Analgesic effects of serotonergic, noradrenergic or dual reuptake inhibitors in the carrageenan test in rats: Evidence for synergism between serotonergic and noradrenergic reuptake inhibition

Neuropharmacology 51 (2006) 1172e1180 www.elsevier.com/locate/neuropharm

Analgesic effects of serotonergic, noradrenergic or dual reuptake inhibitors in the carrageenan test in rats: Evidence for synergism between serotonergic and noradrenergic reuptake inhibition Carrie K. Jones1, Brian J. Eastwood, Anne B. Need, Harlan E. Shannon* Lilly Research Laboratories, Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN 46285, USA Received 12 March 2006; received in revised form 27 July 2006; accepted 3 August 2006

Abstract The efficacy of antidepressant drugs with serotonergic, noradrenergic, or dual reuptake inhibition was evaluated in reversing carrageenaninduced thermal hyperalgesia and mechanical allodynia in rats. Duloxetine (1e30 mg/kg, i.p.), a balanced serotonergicenoradrenergic reuptake inhibitor (SNRI), was equiefficacious and more potent than the SNRI venlafaxine (3e100 mg/kg, i.p.) in reversing both thermal hyperalgesia and mechanical allodynia induced by carrageenan. In addition, the selective noradrenergic reuptake inhibitors (NRIs) thionisoxetine (0.03e 10 mg/kg, i.p.) and desipramine (1e30 mg/kg, i.p.) also produced complete reversals of carrageenan-induced thermal hyperalgesia. However, only thionisoxetine exhibited a greater than 80% reversal of the carrageenan-induced mechanical allodynia. In contrast, the selective serotonergic reuptake inhibitors (SSRIs) paroxetine, sertraline, and fluoxetine (0.3e10 mg/kg i.p.) had little or no effect in the carrageenan model. In order to understand whether the observed enhanced effectiveness of the dual SNRIs was due to a possible synergism between serotonergic and noradrenergic reuptake inhibition, the effects of the NRI thionisoxetine alone and in combination with an inactive dose of the SSRI fluoxetine were determined. In the presence of fluoxetine, the potency of thionisoxetine in reversing carrageenan-induced hyperalgesia and allodynia was significantly increased by approximately 100-fold and brain concentrations of thionisoxetine were increased by 1.1- to 5-fold. The present data indicate fluoxetine pharmacodynamically potentiated the analgesic effects of thionisoxetine over and above a metabolic interaction between these two drugs. The present findings thus indicate that, in the carrageenan model, dual serotonergicenoradrenergic reuptake inhibition by dual SNRIs, or SSRIeNRI combinations, produces synergistic analgesic efficacy. Ó 2006 Elsevier Ltd. All rights reserved. Keywords: Duloxetine; Noradrenergic reuptake inhibition; Serotonergic reuptake inhibition; Mixed serotonergicenoradrenergic reuptake inhibition; Carrageenan test; Persistent pain

1. Introduction Antidepressant drugs with serotonergic and/or noradrenergic reuptake inhibition activity are frequently used for the treatment of persistent pain associated with various clinical conditions including diabetic neuropathy (e.g., Max, 1995; Sindrup and * Corresponding author. Tel.: þ1 317 276 4360; fax: þ1 317 651 6346. E-mail address: [email protected] (H.E. Shannon). 1 Present address: Department of Pharmacology, Vanderbilt University Medical Center, 23rd Avenue South at Pierce, 417C Preston Research Building, Nashville, TN 37232-6600, USA. 0028-3908/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuropharm.2006.08.005

Jensen, 1999; Rowbotham et al., 2004), postherpetic neuralgia (Kishore-Kumar et al., 1990), back injury (e.g., Atkinson et al., 1999), and fibromyalgia (e.g. Goldenberg et al., 1996). However, in comparison, selective serotonergic reuptake inhibitors appear to have marginal efficacy in reducing pain symptoms (Sindrup et al., 1990; Max, 1995; Atkinson et al., 1999). While persistent pain arises from multiple etiologies, it is commonly characterized by the phenomenon of central sensitization, an altered responsiveness of dorsal horn neurons and expansion of their receptive fields to afferent stimulation (Devor and Wall, 1978; Woolf et al., 1994), resulting in hyperalgesia (an increased sensitivity to noxious stimuli) and/or allodynia

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

(a nocifensive response to previously non-noxious stimuli) (Woolf and Doubell, 1994). Under normal conditions, descending serotonergic and noradrenergic inhibitory neuronal pathways, projecting from regions of the brainstem to primary afferent fiber terminals in spinal cord, dorsal horn neurons, and inhibitory interneurons within the dorsal horn of the spinal cord, provide a supraspinal control on the spinal transmission of nociceptive information (Fields and Basbuam, 1994; Jones, 1991; Millan, 2002). However, under conditions of inflammation or nerve injury, dysfunction of descending serotonergic and noradrenergic inhibitory pathways appears to contribute to the initiation and propagation of central sensitization and subsequent development of persistent pain (Fields and Basbaum, 1994; Traub, 1997; Millan et al., 2002). One possible mechanism of action of antidepressants in the treatment of persistent pain is the enhancement of serotonergic and noradrenergic descending inhibition by the blockade of serotonergic and/or noradrenergic reuptake. To date, however, the relative contribution of noradrenergic and serotonergic reuptake inhibition required for effective analgesia remains uncertain. Duloxetine, a dual serotonergicenoradrenergic reuptake inhibitor (SNRI), originally developed for the treatment of major depression (Detke et al., 2002, 2004), has also been demonstrated to be clinically efficacious in alleviating painful symptoms of fibromyalgia (Arnold et al., 2004) and painful diabetic neuropathy (Goldstein et al., 2005). In addition, duloxetine has been shown to be efficacious in the prevention and/or reversal of pain in several preclinical models of inflammatory and neuropathic pain in rodents. Iyengar et al. (2004) demonstrated that duloxetine significantly attenuated late phase paw-licking behavior in the formalin model and also reversed mechanical allodynia behavior in the L5/L6 spinal nerve ligation model of neuropathic pain. Moreover, in the formalin model, inactive doses of the NRI thionisoxetine were efficacious when combined with an inactive dose of the SSRI paroxetine. Similarly, Bomholt et al. (2005) demonstrated that duloxetine attenuated flinching behavior in the late phase of the formalin model and attenuated thermal and mechanical hyperalgesia, but not

1173

mechanical allodynia, in the chronic constriction injury model. We (Jones et al., 2005) have previously demonstrated that duloxetine was efficacious in the acetic acid-induced writhing test and reversed carrageenan-induced thermal hyperalgesia and mechanical allodynia as well as capsaicin-induced mechanical allodynia. Duloxetine was efficacious in these models at doses (1e30 mg/kg) that produced little or no effect on either acute nociception or motor output (Iyengar et al., 2004; Bomholt et al., 2005; Jones et al., 2005). Duloxetine also has little or no affinity for other neurotransmitter receptors, including H1 histamine, a1-adrenergic and muscarinic receptors, or ion channels (Wong et al., 1988, 1993; Fuller et al., 1994; Bymaster et al., 2001) which might produce dose-limiting side effects, such as sedation, orthostatic hypotension, and cardiac conduction abnormalities, observed with the tricyclic antidepressants desipramine and amitriptyline (Max, 1995). In order to further pharmacologically characterize the underlying mechanisms of action of duloxetine in the treatment of persistent pain, we evaluated the effects of antidepressant drugs with varying affinities for serotonin and/or noradrenaline transporters (see Table 1) in reversing persistent inflammatory pain as ascertained by carrageenan-induced thermal hyperalgesia and mechanical allodynia. To this end, dosee response curves were determined for the dual transporter inhibitor duloxetine in comparison with another dual transporter inhibitor venlafaxine, with preferential selectivity for the serotonin transporter, in reversing both carrageenan-induced thermal hyperalgesia and mechanical allodynia. For comparison, doseeresponse curves were determined for the selective noradrenergic reuptake inhibitors (NRIs) desipramine and thionisoxetine, as well as the selective serotonin reuptake inhibitors (SSRIs) fluoxetine, paroxetine, and sertraline in reversing both carrageenan-induced thermal hyperalgesia and mechanical allodynia. In addition, to ascertain whether the mechanisms of serotonergic and noradrenergic reuptake inhibition act synergistically, the effects of the selective NRI thionisoxetine administered alone or concomitantly with the SSRI fluoxetine were evaluated. We hypothesized that if these two mechanisms

Table 1 Affinities (Ki, nM) of the antidepressant drugs tested in the present study for the serotonergic and noradrenergic transporters and the H1 histamine, a1-adrenergic, a2-adrenergic, and muscarinic (nonselective) receptors in rat brain tissue Antidepressant

Thionisoxetine Desipramine Duloxetine Venlafaxine Fluoxetine Paroxetine Sertraline

Monoamine transporters (Ki, nM)

Receptors (Ki, nM)

Clinical efficacy

5-HT

NE

Ratio NE/5-HT

H1

a1

a2

Muscarinic

44.1a 129  7b 0.8  0.04c 19  1.7b 2  0.1b 0.05  0.0003b 0.29  0.01b

0.2a 0.31  0.01b 7.5  0.3c 1067  29b 473  11b 59  0.7b 1597  44b

0.005 0.002 9.38 56.16 236.5 1180 5506

>1000a 31  1b 2300d 12909  1075b 933  23b 13746  404b 5042  165b

>1000a 23  1b 8300d 39921  810b 1353  17b 995  35b 36  2b

>1000a 1379  39b 8600c >100000b 3090  121b 3915  114b 477  17b

>1000a 37  1b 3000d 29966  1464b 512  12b 42  2b 232  1b

SE, dose-limiting side effects; NA, not available. a Gehlert et al., 1995; personal communication. b Owens et al., 1997. c Bymaster et al., 2001. d Wong et al., 1988, 1993. e See Max, 1995; Sindrup and Jensen, 1999; and text. f Arnold et al., 2004; Goldstein et al., 2005.

NA Yese/SE Yesf Yese Nonee Nonee Nonee

1174

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

are synergistic, then the effects of thionisoxetine in reversing carrageenan-induced hyperalgesia and allodynia would be potentiated in the presence of fluoxetine. Since fluoxetine could potentially alter the metabolism of thionisoxetine, we also measured the brain levels of thionisoxetine alone and in the presence of fluoxetine to address any pharmacokinetic contributions to the drug combination. Our present findings demonstrate that the analgesic potency of thionisoxetine was significantly enhanced and that brain levels of thionisoxetine were also increased in the presence of fluoxetine. However, the shift in the analgesia doseeresponse curve was significantly greater in magnitude than can be explained by the increase in brain levels alone, indicating that the mechanisms of serotonergic and noradrenergic reuptake inhibition act in a pharmacologically synergistic manner to produce analgesia in this model. 2. Methods 2.1. Subjects Male SpragueeDawley rats (Harlan Sprague Dawley, Indianapolis, IN) weighing 70e90 g were used in all the experiments. Rats were housed in groups up to six per cage in a large colony room on a 12:12-h light/dark cycle (lights on 06:00 h), with food and water provided ad libitum. Each animal was used only once. Test sessions were performed between 08:00 and 18:00 h. All dose groups consisted of 6e12 rats. All experiments were conducted in accordance with the NIH regulations of animal care covered in ‘‘Principles of Laboratory Animal Care’’, NIH publication 85-23, and were approved by the Institutional Animal Care and Use Committee.

2.2. Carrageenan-induced thermal hyperalgesia and mechanical allodynia Groups of 6 to 12 rats were injected s.c. with l-carrageenan (100 ml of a 1.5% solution) into the plantar surface of the right hind paw at time zero followed 90 min later by an i.p. injection of vehicle or a dose of drug. In the thionisoxetineefluoxetine interaction experiments, an i.p. injection of vehicle or dose of thionisoxetine alone or in the presence of a 10 mg/kg dose of fluoxetine was administered 90 min after carrageenan injection. The 10 mg/kg dose of fluoxetine was chosen because it was the highest dose tested that did not produce significant effects. Withdrawal responses to mechanical and thermal stimuli were determined after an additional 30 and 40 min, respectively. Mechanical allodynia was evaluated with a calibrated series of von Frey filaments as previously described (Chaplan et al., 1994; Jones et al., 2005). Briefly, rats were placed in clear plastic cages (17.5  15  15 cm) fitted with wire mesh flooring and allowed to acclimate for approximately 5 min; the withdrawal thresholds were determined by the up and down method by applying each filament to the midplantar surface of each hind paw in a perpendicular fashion and depressed slowly (4e5 s) until bending occurred and the maximum force of the fiber was exerted. A paw withdrawal response was scored as a response to the filament. Withdrawal latencies to a nociceptive thermal stimulus were assessed as previously described (Jones et al., 2005). Briefly, each rat was placed in a Plexiglas cubicle with a glass floor through which an infrared photobeam was shown onto the plantar surface of the hind paws and the latency to withdrawal from the thermal stimulus was determined. The intensity of the infrared photobeam from the plantar reflex device (Plantar Test, Ugo Basile) was adjusted to produce a mean response latency in untreated rats of approximately 12e15 s. The response latency was determined using a timer linked to the photodiode motion sensors in the plantar reflex device. Response latency was defined as the time from the onset of exposure to the infrared photobeam to the cessation of the photobeam when the photodiode motion sensors detected the withdrawal response of the paw of the rat. Response to the thermal stimulus

was reported as the withdrawal latency differences between the treated and untreated paws in seconds.

2.3. Determination of brain concentrations of thionisoxetine alone and in the presence of fluoxetine Separate groups of rats were administered i.p. doses of thionisoxetine with or without simultaneous doses of fluoxetine in order to assess the possibility of a pharmacokinetic interaction between thionisoxetine and fluoxetine. Groups of eight rats were administered thionisoxetine (0.03, 0.1, 0.3, or 1 mg/kg) and either 10 mg/kg of fluoxetine or vehicle. Only doses of thionisoxetine related to thermal hyperalgesia were studied in order to reduce the number of animals used. All rats were also dosed with carrageenan 90 min prior, as above. The animals were killed by decapitation 30 min after thionisoxetine dosing. Brains were removed and frontal cortex (anterior pole) was dissected and placed in a centrifuge tube on ice. The tissue was homogenized in four volumes of acetonitrile containing 0.1% formic acid using an ultrasonic probe and centrifuged at 14,000 rpm for 16 min (Eppendorf model 5417R, Hamburg, Germany). The supernatant was diluted 4-fold with water and injected by autosampler onto an HPLC (Shimadzu, Columbia, MD). Gradient elution was employed, beginning with 20% acetonitrile in water and rising linearly over 2.5 min to a final concentration of 90% acetonitrile, all containing 0.1% formic acid. The column was a 3.5-mm Zorbax Eclipse XDB-C18 in a 2.1  50 mm format (part number 971700-902, Agilent Technologies, Palo Alto, CA). A triple quad mass spectrometer (Model API3000, Applied Biosystems, Foster City, CA) in positive MRM mode monitored the ion transition from parent to daughter ion (288.1/148.1). A standard curve was generated by adding known quantities of thionisoxetine to brain tissue from untreated rats and processing as described above.

2.4. Drugs Duloxetine HCl, fluoxetine HCl, paroxetine, and thionisoxetine (Lilly Research Laboratories, Indianapolis, IN), sertraline HCl and desipramine HCl (RBI Inc., Natick, MA), and venlafaxine (Bergen Brunswig Drug Co., Louisville, KY) were dissolved in double deionized water. Doses refer to the form of the drug listed. All drugs were administered i.p. in a volume of 1.0 ml/kg in rats. l-Carrageenan (Sigma, St. Louis, MO) was dissolved in double deionized water and injected s.c. (100 ml of a 1.5% solution) into the plantar surface of the right hind paw.

2.5. Statistical analysis Data are expressed as mean  S.E.M.. In all behavioral experiments, treatment groups were compared to appropriate control groups using one-way ANOVA and Dunnett’s t-test. A probability of p < 0.05 was taken as the level of statistical significance. ED50 values were determined using a 4-parameter logistic (4PL) equation with the bottom fixed at the vehicle/vehicle mean response (DeLean et al., 1978). Mean brain concentration levels of thionisoxetine in the presence or absence of 10 mg/kg of fluoxetine were compared by two-way ANOVA with interaction. The main effect term for fluoxetine was used to test the overall effect of the fluoxetine dose on thionisoxetine concentration levels. Bonferroni-adjusted contrasts were used to test the effect of fluoxetine on specific concentrations of thionisoxetine. A log transformation was applied to the concentrations to stabilize the variation. Dose proportionality was examined by linear regression between log-concentration versus logdose. In the thionisoxetineefluoxetine interaction experiment, the dosee response curves were estimated simultaneously by a 4PL, with the bottom fixed at the vehicle/vehicle mean response, in order to estimate the relative potency of thionisoxetine in the presence and absence of fluoxetine. In addition, the brain-concentration response curves for thionisoxetine in the presence and absence of fluoxetine were estimated simultaneously by a 4PL in order to estimate the relative potency of thionisoxetine based on brain concentrations of thionisoxetine in the presence and absence of fluoxetine. Relative potency was defined as the ratio of ED50 values for the simultaneously fitted 4PL systemic doseeresponse curves, and the ratio of EC50 values of the simultaneously fitted 4PL brain concentrationeresponse curves. Relative potency differences

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180 were considered significant when the 95% confidence interval (CI) for the relative potency did not include the value of 1.0. Statistical analyses were performed using JMP statistical software (SAS Institute Inc., Cary, NC).

3. Results

1175

doses of 3 and 10 mg/kg. Desipramine (1e30 mg/kg i.p.) also produced a full reversal of the carrageenan-induced thermal hyperalgesia (left panel, Fig. 2), which was significant at doses of 10 and 30 mg/kg, but produced only a 50% reversal of the carrageenan-induced mechanical allodynia, significant at 10 and 30 mg/kg.

3.1. Efficacy of SNRIs 3.3. Lack of efficacy of SSRIs Duloxetine (1e30 mg/kg, i.p.) produced a complete, dosedependent reversal of carrageenan-induced thermal hyperalgesia, which was significant at doses of 3, 10, and 30 mg/kg (left panel, Fig. 1). Venlafaxine also produced a dose-related reversal of the carrageenan-induced thermal hyperalgesia, with significant effects at doses of 10, 30, and 100 mg/kg (left panel, Fig. 1). The rank order of potencies and ED50 (95% CI) values in mg/kg of compounds in reversing the carrageenan-induced thermal hyperalgesia (left panel, Fig. 1) was duloxetine [4.5 (1.3e12.8)] > venlafaxine [26 (5.8e42.6)]. Duloxetine also produced a complete reversal in the carrageenan-induced mechanical allodynia, significant at doses of 10 and 30 mg/kg (right panel, Fig. 1). In the case of venlafaxine, doses of 30 and 100 mg/kg were required to significantly inhibit the carrageenan-induced mechanical allodynia (right panel, Fig. 1). The rank order of potencies and ED50 values in mg/kg of the two compounds in reversing the carrageenan-induced mechanical allodynia (right panel, Fig. 1) was duloxetine [10.8 (8.6e12.9)] > venlafaxine [29.6 (22.3e41.8)]. 3.2. Efficacy of NRIs The selective NRIs thionisoxetine and desipramine also produced dose-related reversals of both carrageenan-induced thermal hyperalgesia (left panel, Fig. 2) and mechanical allodynia (right panel, Fig. 2). In particular, thionisoxetine (0.03e 10 mg/kg i.p.) produced a complete, dose-dependent reversal of the carrageenan-induced thermal hyperalgesia (left panel, Fig. 2), which was significant at 0.3, 1, 3, and 10 mg/kg and mechanical allodynia (right panel, Fig. 2), significant after 4

3.4. Thionisoxetine in the presence of fluoxetine Finally, in order to directly assess whether the mechanisms of serotonergic and noradrenergic reuptake inhibition were synergistic in reversing the carrageenan-induced thermal hyperalgesia and mechanical allodynia, the effects of the NRI thionisoxetine alone and in combination with an inactive dose of the SSRI fluoxetine were evaluated. As previously shown in Fig. 2, thionisoxetine when administered alone, over a dose range of 0.01 to 10 mg/kg i.p., produced a robust dose-dependent reversal of the carrageenan-induced thermal hyperalgesia (left panel, Fig. 4) and mechanical allodynia (right panel, Fig. 4). When administered alone, 10 mg/kg of fluoxetine had no effect on either the carrageenan-induced thermal hyperalgesia or mechanical allodynia (see points above 10F in left and right panels, Fig. 4). However, the effects of thionisoxetine were robustly enhanced in the presence of 10 mg/kg fluoxetine in reversing both the carrageenan-induced thermal hyperalgesia and mechanical allodynia. In particular, there was a significant leftward shift 16

*

0

*

-2

*

*

-4 -6 -8 -10 -12

Withdrawal Threshold (gms)

*

2

Withdrawal Latency Difference (secs)

In contrast with the NRIs, the SSRIs paroxetine, sertraline, and fluoxetine had no effect in reversing the carrageenaninduced thermal hyperalgesia over the dose range of 0.3 to 10 mg/kg i.p. (left panel, Fig. 3). Fluoxetine also had no effect in reversing the carrageenan-induced mechanical allodynia, while both paroxetine and sertraline produced modest effects that were significant at 3 mg/kg and 10 mg/kg of sertraline (right panel, Fig. 3).

*

14

Venlafaxine i.p. Duloxetine i.p.

*

12 10

*

*

8 6 4 2 0

-14 Veh

1

3

10

Dose (mg/kg)

30 100

Veh

1

3

10

30

100

Dose (mg/kg)

Fig. 1. The dual serotonin and norepinephrine reuptake inhibitor duloxetine was more potent than venlafaxine in reversing carrageenan-induced thermal hyperalgesia (left panel) and mechanical allodynia (right panel) in rats. Vertical lines represent S.E.M. and are absent when less than the size of the point. Abscissa: Dose of drug in mg/kg. Ordinate: Left panel, withdrawal latency difference in seconds to a thermal stimulus; right panel, withdrawal threshold in grams to a mechanical stimulus. *p < 0.05 vs. Veh, Dunnett’s t-test.

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

1176

16

* *

0

*

-2

*

*

*

-4 -6 -8 -10 -12 -14

14

*

Desipramine i.p. Thionisoxetine i.p.

12

*

10

*

8

*

6 4 2

Dose (mg/kg)

30

3

10

1

Ve h

30

3

10

1

0. 03 0. 1 0. 3

Ve h

0 0. 03 0. 1 0. 3

Withdrawal Latency Difference (secs)

2

Withdrawal Threshold (gms)

4

Dose (mg/kg)

Fig. 2. The selective norepinephrine reuptake inhibitors desipramine and thionisoxetine produced dose-related reversals of carrageenan-induced thermal hyperalgesia (left panel) and mechanical allodynia (right panel) in rats. Vertical lines represent S.E.M. and are absent when less than the size of the point. Abscissa: Dose of drug in mg/kg. Ordinate: Left panel, withdrawal latency difference in seconds to a thermal stimulus; right panel, withdrawal threshold in grams to a mechanical stimulus. *p < 0.05 vs. Veh, Dunnett’s t-test.

in the relative potency (RP) of thionisoxetine in reversing the carrageenan-induced thermal hyperalgesia [RP ¼ 38 (11e 138)] and in reversing carrageenan-induced mechanical allodynia [RP ¼ 19.1 (14.5e27)] in the presence of fluoxetine. 3.5. Effect of fluoxetine on brain concentrations of thionisoxetine When administered alone, the concentration of thionisoxetine in brain tissue increased in a dose-proportional manner as evidenced by a linear relationship between log-concentration and log-dose with the estimate for the slope within 1 standard error of unity (slope estimate ¼ 1.06, standard error ¼ 0.070, R2 ¼ 0.991) (Fig. 5). In the presence of 10 mg/kg of fluoxetine, brain levels of thionisoxetine were significantly increased in a dose-dependent manner (drug  dose interaction F(3,54), p ¼ 0.0304) (Fig. 5), ranging from 1.09-fold at the 0.3 mg/kg dose of thionisoxetine ( p ¼ 0.131) to 5.1-fold at 1 mg/kg

( p ¼ 0.0002). Overall, fluoxetine (10 mg/kg) increased thionisoxetine levels by 3.1-fold (95% CI 2.1e4.6 fold, p < 0.0001). 3.6. Relative potency of thionisoxetine alone versus thionisoxetine þ fluoxetine The doseeresponse curve for the analgesic effects of thionisoxetine in ameliorating thermal hyperalgesia was shifted to the left in the presence of fluoxetine when expressed as a function of the systemic dose (Fig. 4, left panel; only doses related to thermal hyperalgesia were studied in order to reduce the number of animals used). Similarly, although smaller in magnitude, the brain concentrationeresponse curve was also shifted to the left in the presence of fluoxetine (Fig. 6). For the pair of systemic doseeresponse curves, the potency ratio for thionisoxetine in the absence and presence of fluoxetine was 92.5 (95% CI 16.2e2940), and for the pair of brain concentrationeresponse curves the potency ratio for thionisoxetine in the absence and

4

Withdrawal Threshold (gms)

16

Withdrawal Latency Difference (secs)

2 0 -2 -4 -6 -8 -10 -12 -14

14

Paroxetine i.p. Sertraline i.p. Fluoxetine i.p.

12 10 8 6

*

4

*

*

2 0

Veh

0.3

1

Dose (mg/kg)

3

10

Veh

0.3

1

3

10

Dose (mg/kg)

Fig. 3. The selective serotonin reuptake inhibitors paroxetine, sertraline and fluoxetine had little or no effect in reversing carrageenan-induced thermal hyperalgesia (left panel) and mechanical allodynia (right panel) in rats. Vertical lines represent S.E.M. and are absent when less than the size of the point. Abscissa: Dose of drug in mg/kg. Ordinate: Left panel, withdrawal latency difference in seconds to a thermal stimulus; right panel, withdrawal threshold in grams to a mechanical stimulus. *p < 0.05 vs. Veh, Dunnett’s t-test.

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180 4

Withdrawal Latency Difference (secs)

* * * * * * * * * *

0 -2 -4 -6 -8 -10

Thionisoxetine alone i.p.

-12

Withdrawal Threshold (gms)

16

2

1177

* * * *

14 12

*

10 8 6

* *

4 2

+ 10 mg/kg Fluoxetine i.p.

-14

Thionisoxetine (mg/kg)

3

10

10 F 0. 01 0. 03 0. 1 0. 3 1

Ve h

3 10

1

10 F 0. 01 0. 03 0. 1 0. 3

Ve h

0

Thionisoxetine (mg/kg)

Fig. 4. The effects of norepinephrine reuptake inhibitor thionisoxetine were potentiated in combination with an inactive dose of the selective serotonin reuptake inhibitor fluoxetine in reversing carrageenan-induced thermal hyperalgesia (left panel) and mechanical allodynia (right panel) in rats. Vertical lines represent S.E.M. and are absent when less than the size of the point. Abscissa: Dose of drug in mg/kg. Ordinate: Left panel, withdrawal latency difference in seconds to a thermal stimulus; right panel, withdrawal threshold in grams to a mechanical stimulus. *p < 0.05 vs. respective Veh, Dunnett’s t-test.

presence of fluoxetine was 11.9 (95% CI 2.5e113.8). Both shifts are statistically significant since the confidence intervals do not include 1.0, indicating there is a significant increase in the potency when the analgesic effects of thionisoxetine are expressed either as a function of systemic dose or brain concentration. If there were no pharmacodynamic or pharmacokinetic interactions, the relative potencies for both sets of curves would both be equal to 1.0. If the pharmacokinetic interaction were the sole determinant of the enhanced potency then the relative potency would still be equal to 1.0 when the analgesic effects are expressed as a function of brain concentration. However, the magnitude of the analgesic effects of thionisoxetine was significantly increased in the presence of fluoxetine even when expressed as a function of brain concentrations (95% CI did not include 1). We interpret these data to indicate that there is a pharmacodynamic synergism between the NRI thionisoxetine and

Brain Concentration Thionisoxetine (ng/gm)

10000

1000

100 Thionisoxetine alone + 10 mg/kg Fluoxetine

10 0.03

0.1

0.3

1

Thionisoxetine (mg/kg IP) Fig. 5. Mean brain levels of thionisoxetine administered alone and in the presence of 10 mg/kg of fluoxetine. Vertical lines represent S.E.M. and are absent when less than the size of the point. Abscissa: Dose of drug in mg/ kg on a log scale. Ordinate: Brain concentration of thionisoxetine in ng/g on a log scale.

the SSRI fluoxetine over and above a pharmacokinetic interaction.

4. Discussion In the present study, the efficacy of antidepressant drugs with serotonergic and/or noradrenergic reuptake inhibition in reversing carrageenan-induced thermal hyperalgesia and mechanical allodynia were evaluated in rats. The dual serotoninenorepinephrine transporter inhibitors duloxetine and venlafaxine were efficacious in reversing both thermal hyperalgesia and mechanical allodynia. The NRIs thionisoxetine and desipramine were also efficacious. On the other hand, the SSRIs fluoxetine, paroxetine and sertraline had little or no efficacy. However, the SSRI fluoxetine potentiated the analgesic effects of the NRI thionisoxetine, an effect which was produced by a pharmacodynamic potentiation over and above a concomitant metabolic interaction. Since both serotonergic and noradrenergic descending inhibitory pathways are known to modulate painrelated neurotransmission, and chronic inflammatory pain directly suppresses descending serotonergic and noradrenergic pathways (Traub, 1997), the present results are consistent with the interpretation that enhancement of neurotransmission of these two descending inhibitory systems by blockade of neurotransmitter reuptake can act in a synergistic manner to produce analgesic efficacy which is greater than inhibition of either uptake mechanism alone. The present findings replicate and extend previous reports that the dual reuptake inhibitor duloxetine is efficacious in persistent pain models. As mentioned above, duloxetine significantly attenuated nocifensive behavior in the late phase of the formalin model (Iyengar et al., 2004; Bomholt et al., 2005). In neuropathic pain models, duloxetine reversed mechanical allodynia in the L5/L6 spinal nerve ligation model (Iyengar et al., 2004) as well as thermal and mechanical hyperalgesia, but not mechanical allodynia, in the chronic constriction injury model (Bomholt et al., 2005). In inflammatory pain

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

1178

models, duloxetine reversed acetic acid-induced writhing as well as carrageenan-induced thermal hyperalgesia and mechanical allodynia (Jones et al., 2005). Further, duloxetine reversed capsaicin-induced mechanical allodynia, which results from the direct stimulation of C-fibers by capsaicin (Jones et al., 2005). The reasons why duloxetine reversed mechanical allodynia in the L5/L6, carrageenan and capsaicin models but not the chronic constriction injury model are most likely due to methodological differences. Iyengar et al. (2004) and Jones et al. (2005) applied each filament only once according to the up-and-down method described by Chaplan et al. (1994), whereas Bomholt et al. (2005) applied each von Frey filament for 8 to 10 times and defined the threshold by the filament that produced a withdrawal response on 50% of the applications and therefore may have been more difficult to reverse. At the doses efficacious in persistent pain models (3e30 mg/ kg), duloxetine has been demonstrated to increase extracellular concentrations of both serotonin and norepinephrine by as much as 200e300% (5e15 mg/kg; Koch et al., 2003). Further, duloxetine was efficacious in pain models at doses that produced little or no effect on motor output as measured by the rotorod (Iyengar et al., 2004; Bomholt et al., 2005; Jones et al., 2005). Thus, at doses which increase extracellular levels of both serotonin and norepinephrine and which do not impair motor performance, duloxetine is efficacious in neuropathic and inflammatory persistent pain models. The present study represents the first direct demonstration of a synergistic interaction between the mechanisms of serotonergic and noradrenergic reuptake inhibition in reversing carrageenan-induced inflammatory pain in rats. In particular, the analgesic effects of the NRI thionisoxetine were potentiated in

4

Withdrawal Latency Difference (sec)

w/o Fluoxetine

2

w/ Fluoxetine

0 -2 -4 -6 -8 -10 -12 V

10

100

1000

10000

Thionisoxetine (ng/g, brain) Fig. 6. Best-fit curves (solid lines) of the analgesic effects of thionisoxetine administered alone and in the presence of 10 mg/kg of fluoxetine in reversing carrageenan-induced thermal hyperalgesia. Data are plotted as a function of the brain concentration of thionisoxetine. The relative potency of thionisoxetine þ fluoxetine to thionisoxetine alone is 11.9 (95% CI 2.5e113.8) by brain concentration (this figure). Abscissa: Withdrawal latency difference in seconds to a thermal stimulus. Ordinate: Average brain concentration in ng/g of tissue.

the presence of an inactive dose of the SSRI fluoxetine in reversing both carrageenan-induced thermal hyperalgesia and mechanical allodynia. Moreover, this potentiation constituted a true pharmacodynamic interaction over and above a metabolic interaction between the NRI thionisoxetine and the SSRI fluoxetine. The present findings extend those of Iyengar et al. (2004) who reported that inactive doses of thionisoxetine became efficacious when administered concomitantly with paroxetine in the formalin model. The synergistic interaction between thionisoxetine and fluoxetine in the carrageenan model in the present study, and additivity with paroxetine in the formalin model (Iyengar et al., 2004), indicates that dual serotonergice noradrenergic reuptake inhibition may provide enhanced analgesic efficacy relative to either mechanism alone. Consistent with this interpretation, the SNRIs duloxetine and venlafaxine exhibited greater potency and efficacy in reversing the carrageenan-induced thermal hyperalgesia and mechanical allodynia than all of the SSRIs tested in the present study, and were equiefficacious to effects observed with desipramine. Taken together, the data indicate that dual SNRIs like duloxetine or NRI-SSRI combinations may also provide superior analgesia over other clinically available antidepressants for the treatment of persistent inflammatory pain in humans. While antidepressant drugs represent one of the principle classes of analgesics for the treatment of persistent pain (Onghena and Van Houdenhove, 1992; McQuay et al., 1996; Lynch, 2001), their underlying mechanism(s) of action, particularly the relative amount of noradrenergic to serotonergic reuptake inhibition required for effective analgesia, is not yet fully understood (Jasmin et al., 2003). In the present study, we evaluated antidepressant drugs with varying affinities for serotonergic and/or noradrenergic transporters (see Table 1; Koch et al., 2003). The SSRIs fluoxetine, paroxetine and sertraline, with low to subnanomolar affinities for the serotonin transporter and ratios of serotonergic to noradrenergic transporter affinities of greater than 200 (see Table 1), demonstrated little or no efficacy in reversing the carrageenan-induced thermal hyperalgesia and mechanical allodynia. In comparison, the NRIs thionisoxetine and desipramine, with nanomolar affinities for the noradrenergic transporter and ratios of noradrenergic to serotonergic transporter affinities of less than 1 (see Table 1), produced moderate to full reversals of the carrageenan-induced thermal hyperalgesia and mechanical allodynia that were greater in magnitude than the SSRIs tested. Overall, the data suggest that ratios of noradrenergic to serotonergic transporter affinities of less than approximately 100 are necessary in order to observe analgesic efficacy in the carrageenan model (see Table 1). However, reuptake inhibitors with similar affinities for transporters do not necessarily increase extracellular levels of monoamines to the same extent (e.g. Koch et al., 2003). Duloxetine and venlafaxine have approximately 10-fold and 55-fold higher affinities for the serotonin transporter relative to the norepinephrine transporter (see Table 1). However, duloxetine increased extracellular concentrations of both serotonin and norepinephrine 200e 300% of control at doses of 5 and 15 mg/kg i.p., while venlafaxine increased extracellular concentrations of both serotonin

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

and norepinephrine to approximately 175% at a dose of 5 mg/ kg i.p. and to approximately 275% and 200%, respectively, at a dose of 15 mg/kg in spite of the differences in affinities for the two transporters. Thus, while a ratio of noradrenergic to serotonergic transporter affinity of less than 100 appears to be important for analgesic activity, the extent to which a compound functionally inhibits the transporter is also critically important. Carrageenan-induced hyperalgesia and allodynia arise as a consequence of the activation and sensitization of primary nociceptive afferents and the subsequent central sensitization of dorsal horn neurons in the spinal cord with the corresponding expansion of their receptive fields and plasticity of their neuronal connections (see e.g., Devor and Wall, 1978; Woolf et al., 1994). Normally, supraspinal inhibitory mechanisms including descending serotonergic and noradrenergic inhibitory pathways function to dampen or reduce the magnitude of excitation and sensitization of dorsal horn neurons by acute noxious stimuli (Jones, 1991: Millan, 2002). These inhibitory functions are balanced by supraspinal descending serotonergic and noradrenergic facilitatory pathways, that while not functionally as well characterized as the descending inhibitory pathways, function normally to enhance the signal to noise ratio for the perception of noxious stimuli, particularly at the level of the spinal cord, and subsequently facilitate evasive behaviors (Fields and Basbaum, 1994). Previously, Millan (2002) pointed out that there exists no clear anatomical separation of the substrates for the descending inhibitory and facilitatory pathways of the serotonergic or noradrenergic systems. In light of this point, inhibition of serotonergic and/or noradrenergic reuptake would more than likely enhance, to varying degrees, both descending inhibition and facilitation of the serotonergic and/or noradrenergic systems, making the degree of analgesia dependent on the relative balance of these antinociceptive and pronociceptive inputs. Under conditions of peripheral inflammation as modeled in the carrageenan test, dorsal horn neurons appear to become disinhibited, due in part to a decrement in tone of descending serotonergic and noradrenergic inhibitory pathways, and are more easily excited and sensitized by incoming noxious stimuli (e.g., Traub, 1997). Thus, one possible mechanism of action for the analgesic effects observed using the SNRIs and SSRIeNRI combinations may be through the enhancement of descending serotonergic and noradrenergic inhibition of spinal nociceptive processing. In addition, enhancement of serotonergic and noradrenergic neurotransmission by reuptake inhibition may also result in pain relief through peripheral mechanisms since peripheral administration of the tricyclic antidepressants desipramine and amitriptyline has been previously shown to produce analgesia in neuropathic and formalin-induced pain models (Esser and Sawynok, 1999; Sawynok et al., 1999). Finally, another possible mechanism of action for the SNRIs may be through the enhancement of the ascending serotonergic and noradrenergic pathways, projecting to areas of the brain such as the thalamus, hypothalamus, and cortex, which modulate the perception of painful stimuli, thereby decreasing the negative affective components of the persistent pain (e.g., Craig, 1994; Max, 1995).

1179

In summary, the present study demonstrated that dual serotonergicenoradrenergic reuptake inhibitors such as duloxetine and venlafaxine produced greater antihyperalgesic and antiallodynic efficacy than SSRIs and some NRIs in the carrageenan-induced inflammatory pain model in rats. Moreover, the mechanisms of serotonergic and noradrenergic reuptake inhibition acted synergistically in the carrageenan model, and provided enhanced analgesic efficacy relative to either SSRIs or NRIs alone. The present findings suggest that noradrenergic systems may have a predominant role in the control of the perception of noxious stimuli and central sensitization, and that serotonergic systems may play a permissive or facilitatory role. Thus, the present findings are consistent with the interpretation that duloxetine provides enhanced analgesic efficacy through a mechanism of action involving the synergistic action of its dual noradrenergic and serotonergic reuptake inhibition. References Arnold, L.M., Lu, Y., Crofford, L.J., Wohlreich, M., Detke, M.J., Iyengar, S., Goldstein, D.J., 2004. A double-blind, multicenter trial comparing duloxetine with placebo in the treatment of fibromyalgia patients with or without major depressive disorder. Arthritis Rheum. 50, 2974e2984. Atkinson, J.H., Slater, M.A., Wahlgren, D.R., Williams, R.A., Zisook, S., Pruitt, S.D., Epping-Jordan, J.E., Patterson, T.l, Grant, I., Abramson, I., Garfin, S.R., 1999. Effects of noradrenergic and serotonergic antidepressants on chronic lower back pain intensity. Pain 83, 137e145. Bomholt, S.F., Mikkelsen, J.D., Blackburn-Munro, G., 2005. Antinociceptive effects of the antidepressants amitriptyline, duloxetine, mirtazapine and citalopram in animal models of acute, persistent and neuropathic pain. Neuropharmacology 48, 252e263. Bymaster, F.P., Dreshfield-Ahmad, L.J., Threlkeld, P.G., Shaw, J.L., Thompson, L., Nelson, D.L., Hemrick-Leucke, S.K., Wong, D.T., 2001. Comparative affinity of duloxetine and venlafaxine for serotonin and norepinephrine transporters in vitro and in vivo, human serotonin receptor subtypes, and other neuronal receptors. Neuropsychopharmacology 25, 871e880. Chaplan, S.R., Bach, F.W., Pogrel, J.W., Chung, J.M., Yaksh, T.L., 1994. Quantitative assessment of tactile allodynia in the rat paw. J. Neurosci. Methods 53, 55e63. Craig, K.D., 1994. Emotional aspects of pain. In: Wall, P.D., Melzack, R. (Eds.), Textbook of Pain. Churchill Livingstone, Edinburgh, pp. 261e274. DeLean, A., Munson, P.J., Rodbard, D., 1978. Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay and physiological doseeresponse curves. Am. J. Physiol. 235, E97eE102. Detke, M.J., Lu, Y., Goldstein, D.J., McNamara, R.K., Demitrack, M.A., 2002. Duloxetine 60 mg once daily dosing versus placebo in the acute treatment of major depression. J. Psychiatric Res. 36, 383e390. Detke, M.J., Wiltse, C.G., Mallinckrodt, C.H., McNamara, R.K., Demitrack, M.A., 2004. Duloxetine in the acute and long-term treatment of major depressive disorder: a placebo- and paroxetine-controlled trial. Eur. Neuropsychopharmacol. 14, 457e470. Devor, M., Wall, P.D., 1978. Reorganization of spinal cord sensory map after peripheral nerve injury. Nature 276, 75e76. Esser, M.J., Sawynok, J., 1999. Acute amitriptyline in a rat model of neuropathic pain: differential symptoms and route effects. Pain 80, 643e653. Fields, H.L., Basbaum, A.I., 1994. Central nervous system mechanisms of pain modulation. In: Wall, P.D., Melzack, R. (Eds.), Textbook of Pain. Churchill Livingstone, Edinburgh, pp. 243e257. Fuller, R.W., Hemrick-Luecke, S.K., Snoddy, H.D., 1994. Effects of duloxetine, an antidepressant drug candidate, on concentrations of monoamines and their metabolites in rats and mice. J. Pharmacol. Exp. Ther. 269, 132e136.

1180

C.K. Jones et al. / Neuropharmacology 51 (2006) 1172e1180

Gehlert, D.R., Hemrick-Luecke, S.K., Schober, D.A., Krushinski, J., Howbert, J.J., Robertson, D.W., Wong, D.T., Fuller, R.W., 1995. (R)thionisoxetine, a potent and selective inhibitor of central and peripheral norepinephrine uptake. Life Sci. 56, 1915e1920. Goldenberg, D., Mayskiy, M., Mossey, C., Ruthazer, R., Schmid, C., 1996. A randomized, double-blind crossover trial of fluoxetine and amitriptyline in the treatment of fibromyalgia. Arthritis Rheum. 39, 1852e1859. Goldstein, D.J., Lu, Y., Detke, M.J., Lee, T.C., Iyengar, S., 2005. Duloxetine vs. placebo in patients with painful diabetic neuropathy. Pain 116, 109e118. Iyengar, S., Webster, A.A., Hemrick-Luecke, S.K., Xu, J.Y., Simmons, R.M., 2004. Efficacy of duloxetine, a potent and balanced serotoninnorepinephrine reuptake inhibitor in persistent pain models in rats. J. Pharmacol. Exp. Ther. 311, 576e584. Jasmin, L., Tien, D., Janni, G., Ohara, P.T., 2003. Is noradrenaline a significant factor in the analgesic effect of antidepressants? Pain 106, 3e8. Jones, S.L., 1991. Descending noradrenergic influences on pain. Prog. Brain Res. 88, 381e394. Jones, C.K., Peters, S.C., Shannon, H.E., 2005. Duloxetine, a selective serotonergic and noradrenergic reuptake inhibitor, is efficacious in persistent and inflammatory, but not acute nociceptive, pain states in rodents. J. Pharmacol. Exp. Ther. 312, 726e732. Kishore-Kumar, R., Max, M.B., Schafer, S.C., Gaughan, A.M., Smoller, B., Gracely, R.H., Dubner, R., 1990. Desipramine relieves postherpetic neuralgia. Clin. Pharmacol. Ther. 47, 305e312. Koch, S., Hemrick-Luecke, S.K., Thompson, L.K., Evans, D.C., Threlkeld, P.G., Nelson, D.L., Perry, K.W., Bymaster, F.P., 2003. Comparison of effects of dual transporter inhibitors on monoamine transporters and extracellular levels in rats. Neuropharmacology 45, 935e944. Lynch, M.E., 2001. Antidepressants as analgesics: a review of randomized controlled trials. J. Psychiatry Neurosci. 26, 30e36. Max, M.B., 1995. Antidepressant Drugs as treatments for chronic pain: efficacy and mechanisms. In: Bromm, B., Desmedt, J.E. (Eds.), Pain and the Brain: From Nociception to Cognition. Advances in Pain Research and Therapy, Vol. 22. Raven Press, New York, pp. 501e514. McQuay, H.J., Tramer, M., Nye, B.A., Carroll, D., Wiffen, P.J., Moore, R.A., 1996. A systematic review of antidepressants in neuropathic pain. Pain 68, 217e227.

Millan, M.J., 2002. Descending control of pain. Prog. Neurobiol. 66, 355e 474. Onghena, P., Van Houdenhove, B., 1992. Antidepressant-induced analgesia in chronic non-malignant pain: a meta-analysis of 39 placebo-controlled studies. Pain 49, 205e219. Owens, M.J., Morgan, W.N., Plott, S.J., Nemeroff, C.B., 1997. Neurotransmitter receptor and transporter binding profile of antidepressants and their metabolites. J. Pharmacol. Exp. Ther. 283, 1305e1322. Rowbotham, M.C., Goli, V., Kunz, N.R., Lei, D., 2004. Venlafaxine extended release in the treatment of painful diabetic neuropathy: a double-blind, placebo-controlled study. Pain 110, 697e706. Sawynok, J., Esser, M.J., Reid, A.R., 1999. Peripheral antinociceptive actions of desipramine and fluoxetine in an inflammatory and neuropathic pain test in rats. Pain 82, 149e158. Sindrup, S.H., Jensen, T.S., 1999. Efficacy of pharmacological treatments of neuropathic pain: an update and effect related to mechanism of drug action. Pain 83, 389e400. Sindrup, S.H., Bjerre, U., Dejgaard, A., Brosen, K., Eshoj, O., Mogensen, E.F., 1990. The selective serotonin reuptake inhibitor paroxetine is effective in the treatment of diabetic neuropathy symptoms. Pain 42, 135e144. Traub, R.J., 1997. Spinal modulation of the induction of central sensitization. Brain Res. 778, 34e42. Wong, D.T., Robertson, D.W., Bymaster, F.P., Krushinski, J.H., Reid, L.R., 1988. LY227942, an inhibitor of serotonin and norepinephrine uptake: biochemical pharmacology of a potential antidepressant drug. Life Sci. 43, 2049e2057. Wong, D.T., Bymaster, F.P., Mayle, D.A., Reid, L.R., Krushinski, J.H., Robertson, D.W., 1993. LY248686, a new inhibitor of serotonin and norepinephrine uptake. Neuropsychopharmacology 8, 23e33. Woolf, C.J., Doubell, T.P., 1994. The pathophysiology of chronic painincreased sensitivity to low threshold AB-fiber inputs. Curr. Opin. Neurobiol. 4, 525e534. Woolf, C.J., Shortland, P., Sivilotti, L.G., 1994. Sensitization of high mechanothreshold superficial dorsal horn and flexor motor neurones following chemosensitive primary afferent activation. Pain 58, 141e155.