Direct Evidence for the Ongoing Brain Activation by Enhanced Dynorphinergic System in the Spinal Cord under Inflammatory Noxious Stimuli

• ❖ Spinal dynorphin release contributes to hypersensitivity to stimuli in rodents, but whether it activates supraspinal structures related to pain is not known

• ❖ Injection of complete Freund's adjuvant into the paw of mice increased dynorphin in the spinal cord and increased activity in several cortical regions associated with pain processing

• ❖ These effects were not present in genetically altered mice lacking dynorphin

ENDOGENOUS opioid peptides are efficient analgesics that bind to opioid receptors. These peptides are typically produced to counteract chronic pain. They are classified as enkephalins, endorphins, and dynorphins.1Dynorphins are neuropeptides that inhibit neuronal activity through κ-opioid receptors.2Dynorphin A (1–17) is one of the major proteolytic fragments of prodynorphin2and has been shown to be distributed widely throughout the central nervous system.2–5Many experimental models of pathologic pain, including inflammatory pain,6–10neuropathic pain,10–12bone cancer pain,13and abnormal pain (hyperalgesia),14show a significant induction of dynorphin A in the spinal cord. Relatively low doses of dynorphin A produce analgesia by virtue of acting as an inhibitory opioid peptide and preferentially activating κ-opioid receptors.10,15,16In contrast, high doses of dynorphin A elicit pronociceptive behaviors, such as biting, licking, and scratching,17or allodynia.3,18,19In such cases, the actions are not sensitive to opioid antagonist but are blocked by intrathecal pretreatment with MK-801, a noncompetitive antagonist of N -methyl-d-aspartic acid (NMDA) receptor channels.19Taken together, increased dynorphin A in the spinal cord may play an important role in the nociceptive state in terms of its paradoxical effects on neurotransmission.

Functional magnetic resonance imaging can be used to investigate spatial and temporal brain activation. Functional magnetic resonance imaging has been used to evaluate pain perception in the central nervous system in healthy humans and in those with various kinds of pain.20,21Noxious heat stimulation in humans or repetitive heat stimulation through peltier elements in animals activates several brain regions, so-called pain matrix.22–25Recent studies have demonstrated that neuroimaging in humans and animals can detect changes in regional activity initiated by the administration of drugs that induce or modulate pain, such as morphine, ketamine, formalin, and capsaicin.26–29Furthermore, these studies have introduced the new term pharmacological functional magnetic resonance imaging (phMRI) for this technique, which promises to become an important new tool for the researcher who is interested in mapping and understanding the pain mechanism. We previously demonstrated that neuropathic pain-like transmission evoked by the spinal activation of protein kinase C caused a significant increase in the activity of several brain regions using a mouse phMRI assay.30 

In this study, we investigated whether a deletion of the gene that encodes the prodynorphin could affect the brain regions activated after unilateral intraplantar injection of complete Freund's adjuvant (CFA) injection using phMRI method in mice and whether direct intrathecal administration of dynorphin A (1–17) at relatively high doses could facilitate brain activity in regions that are associated with pain perception via  an NMDA receptor-mediated pathway.

This study was conducted in accordance with the Guiding Principles for the Care and Use of Laboratory Animals, Hoshi University, as adopted by the Committee on Animal Research of Hoshi University, which is accredited by the Ministry of Education, Culture, Sports, Science and Technology of Japan. This study was approved by the Animal Research Committee of Hoshi University. The first series of experiments using 41 mice, either wild-type (C57BL/6 and 129S4/SvJ mixed genetic background, 21 males, 8–10 weeks old, 18–23 g; The Jackson Laboratory, Bar Harbor, ME) or prodynorphin gene-knockout (C57BL/6 and 129/SvJ mixed genetic backgrounds, 20 males, 8–10 weeks old, 18–23 g; The Jackson Laboratory) mice was also performed. Ninety-seven male mice (8–10 weeks old, 18–23 g) were used for another series of experiments in C57BL/6J mice (CLEA Japan, Inc., Tokyo, Japan). Animals were kept in a room with an ambient temperature of 23°± 1° C and a 12-h light-dark cycle (lights on 8:00 am to 8:00 pm). Food and water were available ad libitum . All animals were individually housed, and all behavioral studies were performed during the light period. At the end of the experiments, animals were humanely killed by a rising concentration of ethyl ether.

Intrathecal injection was performed as described by Hylden and Wilcox31using a 25-μl Hamilton syringe with a 30 ½-gauge needle. The needle was inserted into the intervertebral space between L5 and L6 level of the spinal cord. A reflexive flick of the tail was considered to be an accuracy of each injection. The injection volume was 4 μ l for intrathecal injection.

A persistent inflammatory pain model was produced by unilateral intraplantar injection of CFA (mycobacterium tuberculosis; Sigma-Aldrich Co., St. Louis, MO) in a volume of 50 μ l into the plantar surface of the right hind paw (ipsilateral side) of mice during anesthesia with isoflurane.32–34Control mice were given saline in a volume of 50 μ l into the plantar surface of the right hind paw.

To observe the pain-like behaviors induced by the intrathecal injection of dynorphin A (1–17) in normal mice, groups of mice were individually placed in an observation cage immediately after intrathecal injection of dynorphin A (1–17) (0.3, 3, and 30 pmol/mouse; Peptide Institute, Inc., Osaka, Japan) or saline. Intrathecal pretreatment with MK-801 (0.3 or 1 nmol/mouse) was performed 30 min before the intrathecal injection of dynorphin A (1–17) (30 pmol/mouse).

To assess the pain-like behaviors induced by inflammatory pain, groups of mice were individually placed in an observation cage immediately after the intraplantar injection of CFA or saline. Intrathecal pretreatment with antiserum against dynorphin A (1–17) (1:100; Peninsula Laboratories, Inc., San Carlos, CA) or control serum (normal rabbit serum; Vector Laboratories, Inc., Burlingame, CA) was performed 30 min before the intraplantar injection of CFA.

The number of licking or flinching behaviors was counted, and the duration of licking and flinching behaviors was measured for 20 min after treatment.

Functional imaging was performed, as previously described.30To investigate the effect of the intraplantar injection of CFA, prodynorphin gene-knockout mice or wild-type mice were anesthetized using isoflurane immediately after the intraplantar injection of CFA or vehicle. To investigate the effect of a single intrathecal treatment with dynorphin A (1–17) in C57BL/6J mice, mice were anesthetized using isoflurane immediately after intrathecal injection of dynorphin A (1–17) (30 pmol/mouse). Animals were then transferred to a cradle that was designed to fit inside the probe of the magnetic resonance system and supplied with 1% isoflurane via  a fitted mask. A continuous phMRI scanning protocol was used to study the changes in brain signal intensity using T2 star-weighted blood oxygenation level-dependent (BOLD) contrast. BOLD responses were measured hourly from 30 min to 6 h at all brain levels.

Experiments were performed with a Unity Inova spectrometer (Varian, Palo Alto, CA) that was interfaced to a 9.4-T/31-cm horizontal bore magnet equipped with actively shielded gradients capable of 300 mT/m in a rise time of 500 s (Magnex Scientific, Abingdon, United Kingdom). High-resolution anatomical scans were collected using a fast spin echo pulse sequence (repetition time = 2000 ms, echo time = 45 ms, field of view = 25.6 × 25.6 mm², 1. 0-mm slice thickness, 256 × 256 data matrix).

Functional images were obtained with the two-slice gradient-echo fast imaging sequence (echo time = 25 ms, repetition time = 70 ms, 30-° flip angle, 128 × 128).35One-millimeter-thick slices were simultaneously acquired over a field of view of 25.6 mm², number of average of 2, and an acquisition time of 32 s. Typically, 3–6 images were collected at baseline, followed by intrathecal injection of dynorphin A (1–17) or intraplanter injection of CFA. Intrathecal pretreatment with MK-801 (1 nmol/mouse) was performed 30 min before the intrathecal injection of dynorphin A (1–17). The regions of interest were selected, and statistical analyses were performed using the image-analysis software ImageJ (National Institutes of Health, Bethesda, MD). The regions of interest were drawn according to an atlas of the mouse brain.36BOLD signal intensity values in each regions of interest were extracted and normalized to the time of baseline (expressed as a percent change from baseline). Statistical analysis was performed to compare percent changes in BOLD signal intensity and activated pixels between baseline and each time point after CFA and dynorphin A (1–17) injection.

Total RNA obtained from the spinal cord of mice was extracted using the SV Total RNA Isolation System (Promega Co., Madison, WI). The lumbar spinal cord was quickly removed after mice were decapitated and homogenized in ice-cold lysis buffer containing β-mercaptoethanol following the manufacturer's instructions. First-strand complementary DNA was prepared as described,37and the prodynorphin gene was amplified in 50 μ l of a PCR solution containing MgCl², dNTP mix, and DNA polymerase with either synthesized primers (prodynorphin: 5′-GTG CAG TGA GGA TTC AGG ATG GG-3′[sense] and 5′-GAG CTT GGC TAG TGC ACT GTA GC-3′[antisense], glyceraldehyde-3-phosphate dehydrogenase: 5′-CCC ACG GCA AGT TCA ACG G-3′[sense] and 5′-CTT TCC AGA GGG GCC ATC CA-3′[antisense]). Samples were heated to 94° C for 2 min, 55° C for 2 min, and 72° C for 3 min and cycled 29 times through 94° C for 1 min, 55° C for 2 min, and 72° C for 3 min. The final incubation was at 72° C for 7 min. The mixture was subjected to 1% agarose gel for electrophoresis with the indicated markers and primers for the internal standard (glyceraldehyde-3-phosphate dehydrogenase). Each sample was applied to more than two lanes in the same gel. The agarose gel was stained with ethidium bromide and photographed with ultraviolet transillumination. The intensity of the bands was analyzed and quantified by computer-assisted densitometry using ImageJ (free download software developed by National Institutes of Health). For the control, the different intensities of each band obtained from mice treated with saline were analyzed, and the average intensity was calculated. Each control intensity was then compared again with the average intensity to calculate the standard error. Under these conditions, the intensities of bands for samples obtained from CFA-treated mice were analyzed and compared with the average intensity for mice treated with saline. Finally, the percent of control with standard error for each sample was quantified.

Fast SYBR Green Master Mix (2 ×; Applied Biosystems, Inc., Foster City, CA) was used as the basis for the reaction mixture in the real-time PCR assay. Each gene prepared by the above procedure was amplified in 20 μ l of a PCR solution containing 10 μ l of the Fast SYBR Green Master Mix (2 ×) with synthesized primers for PDYN (sense: 5′-TTT GGC AAC GGA AAA GAA TC-3′, antisense: 5′-CAT AGC GTT TGG CCT GTT TT-3′) or β-actin (sense: 5′-CAG CTT CTT TGC AGC TCC TT-3′, antisense: 5′-TCA CCC ACA TAG GAG TCC TT-3′). In addition to each sample, each test run included a no-target control that contained reaction mixture and PCR-grade water. PCR with a StepOnePlus (Applied Biosystems, Inc., Foster City, CA) was performed with the following cycling conditions: 95° C for 20 s, followed by cycled 45 cycles of 95° C for 3 s and 60° C for 30 s. Fluorescence detection was conducted after each extension step.

Sample preparation and immunohistochemistry were performed following the methods previously described.38Six hours after intraplantar injection, mice were deeply anesthetized with isoflurane and intracardially perfusion fixed with freshly prepared 4% paraformaldehyde in 0.1 m phosphate buffer saline (PBS), pH 7.4. After perfusion, the lumbar spinal cord was quickly removed and postfixed in 4% paraformaldehyde for 2 h and then permeated with 20% sucrose in 0.1 m PBS for 1 day and 30% sucrose in 0.1 m PBS for 2 days with agitation. The L5 lumbar spinal cord segments were then frozen in an embedding compound (Sakura Finetechnical, Tokyo, Japan) on isopentane using liquid nitrogen and stored at −30° C until use. Frozen spinal cord segments were cut with a freezing cryostat (Leica CM 1510; Leica, Wetzlar, Germany) at a thickness of 10 μ m and thaw mounted on poly-l-lysine-coated glass slides.

The spinal cord sections were blocked in 20% normal goat serum with 0.1% Triton X in 0.01 m PBS for 1 h at room temperature. The primary antibody [1:600 dynorphin A (1–17) (Phoenix Pharmaceuticals, Inc., Belmont, CA)] was diluted in 0.01 m PBS containing 20% normal goat serum with 0.1% Triton X and incubated for two nights at 4° C. The samples were then rinsed and incubated with an appropriate secondary antibody conjugated with Alexa 546 for 2 h at room temperature. Because the staining intensity might vary between experiments, control sections were included in each run of staining. The slides were then coverslipped with PermaFluor aqueous mounting medium (Immunon; Thermo Electron, Pittsburgh, PA). All sections were observed with a fluorescence microscope (Olympus BX-80; Olympus, Tokyo, Japan) and photographed with a digital camera (CoolSNAP HQ; Olympus).

Dat a are expressed as the mean with SEM. One-and two-way ANOVAs with independent and repeated measures, as well as planned comparisons or Student t  tests, were used as appropriate for the experimental design. Multiple comparisons were performed using Dunnett or Bonferroni post hoc  test, where appropriate. All statistical analyses were performed with Prism version 5.0a (GraphPad Software, Inc., San Diego, CA).

Genotyping of the offspring from prodynorphin knockout (−/−) mice was confirmed by PCR analysis using DNA extracted from the ear (data not shown). W e investigated whether lack of the prodynorphin gene could influence inflammatory pain using these genotype mice. In wild-type mice, CFA injection caused significant flinching or licking behaviors (fig. 1). These behaviors started just after CFA injection and lasted for more than 20 min. However, the numbers and durations of these behaviors were clearly decreased in prodynorphin knockout mice with CFA injection (fig. 1). Two-way ANOVA showed a significant interaction between genotype and treatment (flinching, F  (1,14) = 4.736, P = 0.0471; licking, F  (1,14) = 6.699, P = 0.0215; duration, F  (1,14) = 31.72, P < 0.0001), a significant effect of treatment (flinching, F  (1,14) = 19.39, P = 0.0006; licking, F  (1,14) = 17.53, P = 0.0009; duration, F  (1,14) = 59.56, P < 0.0001), and a significant effect of genotype (flinching, F  (1,14) = 5.291, P = 0.0373; licking, F  (1,14) = 10.94, P = 0.0053; duration, F  (1,14) = 43.78, P < 0.0001). Post hoc  comparison indicated a significant difference between the saline in wild-type group and the CFA in wild-type group (flinching response: F  (3,12) = 24.87, P < 0.0001; licking response: F  (3,12) = 70.84, P < 0.0001; total duration of responses: F  (3,12) = 58.98; P < 0.0001) and a significant difference between CFA in wild-type group and CFA in prodynorphin knockout group (flinching response: F  (3,12) = 24.87, P = 0.0019; licking response: F  (3,12) = 70.84, P < 0.0001; total duration of responses: F  (3,12) = 58.98, P < 0.0001).

Next, we investigated the changes in BOLD signal intensity in two brain regions after intraplantar injection of CFA. As shown in figures 2A–C, injection of CFA caused positive signal activity in the cingulate cortex (A-i), somatosensory cortex (B-i), and insular cortex (C-i) of wild-type mice compared with prodynorphin knockout mice. In the cingulate cortex, injection of CFA produced a n increase in signal intensity (fig. 2A-ii). In the somatosensory cortex of wild-type mice, injection of CFA also produced an increase in signal intensity after injection (fig. 2B-ii). In the insular cortex of wild-type mice, similar to somatosensory area, injection of CFA produced a n increase in signal intensity after injection (fig. 2C-ii). These effects were not seen in mice that lacked the prodynorphin gene. Statistical analyses were done with two-way ANOVA followed by Bonferroni test (cingulate cortex: interaction between genotype and time: F  (7,48) = 1.138, P = 0.356; effect of genotype, F  (1,48) = 14.28, P = 0.0004; effect of time, F  (7,48) = 1.348, P = 0.2491; somatosensory cortex: interaction between genotype and time: F  (7,48) = 0.996, P = 0.446; effect of genotype, F  (1,48) = 27.77, P < 0.0001; effect of time, F  (7,48) = 0.984, P = 0.4543; insular cortex: interaction between genotype and time: F  (7,48) = 4.799, P = 0.0004; effect of genotype, F  (1,48) = 151.3, P < 0.0001; effect of time, F  (7,48) = 2.933, P = 0.0122) (figs. 2A-ii–C-ii). As shown in figures 2D and E, injection of CFA caused positive signal activity in the medial thalamic nuclei (D-i) and the lateral thalamic nuclei (E-i) of wild-type mice compared with prodynorphin knockout mice. The medial thalamic nuclei and lateral thalamic nuclei were also activated by CFA injection in wild-type mice but not in prodynorphin knockout mice. Statistical analyses were done with two-way ANOVA followed by Bonferroni test (medial thalamic nuclei: interaction between genotype and time: F  (7,48) = 0.730, P = 0.6475; effect of genotype, F  (1,48) = 14.62, P = 0.0004; effect of time, F  (7,48) = 0.856, P = 0.5476; lateral thalamic nuclei: interaction between genotype and time: F  (7,48) = 0.501, P = 0.8289; effect of genotype, F  (1,48) = 11.72, P = 0.0013; effect of time, F  (7,48) = 0.563, P = 0.7821) (figs. 2D-ii and E-ii). Under these conditions, control mice of both genotypes that had been injected with saline failed to show activation in any brain regions (data not shown).

To confirm the possible involvement of spinal dynorphin A (1–17) in the pain-related response induced by CFA injection, we performed intrathecal injection of antiserum against dynorphin A (1–17) and evaluated its effect on CFA-mediated pain-related responses. Dynorphin A (1–17) antiserum significantly decreased the number of and shortened the duration of flinching or licking responses compared with those in control serum-treated mice (fig. 3). The two-way ANOVA indicated a significant interaction between pretreatment and posttreatment (flinching, F  (1,23) = 48.89, P < 0.0001; licking, F  (1,23) = 14.83, P = 0.0008; duration, F  (1,23) = 14.19, P = 0.001), a significant effect of pretreatment (flinching, F  (1,23) = 122.8, P < 0.0001; licking, F  (1,23) = 42.02, P < 0.0001; duration, F  (1,23) = 22.43, P < 0.0001), and a significant effect of posttreatment (flinching, F  (1,23) = 56.6, P < 0.0001; licking, F  (1,23) = 12.03, P = 0.0021; duration, F  (1,23) = 8.825, P = 0.0068). Post hoc  comparison indicated a significant difference between the control serum-saline group and control serum-CFA group (flinching response: F  (3,23) = 76.95, P < 0.0001; licking response: F  (3,23) = 23.04, P < 0.0001; total duration of responses: F  (3,23) = 15.38; P < 0.0001) and a significant difference between control serum-CFA group and dynorphin A (1–17) antiserum-CFA group (flinching response: F  (3,23) = 76.95, P < 0.0001; licking response: F  (3,23) = 23.04, P < 0.0001; total duration of responses: F  (3,23) = 23.04, P = 0.00025).

In the reverse transcription–PCR assay, the expression of prodynorphin messenger RNA (mRNA) on the ipsilateral side of the spinal cord obtained from CFA-treated mice was significantly increased at 1, 3, and 6 h after CFA injection compared with that in saline-treated mice (1 h: P = 0.014, 3 h: P = 0.0002, 6 h: P = 0.0008 vs.  saline-treated group; figs. 4A and B). In the contralateral spinal cord obtained from CFA-treated mice, there were no significant differences in the expression level of prodynorphin mRNA compared with those on the contralateral side of saline-treated mice (data not shown). In agreement with the results of reverse transcription–PCR, we confirmed that the expression of PDYN mRNA was significantly increased in the spinal cord of mice with CFA treatment compared with that in saline-treated mice at 6 h after CFA injection using real-time PCR (P = 0.042 vs.  saline-treated group; fig. 4C).

To investigate a possible change in the level of dynorphin A (1–17) in the superficial dorsal horn of CFA-treated mice, immunohistochemical studies were performed. Dynorphin A (1–17)-like immunoreactivity was detected on the superficial laminae of the ipsilateral side of the L5 lumbar spinal dorsal horn with saline treatment (fig. 4D-i). Six hours after CFA injection, dynorphin A (1–17)-like immunoreactivity on the superficial laminae of the ipsilateral side of the L5 lumbar spinal dorsal horn was markedly increased compared with that with saline injection (fig. 4D-ii).

Next, we observed pain-like behaviors induced by intrathecal dynorphin A (1–17) in mice. As shown in figure 5, intrathecal injection of dynorphin A (1–17) produced a significant increase in both the number of biting or licking responses and the duration of these responses in a dose-dependent manner. Statistical analyses were done with one-way ANOVA followed by Dunnett test (biting or licking responses: F  (3,17) = 6.084, 0.3 pmol: P = 0.6584, 3 pmol: P = 0.039, 30 pmol: P = 0.0033; total duration of responses: F  (3,17) = 10.94, 0.3 pmol: P = 0.2108, 3 pmol: P = 0.0018, 30 pmol: P = 0.0002; vs.  intrathecal vehicle group; figs. 5A and B). These responses started a pproximately 4 or 5 min after intrathecal injection and lasted for more than 20 min. The control group-injected intrathecal with vehicle did not exhibit these behaviors.

We also investigated the changes in BOLD signal intensity in different brain regions after the direct injection of dynorphin A (1–17) in the spinal cord of mice. Intrathecal injection of 30-pmol dynorphin A (1–17), which induced maxim u m pain-related responses in the behavior study, caused a significant increase in BOLD signal intensity in the cingulate cortex, somatosensory cortex, and insular cortex of mice compared with the basal activity. Although each of the signal intensities was increased compared with the basal intensity, there were slight differences in the time course of the increase between these regions. The cingulate cortex was significantly activated after intrathecal injection of dynorphin A (1–17) (fig. 6A). In the somatosensory cortex, intrathecal injection of dynorphin A (1–17) produced a significant increase in signal intensity (fig. 6B). In the insular cortex, the activation appeared after intrathecal injection of dynorphin A (1–17) (fig. 6C). In the medial and lateral thalamic nuclei, both BOLD signal intensities were significantly changed after intrathecal injection of dynorphin A (1–17) (figs. 6D and E). These activations reverted to their basal levels at 6 h after dynorphin A (1–17) injection. Control mice that had been injected intrathecal with saline did not show such activation in these regions (fig. 6). Statistical analys e s were done with two-way ANOVA followed by Bonferroni test (cingulate cortex: interaction between treatme n t and time: F  (7,64) = 0.965, P = 0.4642; effect of treatment, F  (1,64) = 19.70, P < 0.0001; effect of time, F  (7,64) = 0.938, P = 0.4837; somatosensory cortex: interaction between treatment and time: F  (7,64) = 5.413, P < 0.0001; effect of treatment, F  (1,64) = 19.85, P < 0.0001; effect of time, F  (7,64) = 6.933, P < 0.0001; insular cortex: interaction between treatment and time: F  (7,64) = 2.474, P < 0.026; effect of treatment, F  (1,64) = 15.79, P = 0.0002; effect of time, F  (7,64) = 2.977, P = 0.0091; medial thalamic nuclei: interaction between treatment and time: F  (7,64) = 1.655, P = 0.1362; effect of treatment, F  (1,64) = 28.97, P < 0.0001; effect of time, F  (7,64) = 2.242, P = 0.0419; lateral thalamic nuclei: interaction between treatment and time: F  (7,64) = 1.500, P = 0.1834; effect of treatment, F  (1,64) = 12.78, P = 0.0007; effect of time, F  (7,64) = 1.581, P = 0.1571).

To determine the possible involvement of spinal NMDA receptors in dynorphin A (1–17)-induced pain-like behaviors, we observed the influence of intrathecal pretreatment with the NMDA receptor antagonist, MK-801. As shown in figure 7, MK-801 pretreatment produced a dose-dependent inhibition of pain-like responses that could be caused by after intrathecal dynorphin A (1–17) injection. Statistical analys e s were done with one-way ANOVA followed by Dunnett test (biting or licking responses: F  (2,14) = 15.65, 0.3 nmol: P = 0.0022, 1 nmol: P = 0.0002; total duration of responses: F  (2,14) = 13.28; 0.3 nmol: P = 0.0052, 1 nmol: P = 0.0003; vs.  vehicle–dynorphin A (1–17)-treated group). Under these conditions, control mice of both, that is, pretreatment with saline and MK-801 that had been injected with saline, failed to show these behaviors (data not shown).

Because the cingulate cortex, somatosensory cortex, insular cortex, and the medial and lateral thalamic nuclei were all activated by intrathecal injection of dynorphin A (1–17) in a time-dependent manner, we further investigated whether the activation of these brain regions could be changed by pretreatment with MK-801. After intrathecal injection of MK-801 (1 nmol) and before dynorphin A (1–17) administration, we compared the changes in the increased levels of signal intensity in brain regions after intrathecal injection of dynorphin A (1–17), when the most significant increase in BOLD signal intensity was observed. As for the cortex regions, the increased levels of BOLD intensity at 2 h (in the cingulate cortex) or 1 h (in somatosensory cortex and the insular cortex) induced by intrathecal dynorphin A (1–17) were significantly suppressed by pretreatment with MK-801 (figs. 8A–C). In the medial and lateral thalamic nuclei, increased levels of BOLD intensity induced by dynorphin A (1–17) were also blocked by pretreatment with MK-801 (figs. 8D and E). Two-way ANOVA showed a significant interaction between phase and pretreatment (cingulate cortex, F  (1,16) = 6.517, P = 0.0213; somatosensory cortex, F  (1,16) = 13.06, P = 0.0023; insular cortex, F  (1,16) = 26.31, P = 0.0001; medial thalamic nuclei, F  (1,16) = 9.884, P = 0.0063; lateral thalamic nuclei, F  (1,16) = 8.402, P = 0.0105), a significant effect of phase (cingulate cortex, F  (1,16) = 10.17, P = 0.0057; somatosensory cortex, F  (1,16) = 29.52, P < 0.0001; insular cortex, F  (1,16) = 34.81, P < 0.0001; medial thalamic nuclei, F  (1,16) = 15.30, P = 0.0012; lateral thalamic nuclei, F  (1,16) = 6.496, P = 0.0215), and a significant effect of pretreatment (cingulate cortex, F  (1,16) = 6.427, P = 0.0221; somatosensory cortex, F  (1,16) = 13.35, P = 0.0021; insular cortex, F  (1,16) = 25.20, P = 0.0001; medial thalamic nuclei, F  (1,16) = 11.64, P = 0.0036; lateral thalamic nuclei, F  (1,16) = 9.030, P = 0.0084). Post hoc  comparison indicated a significant difference between the basal intensity of saline-pretreated groups versus  dynorphin challenge in saline-pretreated groups (cingulate cortex, F  (3,16) = 7.703, P = 0.0048; somatosensory cortex, F  (3,16) = 18.64, P < 0.0001; insular cortex, F  (3,16) = 28.77, P < 0.0001; medial thalamic nuclei, F  (3,16) = 12.27, P = 0.0008; lateral thalamic nuclei, F  (3,16) = 7.976, P = 0.0085) and a significant difference between the dynorphin challenge in saline-pretreated groups versus  dynorphin challenge in MK-801-pretreated groups (cingulate cortex, F  (3,16) = 7.703, P = 0.0144; somatosensory cortex, F  (3,16) = 18.64, P < 0.0001; insular cortex, F  (3,16) = 28.77, P < 0.0001; medial thalamic nuclei, F  (3,16) = 12.27, P = 0.00165; lateral thalamic nuclei, F  (3,16) = 7.976, P = 0.0043).

Many studies have confirmed that endogenous opioid neuropeptides are associated with nociceptive transmission. In addition to endogenous opioid peptides, dynorphin A has been implicated in both pronociceptive and antinociceptive actions and is widely expressed throughout the central nervous system.3Several studies have provided evidence that the nociceptive action of spinal dynorphin plays a role in the long-term maintenance of chronic pain, such as inflammatory pain8,21and neuropathic pain.39However, little, if any, is known about the role of dynorphin A in the acute phase of inflammatory pain. Therefore, in this study, we first investigated whether dynorphin A could be involved in pain-associated behaviors and brain activity in the acute phase of the peripheral inflammatory pain-like state using prodynorphin gene knockout mice and their wild-type mice.

Prodynorphin gene knockout mice were indistinguishable from wild-type mice in terms of general behavior and development. Intraplantar injection of CFA increased the duration and number of spontaneous pain-associated behaviors such as flinching or licking in wild-type mice but not in prodynorphin gene knockout mice. Under the present condition, phMRI showed that intraplantar injection of CFA caused robust positive signal activity in the cingulate cortex, somatosensory cortex, insular cortex, and medial and lateral thalamic nuclei of wild-type mice. In contrast, these effects were diminished in prodynorphin knockout mice. Because cortical areas are activated by the receipt of noxious information from the spinothalamic tract, many neuroimaging studies have shown activity by demonstrating brain circuitry.40–42These cortical representations of pain are called the pain matrix, including the somatosensory cortex, insular cortex, cingulate cortex, and prefrontal cortex.43Among these areas, the cingulate cortex area is an affective-motivational component of pain and mainly receives information from the medial system of the spinothalamic tract.44,45On the other hand, the somatosensory cortex area is a sensory-discriminative component of pain and mainly receives information from the lateral system of the spinothalamic tract. The insular cortex is basically considered to be not only in charge of the sensory-integrative component via  the lateral system but also involved in limbic integration by virtue of the medial system.43The medial and lateral thalamic nuclei are also categorized as centers for pain perception, which relay sensory information to those cortical areas. Taken together, the present findings suggest that dynorphin A regulates spontaneous pain-associated behaviors and may play a crucial role in the activation of these pain-related brain regions in the acute phase of inflammatory pain.

In the spinal cord, dynorphin A is principally localized in laminae 1 and 2, and in an inflammatory state it spreads to laminae 3 and 4 and deeper laminae.20,46Furthermore, it has been reported that spinal dynorphin A immunoreactivity is increased in CFA-injected mice, 20,46and the upregulation of spinal preprodynorphin mRNA is induced by peripheral inflammation.6,8,21,47,48It was reported that in rats, incisional surgery increased the spinal cord dynorphin expression but did not drive microglial prostaglandin production or mechanical hypersensitivity.49Although further investigation is required, these findings suggest that the involvement of spinal dynorphin in the different types of pain might be, in some cases, the species difference. In this study, the level of prodynorphin mRNA in the lumbar spinal cord of mice was increased at 1, 3, and 6 h after CFA injection. Furthermore, 6 h after intraplantar injection, CFA produced a marked increase in dynorphin A (1–17)-like immunoreactivity on the superficial layers of the ipsilateral side of the spinal cord. In this study, we demonstrated that CFA-induced pain-associated behaviors were suppressed by intrathecal injection of a specific antiserum for dynorphin A (1–17). These findings suggest that peripheral inflammation may release dynorphin A (1–17) with an increase in the expression of prodynorphin in the spinal cord, resulting in the induction of an early phase of an inflammatory pain-like state.

Prodynorphin transcription is regulated by several factors in a tissue-specific manner.50It has been proposed that dynorphin expression is determined by a balance between gene transactivation and repression.51In the spinal cord, the binding of transactivators including cyclic adenosine monophosphate response element-binding protein to cyclic adenosine monophosphate response element,52the binding of c-fos  with c-jun  to an AP-1 element in the dynorphin promoter,53and the derepression of DREAM54play important roles in prodynorphin gene transcription. Several studies have reported that dynorphin A can affect several ion channels or nociceptive receptors through nonopioidergic mechanisms.3,15,18,19,55In addition, those mechanisms involve the theory that dynorphin A may play a role in both antinociceptive and pronociceptive effects depending on its concentration: the former is due to endogenous κ-opioidergic action and the latter is due to glutaminergic actions through NMDA receptors.3At physiologically lower concentrations, dynorphin A reduces calcium influx via κ-opioid receptor activity, whereas at pathologically higher concentrations, dynorphin A causes increased intracellular calcium levels by activating NMDA receptors.15Although further experiment is required for the measurement of released dynorphin A in the present case after CFA injection, we propose here that the noxious stimuli after CFA injection into the hind paw may consistently release high concentrations of dynorphin A in the spinal cord, leading to excitatory neurotransmission.

It is well known that NMDA receptors initiate increases in the excitability of spinal neurons and cause the central sensitization of pain. Furthermore, a previous study suggested that NMDA receptors in the spinal cord play a major role in the increased BOLD signal in the somatosensory cortex by way of the spinothalamic pathway.56Therefore, a subsequent study was undertaken to investigate whether NMDA receptors could be involved in spontaneous pain-associated behaviors or changes in brain activation induced by the intrathecal injection of dynorphin A. A single intrathecal injection of dynorphin A (1–17) caused an increase in pain-associated behaviors, and this effect was suppressed by pretreatment with MK-801. Under these conditions, we found here for the first time that dynorphin A produced a significant increase in signal intensity in the cingulate cortex, somatosensory cortex, insular cortex, and medial and lateral thalamic nuclei at 0.5–2 h after injection compared with the basal intensity, and this effect was abolished by pretreatment with MK-801. If we consider these findings, the present data suggest that an increase in spinal dynorphin A (1–17) may activate the transmission of ascending nociceptive information in the central nervous system via  NMDA receptors. Although additional research is needed to understand the mechanism of the activation of dynorphin A in the early phase of inflammatory pain, it seems likely that increased spinal dynorphin A may contribute not only to signals of acute pain but also to induction of chronic pain via  spinal NMDA receptors.

In view of the clinical application of the present information, we propose that spinal dynorphin may be an unique biomarker for pain.

The present data constitute novel evidence that loss of the prodynorphin gene prevents the ascending transmission of nociceptive information from the dorsal horn of the spinal cord to brain areas after the intraplantar injection of CFA. Moreover, dynorphin A within the spinal cord is directly involved in the early phase of an inflammatory pain-like state through NMDA receptors. This study is the first to clarify the ongoing brain activation in the early phase of inflammatory pain associated with an endogenous dynorphinergic pathway in an animal model of pain by means of pharmacological neuroimaging technology.

The authors thank Kazuhiko Miyashita, B.S., and Atsuo Suzuki, B.S. (Graduate Students, Department of Toxicology, Hoshi University School of Pharmacy and Pharmaceutical Sciences, Shinagawa-ku, Tokyo, Japan), for their expert technical assistance.