, Newton, CT). At the end of centrifugation, the column filtrate was collected and the retentate was re-suspended in RPMI 1640 medium supplemented with 20% FBS, 100 U penicillin, and 100 μg/mL streptomycin to the original volume. The CM was then overlaid onto sub-confluent SH-SY5Y cells for 24 h.
TPA-HL-60 cells were grown with and without LPS (0.5 mg/mL) for 24 h. After LPS treatment, TPA-HL-60 supernatants were pooled, filter sterilized, and treated with either anti-TNF-α (10 μg/mL), anti-GM-CSF (2 μg/mL), or both anti-TNF-α and anti-GM-CSF for 1 h at 37° C (Gomes et al., 2005). After neutralization, the CM was overlaid onto actively growing SH-SY5Y cells for 24 h.
Data in this study are represented as the mean ± SD. Differences among treatment groups were analyzed by a one-way ANOVA, followed by a Tukey’s post hoc test or a Bonferroni’s Multiple Comparison Test.
HL-60 cells, differentiated for 48 h with 16 nM TPA (TPA-HL-60 cells), were treated with 0.125, 0.250, or 0.500 LPS for 24 h. Increased ROS production and intracellular accumulation of ROS have been reported in rat cardiomyocytes (Yuan et al., 2009) and rat microglia (Wang et al., 2004) using these concentrations of LPS. After LPS treatment, 20 μM of the ROS indicator, DHR123, was added for 30 min. DHR123 passively diffuses into cells where it undergoes oxidation to form the fluorophore, rhodamine 123 (Henderson and Chappell, 1993). CLSM was used to examine intracellular accumulation of ROS in the TPA-HL-60 cells. Saline solution (0.9% w/v), added at an equivalent volume, served as a vehicle control. Representative confocal images of TPA-HL-60 cells treated with LPS, then DHR123, in order to assess intracellular ROS accumulation are shown in Fig. 1A. TPA-HL-60 cells treated with 0.250 and 0.500 mg/mL LPS significantly increased intracellular ROS when compared to the vehicle control (Fig. 1B).
To further investigate if ROS accumulation in TPA-HL-60 cells was directly related to LPS treatment, we then determined whether pre-treatment with the anti-oxidant, Vitamin E (VE), could block the LPS-induced accumulation of ROS in the TPA-HL-60 cells. The concentration of VE used in this study to block the accumulation of ROS is in the concentration range reported by Wu et al. (2007). Figure 2A shows representative confocal images of TPA-HL-60 cells treated with 0.500 mg/mL LPS alone (LPS), 100 μM VE alone for 3 h (VE), or pre-treated with 100 μM VE for 3 h prior to treatment with 0.500 mg/mL LPS (VE+LPS). As expected, LPS treatment alone significantly increased ROS by approximately 60% when compared to the vehicle control (0.9% saline + 0.1% ethanol) [Fig. 2B]. There was no significant difference in ROS in the cells treated with VE alone compared to control. ROS accumulation was inhibited in the cells treated with VE+LPS compared to LPS alone (Fig. 2B).
TPA-HL-60 cells were treated with 0.500 mg/mL LPS alone (LPS) for 24 h, or pretreated with 100 μM VE for 3 h prior to LPS treatment (VE+LPS). Relative real-time PCR showed that LPS at that concentration increased MOR expression by approximately 50% compared to control (Fig. 3). VE alone did not have a significant effect on MOR expression; however, VE pre-treatment (VE+LPS) attenuated LPS-induced MOR expression compared to LPS alone (Fig. 3).
We next used relative real-time PCR to examine if LPS-induced ROS accumulation in the TPA-HL-60 cells would have an effect on MOR expression in a neuronal cell line (SH-SY5Y). The CM from the TPA-HL-60 cells treated with LPS alone, VE alone, or VE+LPS was overlaid onto actively growing SH-SY5Y cells for 24 h. CM from the LPS-treated (CM, LPS) TPA-HL-60 cells significantly increased MOR expression in the SH-SY5Y cells by approximately 50% when compared to the vehicle control (Fig. 4), whereas VE treatment alone (CM, VE) had no significant effect on MOR expression compared to control. MOR expression in SH-SY5Y cells incubated with CM from the VE+LPS treatment group (CM, VE+LPS) was not significantly different from the CM, LPS group (Fig. 4). Worth noting, initial pilot experiments were performed to examine the effect of LPS on MOR expression in SH-SY5Y cells. We found that LPS at 0.500 mg/mL did not significantly change the level of MOR expression in SH-SY5Y cells when compared to vehicle control (data not shown).
We next used an ultrafiltration process to determine if the increased MOR expression in the SH-SY5Y cells incubated with CM from LPS-treated TPA-HL-60 cells was a cytokine mediated event. Cytokines that are known to modulate MOR expression are larger than 3 kDa; therefore, ultrafiltration serves as an effective method for collection of CM void of cytokine mediators (Curfs et al., 1997). TPA-HL-60 cells were treated with 0.500 mg/mL LPS for 24 h. CM from both the vehicle control (0.9% saline) and LPS-treated cultures were separated into retentate (molecules > 3 kDa) and filtrate (molecules < 3 kDa) fractions. These fractions were then overlaid onto actively growing SH-SY5Y cells for 24 h. There was a significant 40% increase in SH-SY5Y MOR expression in the CM, LPS retentate compared to the CM, vehicle control retentate. However, the CM, LPS filtrate did not significantly affect SH-SY5Y MOR expression when compared to the CM, vehicle control filtrate CM (Fig. 5).
The presence of pro-inflammatory cytokines in the CM from the TPA-HL-60 cells was determined using a sandwich immunoassay in conjunction with an electrochemiluminescent compound. TPA-HL-60 cells treated for 24 h with 0.500 mg/mL LPS exhibited increased IL-1β, IL-2, IL-6, IL-10, IL-12p70, IFN-γ, TNF-α (Fig. 6A), and GM-CSF (Fig. 6B) secretion, but decreased IL-8 secretion (Table 1) compared to vehicle control. Pre-treatment of TPA-HL-60 cells with VE (VE+LPS) for 3 h significantly attenuated the LPS-induced secretion of both TNF-α and GM-CSF (Figs. 6A and 6B) compared to LPS treatment alone. CM from the VE alone treatment significantly increased IL-8 and IL-12p70 levels when compared to the vehicle control (Table 1).
We next determined the effects of neutralization of the TNF-α and GM-CSF present in TPA-HL-60 CM on MOR expression in SH-SY5Y cells. In these experiments, anti-TNF-α, anti-GM-CSF, or both were added to the CM from the TPA-HL-60 cells, with and without LPS treatment, for 1 h at 37° C prior to being overlaid onto actively growing SH-SY5Y cells.
As expected, CM from the LPS-stimulated TPA-HL-60 cells significantly increased MOR expression in the SH-SY5Y cells (Fig. 7A, B, C). SH-SY5Y MOR expression was significantly increased by approximately 50% following neutralization of GM-CSF present in the CM from both the vehicle control (vehicle control+anti-GM-CSF) and LPS treatment (LPS+anti-GM-CSF) groups compared to vehicle control alone (Fig. 7A). However, neutralization of TNF-α had no effect on SH-SY5Y MOR expression in the vehicle control+anti-TNF-α group, and only a slight, non-significant increase in MOR expression in the LPS+anti-TNF-α group (Fig. 7B). Simultaneous neutralization of the GM-CSF and TNF-α present in the TPA-HL-60 CM significantly increased SH-SY5Y MOR expression by 60% in the vehicle control-anti-GM-CSF+anti-TNF-α group and by 50% in the LPS-anti-GM-CSF+anti-TNF-α group when compared to the vehicle control (Fig. 7C).
In this study, we found that LPS stimulated the intracellular accumulation of ROS and the expression of the MOR in TPA-HL-60 cells, a macrophage cell model, and that CM from those cultures significantly increased MOR expression in SH-SY5Y cells, a neuronal cell model. These findings suggest that an indirect ROS signaling mechanism could be responsible, at least in part, for the modulation of LPS-stimulated MOR expression in SH-SY5Y cells.
Cytokines have been shown to regulate MOR expression (Borner et al., 2004; Chang et al., 1998; Kraus, 2009; Kraus et al., 2003; Wei and Loh, 2002). Our lab previously demonstrated that co-treatment of microvascular endothelial cells with IL-1α and IL-1β increases MOR expression (Vidal et al., 1998), and that IL-1 is the cytokine responsible for LPS-induced up-regulation of MOR expression in the rat mesentery (Chang et al., 2001). In this study, our data indicated that cytokines present in the CM from LPS-treated TPA-HL-60 cells are involved in the modulation of MOR mRNA expression in the SH-SY5Y cells. However, it remains to be determined whether the increases in MOR mRNA observed in this study correlate to changes at either the translational or functional level.
To test our hypothesis that ROS plays a role in the LPS-stimulated cytokine secretion in TPA-HL-60 cells, we examined the levels of pro-inflammatory cytokines secreted by TPA-HL-60 cells in response to LPS-induced ROS accumulation. Of the cytokines measured, both TNF-α and GM-CSF levels were found to be significantly decreased when ROS accumulation was blocked by the antioxidant, VE.
TNF-α appears to have a positive effect on SH-SY5Y MOR expression. TNF-α stimulates transcription initiation sites on the MOR gene (Borner et al., 2002), and TNF-α increases MOR expression in human T lymphocytes, Raji B cells, U937 monocytes, primary human polymorphonuclear leukocytes, and mature dendritic cells (Kraus et al., 2003).
GM-CSF down-regulates MOR expression in dendritic cells, which may involve inhibitory actions on IL-4 (Kraus et al., 2003). We observed an increase in MOR expression in response to the neutralization of GM-CSF in vehicle control+anti-GM-CSF and LPS+anti-GM-CSF treatment groups when compared to the vehicle control. These data suggest that only low levels of GM-CSF, present in the CM, are necessary to modulate MOR expression since the concentration of GM-CSF in the vehicle control was significantly less than that of the LPS treatment group (Table 1).
We also observed that neutralization of TNF-α in the LPS treatment group only partially attenuated MOR expression. This partial attenuation of MOR expression may be due to the effects of pro-inflammatory cytokines other than TNF-α in the up-regulation of the MOR. Interestingly, when both TNF-α and GM-CSF were neutralized simultaneously (LPS+anti GM-CSF+anti TNF-α), MOR expression was not significantly decreased compared to the LPS treatment group. This indicates that GM-CSF’s inhibitory effects on MOR expression only becomes apparent when cytokines capable of increasing MOR expression are neutralized, e.g., TNF-α. This again underscores the possibility that cytokines capable of increasing MOR expression can compensate for one another since the simultaneous neutralization of both GM-CSF and TNF-α (LPS+anti GM-CSF+anti TNF-α) did not significantly reduce MOR expression when compared to the LPS treatment group.
We hypothesized that, since morphine potentiates LPS cytotoxicity, activation of the opioid pathway by morphine in a clinical setting could cause an adverse physiological response, i.e., acceleration of sepsis to septic shock (Chang et al., 2001; Ocasio et al., 2004). The damaging consequences of endotoxic shock resulting from exposure to LPS, and the subsequent signaling actions mediated by ROS may, in fact, be exacerbated as a result of the immunosuppressive effects associated with MOR activation in neuronal and non-neuronal cells (Gaveriaux-Ruff, Matthes et al., 1998; Wang, Charboneau et al., 2002).
Our findings suggest that the LPS-induced ROS signaling that occurs in immune cells may indirectly regulate the opiodergic pathway by modulating MOR expression in neurons. Our data also indicate that ROS, produced in LPS challenged TPA-HL-60 cells, is involved in modulating the secretion of TNF-α and GM-CSF, two cytokines that have previously been shown to modulate MOR expression. This mechanism, i.e., LPS-induced ROS production coupled to cytokine secretion in immune cells, can impact molecular events in neurons, and highlights one possible way endotoxin exposure resulting from bacterial infection promotes an interaction between the nervous and immune systems.
The authors would like to express their thanks to Dr. Louaine Spriggs for her editorial assistance on this manuscript.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1This study was supported, in part, by the National Institutes of Health/National Institute on Drug Abuse (R01 DA007058 and K02 DA016149 to SLC).