SEARCH

SEARCH BY CITATION

Keywords:

  • antisense oligonucleotide;
  • microglia;
  • p38;
  • pain;
  • spinal cord;
  • substance P

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Antagonist studies show that spinal p38 mitogen-activated protein kinase plays a crucial role in spinal sensitization. However, there are two p38 isoforms found in spinal cord and the relative contribution of these two to hyperalgesia is not known. Here we demonstrate that the isoforms are distinctly expressed in spinal dorsal horn: p38α in neurons and p38β in microglia. In lieu of isoform selective inhibitors, we examined the functional role of these two individual isoforms in nociception by using intrathecal isoform-specific antisense oligonucleotides to selectively block the expression of the respective isoform. In these rats, down-regulation of spinal p38β, but not p38α, prevented nocifensive flinching evoked by intraplantar injection of formalin and hyperalgesia induced by activation of spinal neurokinin-1 receptors through intrathecal injection of substance P. Both intraplantar formalin and intrathecal substance P produced an increase in spinal p38 phosphorylation and this phosphorylation (activation) was prevented when spinal p38β, but not p38α, was down-regulated. Thus, spinal p38β, probably in microglia, plays a significant role in spinal nociceptive processing and represents a potential target for pain therapy.

Abbreviations used
ACSF

artificial cerebrospinal fluid

AS

antisense oligonucleotides

IT

intrathecal

MS

missense oligonucleotide

NK-1

neurokinin-1

PGE

Prostaglandin E

SP

substance P

p38 is a member of the MAPK family which is known to regulate events associated with cellular stress (Shi and Gaestel 2002). To date, four different p38 isoforms, α, β, γ and δ, have been identified (Kumar et al. 2003). These isoforms have been found in peripheral tissues and p38α in particular has been associated with local inflammatory cascades (Hale et al. 1999; Kumar et al. 2003). The isoforms differ in their substrate preference, activation modes and response to inhibitors (Goedert et al. 1997; Vachon et al. 2002; Pramanik et al. 2003; Pratt et al. 2003; Kuma et al. 2004). In the mature CNS, only p38α and p38β are constitutively expressed (Jiang et al. 1996, 1997; Carboni et al. 1998; Hu et al. 1999; Lee et al. 2000) and it has been shown that, in mouse brain, both p38α and p38β are present in neurons, while p38β is also expressed in glial cells (Lee et al. 2000).

Accumulating evidence now suggests that p38 plays an important role in nociceptive processing. Pain states arising from tissue injury and inflammation are characterized by an enhanced response to subsequent afferent stimulation. This hyperalgesia arises in large part from a facilitated processing of noxious input at the spinal level. Thus, injury-evoked afferent input leads to spinal release of peptides [i.e. substance P (SP)] and excitatory amino acids (i.e. glutamate) that activate, through their respective receptors, signaling pathways that generate spinal sensitization. For example, spinal p38 can activate phospholipase A2 (Lin et al. 1993; Borsch-Haubold et al. 1997; Hiller and Sundler 1999; Zhu et al. 2001) that liberates arachidonic acid. Arachidonic acid is converted by spinal constitutive cyclooxygenase to prostaglandins which have been shown to facilitate dorsal horn activity (Nicol et al. 1992; Baba et al. 2001). An important element of the induction and maintenance of hyperalgesia appears to involve non-neuronal cells such as microglia and astrocytes (Watkins et al. 2001). Several lines of evidence suggest that activation of spinal p38 is an important component in this process. Thus, afferent input generated by tissue and nerve injury (Watkins et al. 1997; Kim et al. 2002; Jin et al. 2003; Milligan et al. 2003; Schafers et al. 2003; Tsuda et al. 2004) or by the direct activation of spinal neurokinin-1 (NK-1) (Svensson et al. 2003b) or NMDA receptors (Svensson et al. 2003a) leads to phosphorylation (activation) of p38 in spinal microglia. The hyperalgesia observed in these various behavioral models was prevented by spinal administration of p38 inhibitors, indicating a behavioral consequence of spinal p38 activation. An important issue that has not yet been addressed is the relative contribution of the two p38 isoforms to the induction and maintenance of hyperalgesia.

Structurally the MAPKs (ERK, JNK and p38) display significant homology, although the ATP binding site necessary for MAPK function is sufficiently distinct as to permit the synthesis of inhibitors selective for p38 (Fitzgerald et al. 2003). However, the structural similarities between the p38α and p38β isoforms makes it difficult to target either one selectively. Thus, the majority of currently available inhibitors display activity against both isoforms. Even those reported to be isoform selective do not exceed IC50 ratios of 15–30 (Ju et al. 2002; Li et al. 2004; Sweitzer et al. 2004a,b). In addition, viable p38 isoform-deficient mice are not available (Adams et al. 2000; Mudgett et al. 2000).

In order to define the role of spinal p38α/β isoforms in nociception, we undertook a series of experiments with several specific aims: (i) to examine the anatomical and cellular distribution and location of p38α/β isoforms in rat spinal cord; (ii) to verify the efficacy and specificity of intrathecal (IT) treatment with isoform-specific antisense oligonucleotides (AS) on spinal expression of these two isoforms; (iii) to determine which isoform contributes to the stimulus-induced spinal p38 phosphorylation; (iv) to examine the effect of down-regulation of individual spinal isoforms on nocifensive behaviors in two experimental models of hyperalgesia and (v) to study possible mechanisms of p38-mediated spinal sensitization.

All experiments were carried out according to protocols approved by the Institutional Animal Care Committee of University of California, San Diego.

Animals

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Male Holzman Sprague-Dawley rats (300–350 g) were housed individually in micro-isolator filter cages and maintained on a 12-h light/dark cycle with free access to food and water. To permit repeated bolus IT drug delivery, chronic lumbar IT injection catheters (single lumen PE-5, 8.5 cm in length, Intramedic and Clay Adams; Becton Dickson Co., Sparks, MD, USA) were implanted through a cisternal exposure under isoflurane anesthesia and externalized as described elsewhere (Hayes et al. 2003). To permit IT injection and dialysis of the lumbar IT space, rats were prepared with chronic triple lumen loop dialysis catheters advanced 8.5 cm through a cisternal incision to the lumbar enlargement under isoflurane anesthesia and externalized (Marsala et al. 1995; Koetzner et al. 2004). The IT portion of the dialysis probe consists of tubular 3-cm cellulose dialysis fibers (Filtral AN69HF; Cobe Laboratories, Lakewood, CO, USA) bent double and connected at their ends to 7 cm of two lumens of the triple lumen catheter (Spectranetics, Colorado Springs, CO, USA); the third lumen permits the delivery of IT drug without interrupting dialysis. Preparation of the triple lumen loop dialysis catheter is described elsewhere (Koetzner et al. 2004). Studies involving rats with chronic IT dialysis catheters or single lumen injection catheters were undertaken 4–5 days after surgery. For bolus IT injection, all agents were prepared to be delivered in 10 µL followed by 10 µL saline to flush the catheter. Rats were monitored daily and removed from the study if any neurological dysfunction was noted, if there was greater than 10% weight loss over 5 days or if the catheter was occluded. Less than 5% of the animals prepared were so excluded.

Drugs: preparation and administration

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

For bolus IT injection, all agents were prepared to be delivered in 10 µL followed by 10 µL saline to flush the catheter. SP (30 nmol; Sigma, St Louis, MO, USA) was dissolved in physiological saline and SB203580 (3–30 µg; EMD Biosciences Inc., La Jolla, CA, USA) was dissolved in 5% dimethylsulfoxide (Sigma) and 5% Cremophor EL (Sigma) in saline. IT injection of vehicles (5% dimethylsulfoxide and 5% Cremophor EL in physiological saline or saline alone) had no effect on nociceptive thresholds/behavior or protein expression/phosphorylation.

Antisense oligonucleotides

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Three oligonucleotides (a kind gift from ISIS Pharmaceuticals Inc., Carlsbad, CA, USA) with lengths of 20 nucleotides were employed in the study: ISIS 101757 (p38α AS) AGGTGCTCAGGACTCCATTT, beginning at position 1081 in the rat p38α mRNA; ISIS 107211 (p38β AS) GTATGTCCTCCTCGCGTGGA, beginning at position 439 in the rat p38β mRNA and ISIS 141923 [a universal missense oligonucleotide (MS) control] CCTTCCCTGAAGGTTCCTCC. These oligonucleotides were synthesized as 2′-methoxyethyl phosphorothioate chimeric oligonucleotides in which the first and last five bases contained 2′-methoxyethyl modifications with a uniform phosphorothioate backbone (Baker et al. 1997; Butler et al. 2002). The oligonucleotides were dissolved in artificial cerebrospinal fluid (ACSF) immediately before IT administration. The ACSF contained (in mm): Na+, 151.1; K+, 2.6; Mg2+, 0.9; Ca2+, 1.3; Cl, 122.7; HCO3, 21.0; HPO4, 2.5 and was bubbled with 95% O2/5% CO2 before use to adjust the final pH to 7.2. AS and MS (10–30 µg) were administrated in 10 µL ACSF daily for 5 days and behavioral experiments and spinal tissue collection performed on day six.

Western blot

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Before killing, rats were deeply anesthetized and, after decapitation, the spinal cords were ejected from the vertebral column by a saline-filled syringe. The lumbar part of the spinal cord was immediately homogenized in extraction buffer [50 mm Tris buffer, pH 8.0, containing 0.5% Triton X-100, 150 mm NaCl, 1 mm EDTA, protease inhibitor cocktail (P-8340, 1 : 100, Sigma) and phosphatase inhibitor cocktail I and II (1 : 100, Sigma)] by sonication. The tissue extracts were subjected to denaturing NuPAGE 4–12% Bis-Tris gel electrophoresis (Invitrogen, Carlsbad, CA, USA) and then transferred to nitrocellulose membranes (Micronic Separation Inc., Westborough, MA, USA). After blocking non-specific binding sites with 5% low-fat milk in Tris-based buffer (50 mm Tris-Cl, 6 mm NaCl) containing 0.1% Tween 20 for 1 h at room temperature (22–24°C), the membranes were incubated with antibodies overnight at 4°C. After washing, the antibody–protein complexes were probed with appropriate secondary antibodies labeled with horseradish peroxidase for 1 h at room temperature and detected with chemiluminescent reagents (SuperSignal; Pierce, Rockford, IL, USA). The nitrocellulose membranes were stripped with a Re-Blot western blot recycling kit (Chemicon, Temecula, CA, USA) and reblotted with different antibodies. The antibodies used in this study were: phosphorylated p38 (1 : 1000); p38α (1 : 500); p38β (1 : 1000; Zymed, San Francisco, CA, USA); total p38 (1 : 1000; Cell Signaling Technology, Beverly, MA, USA) and β-actin (1 : 5000; Sigma). The p38β antibody is currently characterized for mouse protein. To confirm that the p38β antibody recognizes rat p38β, spinal cord samples from adult Holzman Sprague-Dawley rats and C3H/HeJ mice (Jackson Laboratory, Bar Harbor, ME, USA) were subjected to western blotting. The intensity of immunoreactive bands was quantified using ImageQuant software (Molecular Dynamics, Sunnyvale, CA, USA). p38-immunopositive bands were normalized relative to β-actin.

Immunohistochemistry

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Animals were deeply anesthetized with sodium pentobarbital (50 mg/kg body weight) and perfused intracardially with heparinized saline (200 mL) followed by freshly prepared 4% paraformaldehyde in 0.1 m phosphate-buffered saline (pH 7.4; Sigma). The lumbar spinal cord was removed, post-fixed in the same fixative for 6 h and transferred to phosphate-buffered saline containing 20% sucrose for 24 h and then 30% sucrose for 48 h. The lumbar segments L3–6 were dissected and transverse sections (10 µm) were cut with a freezing microtome and mounted on silane-covered glass slides. Non-specific binding was blocked by incubation in 5% normal goat serum in phosphate-buffered saline with 0.2% Triton X-100 followed by incubation with primary p38α antibody (generated in rabbit, 1 : 500; Cell Signaling Technology) overnight at 4°C under gentle agitation. Binding sites were visualized with anti-rabbit IgG antibodies conjugated with Alexa-488 (1 : 250). To determine the cellular distribution of p38α, neurons, astrocytes and microglia were counterstained with primary antibodies raised in mouse against markers for neurons (neuronal N, 1 : 1000; Chemicon), astrocytes (glial fibrillary acidic protein, 1 : 500; Chemicon) and microglia [CD11b (OX-42), 1 : 100; Biosource International, Dallas, TX, USA], respectively. Binding sites were visualized with anti-mouse IgG antibody conjugated with Alexa-594 (1 : 250). To determine the cellular distribution of p38β, antigen retrieval was performed by microwaving slides for 30 s while covered with sodium citrate buffer (pH 6), followed by incubation with p38β antibody (1 : 500; Zymed). Binding sites were visualized with anti-mouse IgG antibodies conjugated with Alexa-488. These sections were counterstained with primary antibodies raised in mouse against neuronal marker (neuronal N, 1 : 1000, conjugated to biotin; Chemicon) detected with avidin conjugated to Cy3 (1 : 1000; Sigma) and astrocyte marker (glial fibrillary acidic protein, 1 : 1000, labeled with Cy3; Sigma). Microglia were detected using primary antibody raised in rabbit against a microglia-specific calcium-binding protein (Iba-1, 1 : 1000; Wako Chemicals, Richmond, VA, USA) and binding sites visualized using rabbit IgG antibody conjugated with Alexa-594 (1 : 250). All antibodies were diluted in 0.5% Triton X-100 and 5% goat serum in phosphate-buffered saline. Reagents conjugated to Alexa fluorophores were purchased from Molecular Probes (Eugene, OR, USA). Coverslips were mounted on the glass slides with ProLong antifade medium (Molecular Probes). Non-specific staining was determined by excluding the primary antibodies. Images were captured using a confocal microscopy system (Nikon, Melville, NY, USA) operated by Lasersharp 2000 software (Biorad, Hemel Hemstead, UK).

Nociceptive models

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

To induce spinal sensitization directly, rats received IT injections of SP (30 nmol) which activates spinal NK-1 receptors and produces thermal hyperalgesia (Piercey et al. 1981; Yashpal and Henry 1983; Hua et al. 1999). Thermally-evoked paw withdrawal responses were assessed using a Hargreaves-type testing device (Dirig et al. 1997). Briefly, rats were placed on a glass surface maintained at 30°C. The thermal nociceptive stimulus originates from a projection bulb below the glass surface and the stimulus is delivered separately to one hindpaw at a time. The animals were allowed to acclimate to the device and basal paw withdrawal latencies were then assessed for left and right paws at time (T) = −40, −30 and −20 min and expressed as the mean of the six measurements. At T = −10 min the animals received IT vehicle or drug and at T = 0 SP was injected. Withdrawal latencies were then assessed at T = 15, 30, 45 and 60 min. In the AS experiment withdrawal latencies were measured at T = 15 and 27 min. At T = 30 min these animals were killed and spinal cords collected for western blot analysis. Paw withdrawal latency for each group was expressed as the mean at each time-point.

An automated sensing system was employed to quantify formalin-induced flinching, (Yaksh et al. 2001b). Briefly, a soft metal band was placed on the hindpaw of the animal being tested. Animals were allowed to acclimate in individual Plexiglas chambers for 30 min before being moved to a test chamber. Just before the animals were placed into the test chamber, they were briefly restrained in a cloth towel and 2.5% formalin (50 µL) was injected into the dorsal side of the banded paw. Nociceptive behavior was quantified by automatically counting the incidences of spontaneous flinching or shaking of the injected paw. The flinches were counted for 1-min periods for 60 min and the flinch data are expressed as number of flinches per min and total flinches observed during phase I (0–9 min) and phase II (10–60 min).

Intrathecal dialysis and prostaglandin E (PGE)2 assay

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Dialysis experiments were conducted in unanesthetized rats 4–5 days after the implant. A syringe pump (Harvard, Natick, MA, USA) was connected and dialysis tubing was perfused with ACSF at a rate of 10 µL/min. The efflux (15 min per fraction) was collected in an automatic fraction collector (Eicom, Kyoto, Japan) at 4°C. Two baseline samples were collected after a 30-min washout and an additional four fractions after IT injection of SP (30 nmol in 10 µL saline followed by 10 µL of saline to flush the injection line). The concentration of PGE2 in spinal dialysate was measured by ELISA using a commecialy available kit (90001; Assay Designs, Ann Arbor, MI, USA). The antibody is selective for PGE2 with less than 2.0% cross-reactivity to PGF, PGF, 6-ketoPGF1α, PGA2 or PGB2 but cross-reacts with PGE1 and PGE3.

Statistics

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

For measurements of protein levels by western blotting, five to six animals were included per group and there were four to eight animals in each group for analysis of behavior and PGE2 release. Differences between groups were compared with one-way anova and a Turkey (San Diego, CA, USA) post-hoc test except for IT SP-evoked PGE2 release where an unpaired Student's t-test was applied (Prism statistical software). For IT SP-evoked thermal hyperalgesia a two-way anova repeated measurements with a Bonferroni post-hoc test was employed (StatView statistical software, Cary, NC, USA).

p38α and p38β are expressed in rat spinal cord

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Both p38α and p38β protein expression were detected in rat lumbar spinal cord homogenates by western blotting (Fig. 1). A single immunoreactive band was detected at about 40 kDa with the p38α antibody (Fig. 1a). Four different p38β antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA and Zymed) were tested in order to identify the antibody that gave the best signal. In our setting the monoclonal mouse anti-p38β antibody from Zymed provided the best result. This antibody has previously been characterized for mouse, but not rat, p38β. To confirm that the antibody cross-reacts with rat p38β, mouse and rat naive spinal cord samples were run concurrently. A band with an approximate molecular size of 40 kDa was detected, at two different protein concentrations for both species (Fig. 1b), supporting the suggestion that this antibody recognizes rat p38β protein and also that p38β is expressed in rat spinal cord. Additional bands were detected in both mouse and rat tissue; a lighter band (20 kDa) was detected in the mouse tissue and a heavier band (80 kDa) in the rat tissue.

image

Figure 1. Western blot showing immunoreactive bands at approximately 40 kDa (arrows) indicating that p38α (a) and p38β (b) are expressed in naive mouse (M) and rat (R) spinal cord (50 and 20 µg total protein loaded). Mouse and rat spinal cord samples were run on the same gel to confirm that the p38β antibody cross-reacts with rat p38β.

Download figure to PowerPoint

p38α is expressed in neurons while p38β is primarily expressed in microglia

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Immunohistochemistry was undertaken in order to determine the cellular distribution of p38α and p38β. Spinal cord sections from naive animals were incubated with antibodies against p38α and p38β and also with cellular markers for neurons, astrocytes and microglia. p38α colocalized with the neuronal marker neuronal N throughout the dorsal and ventral spinal parenchyma. However, the staining was weaker in laminae I and II in the dorsal horn in comparison to deeper laminae or the ventral horn (Figs 2a–c and e). p38α does not appear to be expressed in microglia or astrocytes in naive rat spinal cord, as no colocalization could be detected between this isoform and OX-42 (Fig. 2g) or glial fibrillary acidic protein (Fig. 2h).

image

Figure 2. Distribution and cellular localization of p38α in naive rat spinal cord. (a) p38α immunoreactivity distributed throughout the dorsal horn parenchyma. The signal intensity is higher in deeper than in superficial laminae. (b) Presence of p38α immunoreactivity in the ventral horn of naive rat spinal cord. (c–f) Double immunofluorescence confocal micrographs showing that spinal p38α (c, green) and the neuronal marker neuronal N (NeuN) (d, red) are colocalized in the dorsal horn (e, merged overlap appears as yellow) indicating that p38α is expressed in neurons. (f) Higher magnification of (e) indicating overlap of neuronal marker and p38α staining. (g) Double labeling with antibodies against p38α (green) and the microglia marker OX-42 (red) showed no overlap indicating that p38α is not expressed in microglia. (h) Double immunofluorescence staining p38α (green) and glial fibrillary acidic protein (GFAP) (red) showed no overlap indicating that p38α is not expressed by astrocytes. Size bars, 50 µm. Similar results were observed in spinal cord tissue from a total of four rats.

Download figure to PowerPoint

p38β was distributed homogeneously throughout the spinal parenchyma (Figs 3a and b). In contrast to the neuronal distribution of p38α, p38β was colocalized primarily with the microglia marker Iba-1 in the dorsal horn (Figs 3c–f). However, in the ventral horn p38β was also noted in motor neurons as indicated in Fig. 3(b). In the dorsal horn p38β was not present in cells positive for neuronal N (Fig. 3g) or glial fibrillary acidic protein (Fig. 3h), indicating that p38β is not expressed in dorsal horn neurons or astrocytes.

image

Figure 3. Distribution and cellular localization of p38β in naive rat spinal cord. (a) Micrograph depicting p38β expression throughout the parenchyma of dorsal horn of naive rat spinal cord. (b) Micrograph indicating presence of p38β in glia-like cells as well as large motor neurons in naive rat ventral horn (arrows, motor neuron-like cells; arrowheads, microglia-like cells). (c–f) Double immunofluorescence micrographs demonstrating that antibodies against p38β (c, green) and the microglia marker Iba-1 (d, red) labels the same cell (e) indicating that p38β is expressed in microglia. (f) Higher magnification of microglia containing p38β. (g) Colabeling of p38β (green) and Neuronal N (NeuN) (red) showed overlap in the ventral horn (data not shown) but not dorsal horn. (h) Immunofluorescence double labeling of p38β (green) and the astrocyte marker glial fibrillary acidic protein (GFAP) (red) showing no overlap in the dorsal horn indicating that p38β is not expressed in astrocytes. Size bars, 50 µm. Similar results were observed in spinal cord tissue from a total of four rats.

Download figure to PowerPoint

Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

It has previously been shown that inhibition of spinal p38 prevents both inflammation-evoked hyperalgesia and nerve injury-induced allodynia (Ji et al. 2002; Schafers et al. 2003; Svensson et al. 2003a,b). It remains to be determined whether only one or both of the spinal p38 isoforms are involved in modulation of spinal nociceptive processing. As there are no specific p38 isoform inhibitors available, rats were treated intrathecally with isoform-specific AS in order to inhibit expression of p38α and p38β individually. After 5 days of IT bolus injections with p38α AS (10–30 µg) or p38β AS (10–20 µg) there was a significant down-regulation of spinal p38α(Fig. 4a; 30 µg, p < 0.05) and p38β (Fig. 4b; 10 and 20 µg: p < 0.05) protein expression as compared with the respective protein expression in the control MS-treated animals. There were no differences between the AS-, MS- or vehicle-treated groups in terms of weight gain or motor function (data not shown). We also examined the specificity of the effects of these two AS. Western blot membranes were stripped and reincubated with the alternate p38 isoform antibody. As shown in Table 1, neither p38α AS nor p38β AS produced a reduction of the protein expression of the other isoform, although the p38α protein level was increased in spinal cords of rats that had received p38β AS, as compared with MS-treated rats (Table 1).

image

Figure 4. Effect of intrathecal (IT) delivery of p38 isoform-specific antisense oligonucleotides (AS) on spinal p38α and p38β protein expression. (a and b) p38α and p38β protein expression as percentage change from levels assessed in spinal cords of IT missense oligonucleotide (MS)-treated rats. Two doses of p38α AS (a, 10 and 30 µg) or p38β AS (b, 10 and 20 µg) were administered intrathecally for 5 days which led to a statistically significant down-regulation of p38α and p38β, respectively, *p < 0.05 compared with MS. Representative western blots are shown below graphs. Data were normalized against β-actin protein expression in each sample and the same universal MS oligonucleotide was used as a control. Bars represents mean ± SEM, n = 6 rats per group.

Download figure to PowerPoint

Table 1.  Protein expression after intrathecal (IT) antisense oligonucleotide (AS)
Protein expressionIT AS% expression compared with MS
  1. AS was given intrathecally once daily for 5 days and protein expression assayed in spinal cord at day six and compared with protein level detected in missense oligonucleotide-treated animals (MS = 100%). Each group represents mean ± SEM, n = 6 per group. *p < 0.05 as compared with the MS group.

p38αα-AS (10 µg)59 ± 26
α-AS (30 µg)39 ± 8*
β-AS (20 µg)205 ± 65*
p38ββ-AS (10 µg)45 ± 9*
β-AS (20 µg)39 ± 12*
α-AS (30 µg)111 ± 15

Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

In a recent work we showed that inhibition of spinal p38 attenuated formalin-induced flinching (Svensson et al. 2003b). To examine the contribution of p38α and p38β in this experimental model of hyperalgesia, animals were injected with IT p38α, p38β AS or MS for 5 days. These animals were then examined using the formalin test. Unilateral injection of formalin solution (2.5%) into the dorsal side of the paw caused a biphasic flinching pattern where the first 9 min are referred to as phase I and the following 50 min (min 10–60) as phase II (Yaksh et al. 2001b). Phase II flinching is considered to represent spinal facilitation initiated by afferent input (Dickenson and Sullivan 1987; Puig and Sorkin 1996). The total number of flinches during phase I or II recorded for animals treated with IT p38α AS did not differ from the animals treated with IT saline or MS (Figs 5a and b). In contrast, rats that received p38β AS displayed a statistically significant reduction in the number of flinches during phase two (Figs 5c and d) as compared with saline- and MS-treated animals (p < 0.05). Rats that received IT MS did not differ in number of flinches from IT saline-injected rats (Figs 5a and b). To further confirm the role of p38 in injury-induced pain behavior, a pharmacological inhibitor (SB203580, a p38α/β non-selective inhibitor) was also studied. IT injection of SB203580 (3–30 µg) 10 min before formalin injection into the paw resulted in a dose-dependent attenuation of flinching during phase II (Figs 5e and f; 30 µg; p < 0.05).

image

Figure 5. Effect of intrathecal (IT) p38α and p38β antisense oligonucleotides (AS) and SB203580, a p38 inhibitor, on formalin-induced hyperalgesia. Graphs show number of flinches/min plotted vs. time after injection of formalin (2.5%, 50 µL) into the dorsal side of the right hindpaw in rats treated with (a) IT p38α AS (30 µg), missense oligonucleotide (MS) (3–10 µg) or saline, (c) IT p38β AS (20 µg) or MS (20 µg) daily for 5 days and (e) IT SB203580 (30 µg) or vehicle 10 min before formalin injection. Bar graphs show cumulative number of flinches after paw injection of formalin during phase I (0–9 min) and phase II (10–60 min) in rats treated with different doses of (b) IT p38α AS (3–30 µg), MS (3–10 µg) or saline, (d) IT p38β AS (3–20 µg) or MS (20 µg) daily for 5 days and (f) IT SB203580 (3–30 µg) or vehicle 10 min before formalin injection. *p < 0.5 vs. MS (d) and vehicle (f). Time-points and bars represent mean ± SEM, n = 6–8 for each group.

Download figure to PowerPoint

Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Injection of formalin into the hindpaw leads to phosphorylation of p38 in spinal cord within 5 min of injection (Svensson et al. 2003b). We performed a full time-course (0, 5, 15, 20, 30 and 60 min) for formalin-induced phosphorylation of spinal p38 and found peak p38 phosphorylation at 15 min (Fig. 6a). As shown above, down-regulation of spinal p38β leads to attenuation of formalin-induced hyperalgesia and we hypothesized that down-regulation of spinal p38β would reduce the amount of phosphorylated p38 in the spinal cord after formalin injection. As shown in Figs 6 (b and c), IT treatment with p38β AS results in a reduction of phosphorylated p38 in the spinal cord (assessed 15 min after formalin injection) in comparison to MS-treated rats. Down-regulation of p38α protein expression did not prevent spinal p38 phosphorylation after paw injection of formalin as indicated by the lack of difference in spinal p38 phosphorylation between rats injected with IT p38α AS and MS in this model (Figs 6b and c).

image

Figure 6. Effect of intrathecal (IT) p38 isoform-specific antisense oligonucleotides (AS) on phosphorylation of spinal p38 evoked by injection of formalin into the hindpaw. (a) Representative western blots showing levels of phosphorylated p38 (P-p38), total p38 and β-actin in the contralateral (left side) and ipsilateral (right side) lumbar spinal cord at different time-points after injection of formalin (2.5%, 50 µL) into the dorsal side of the right hindpaw. (b) Rats were naive (no IT treatment, no formalin paw injection) or treated with p38α AS (30 µg), p38β AS (20 µg) or missense oligonucleotide (MS) (30 µg) for 5 days. On day six, formalin was injected into the hindpaw and the level of P-p38 assessed 15 min after injection. Graph displays percentage change in spinal P-p38 as compared with naive spinal cord. Data were normalized against β-actin protein expression in each sample. Time-points and bars represent mean ± SEM, n = 5 for each group. *p < 0.05. (c) Representative western blots depicting levels of spinal P-p38, p38α, p38β and β-actin 15 min after formalin injection into the paw.

Download figure to PowerPoint

Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Hyperalgesia is associated with persistent sensory afferent input-induced spinal sensitization. SP, which is released from central terminals of afferent C-fibers (Yaksh et al. 1980), contributes significantly to this facilitatory state in spinal cord (Yaksh 1999). We have shown that IT injection of SP (30 nmol) produces thermal hyperalgesia (Malmberg and Yaksh 1992) and that inhibition of spinal p38α and β by IT SD-282 blocks this hyperalgesia (Svensson et al. 2003b). In the present study, we demonstrated that there is a time-dependent phosphorylation of spinal p38 in a temporal pattern similar to the thermal hyperalgesia induced by IT SP (Fig. 7b), i.e. starting at 5–15 min with a maximum at 30 min. To investigate the contribution of p38α and p38β to SP-evoked hyperalgesia and p38 phosphorylation, we again undertook spinal delivery of p38α and p38β AS and MS. Thermal withdrawal latencies were assessed 27 min after IT injection of SP and the lumbar spinal cords of these animals collected 30 min after IT SP to compare the degree of spinal p38 phosphorylation at the previously determined time of peak phosphorylation. IT injection of SP into the MS group produced an expected reduction of thermal withdrawal latency in comparison with IT saline injection into vehicle-treated animals (Fig. 7c). The decrease in threshold after IT injection of SP into the MS-treated group was similar to that seen after IT administration of SP in saline-treated animals (Fig. 7a), indicating an absence of effect of IT MS. The rats receiving p38α AS showed a similar degree of thermal hyperalgesia after IT SP as the MS group (Fig. 7f). In contrast, IT SP-induced thermal hyperalgesia was attenuated in rats that had received p38β AS (Fig. 7f). Baseline thermal thresholds did not differ among the groups (ACSF, 10.9 ± 0.4; MS, 11.4 ± 0.5; p38α AS, 11.5 ± 0.3 and p38β AS, 10.8 ± 0.5 s). In accordance with the formalin study, the p38α AS-treated group did not show a reduction in phosphorylation of spinal p38 after SP (Figs 7d and e) while there was a pronounced reduction in spinal p38 phosphorylation in the p38β AS-treated group, as compared with the MS group (Figs 7g and h). Stripping and reprobing the western blot membranes confirmed that p38α protein expression was knocked down in the p38α AS-treated group (Fig. 7d) as well as p38β in the p38β AS-treated group (Fig. 7g).

image

Figure 7. Effect of intrathecal (IT) p38α and p38β antisense oligonucleotides (AS) on thermal hyperalgesia and spinal p38 phosphorylation evoked by IT substance P (SP). (a) Paw withdrawal latency (PWL) plotted vs. time after IT injection of SP (30 nmol) or saline at T = 0. (b) Representative western blots showing levels of phosphorylated p38 (P-p38), total p38 and β-actin in lumbar spinal cord at different time points after IT injection of SP (30 nmol). +, positive control (C-6 anisomycin-treated glioma cell extract; Cell Signaling Technology). PWL assessed 15 and 27 min after bolus injection of IT saline or IT SP injection in rats treated with (c) IT artificial cerebrospinal fluid (ACSF) (10 µL) or missense oligonucleotide (MS) (10 µ g) and (f) p38α AS (10 µg) or p38β AS (10 µg) daily for 5 days. Time-points represent the mean ± SEM, n = 4–5 per group. (d and g) Representative western blots for the bar graph data (e and h) showing the level of P-p38, p38α, p38β and β-actin protein expression assessed 30 min after IT SP injection.

Download figure to PowerPoint

Intrathecal substance P-induced PGE2 release is mediated by activation of p38

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Prostaglandins, including PGE2, have been implicated in playing a role in the spinal sensitization and hyperalgesia produced by IT SP (Hua et al. 1999; Yaksh 1999). It has been reported that phospholipase A2, which generates free arachidonic acid necessary for eicosanoid synthesis, requires phosphorylation to become fully activated and that this phosphorylation can be mediated by p38 (Borsch-Haubold et al. 1997). To address the question of how p38 is involved in spinal sensitization, we examined the effect of spinal p38 inhibition on IT SP-evoked PGE2 release in an in vivo IT dialysis model. In the absence of pre-treatment, baseline concentrations of PGE2 in dialysates were determined to be 367 ± 14 fmol/100 µL (n = 20). IT SP (30 nmol) produced six- to eightfold increases in PGE2 level in IT CSF(Fig. 8a; p < 0.05). Pre-treatment with the p38 inhibitor SB203580, at the dose of 30 µg (IT) that significantly blocked IT SP-induced thermal hyperalgesia (Svensson et al. 2003a), prevented IT SP-evoked PGE2 release (Figs 8a and b). Comparison of area under curve between groups showed that SP-evoked release of PGE2 was significantly lower in the group that received SB203580 before SP (unpaired t-test, p < 0.05, Fig. 8b). Because of the limitation of viability of long-term IT dialysis probes (Koetzner et al. 2004), we could not perform chronic treatment with IT AS in this dialysis model. However, as our data indicate that IT SP-induced p38 activation is primarily attributed to the p38β isoform (see Fig. 7f), we believe that IT SP-induced PGE2 release is probably mediated by activation of p38β.

image

Figure 8. Inhibition of spinal p38α/β attenuates substance P (SP)-evoked spinal PGE2 release. (a) PGE2 concentration (percentage of baseline) measured in cerebrospinal fluid collected by in vivo spinal dialysis of conscious rats before and after intrathecal (IT) injection of SP (30 nmol) in the presence of SB203580 (30 µg) or vehicle. Spinal dialysate was removed and assayed for PGE2 by ELISA at baseline (b, average of two 15-min fractions collected before IT injection of vehicle or SB203580) and F1, 0–15 min ; F2, 15–30 min ; F3, 30–45 min and F4, 45–60 min after SP injection. Rats received vehicle or IT SB203580 15 min before IT SP. (b) PGE2 release is presented as area under curve (AUC) calculated 60 min after IT injection of SP. Bars represent the average and SEM for four to eight rats per group; *p < 0.05 vs. IT SP.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

The present findings comprise four key observations. First, we demonstrated that p38α and p38β have different cellular locations in the spinal dorsal horn. p38α is expressed in neurons, while p38β is predominantly expressed in microglia. Second, we were successful in selectively down-regulating spinal p38α and p38β protein expression by IT treatment with the respectively targeted AS. Third, block of p38β, but not p38α, protein expression prevented the appearance of phosphorylated p38 otherwise evoked by small afferent input (i.e. intraplantar formalin injection) or by direct activation of NK-1 receptors (i.e. IT SP injection). The essentially complete block of p38 phosphorylation by p38β AS suggests that activation of spinal p38 was largely attributable to the p38β isoform. Fourth, selective down-regulation of p38β, but not p38α, prevented hyperalgesia in both animal models. The importance and novelty of the present findings are not only in defining the role of p38β in the development of spinal sensitization but also in affirming the hypothesis that a non-neuronal origin of this isoform, i.e. microglia, mediates the hyperalgesic phenotype.

Intrathecal antisense oligonucleotides to define spinal p38 isoform function

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

The initial identification of a functional role for spinal p38 in nociception arises from work showing the effect of IT delivery of p38α/β inhibitors at doses considerably lower than those reported to be active after systemic delivery (Jin et al. 2003; Schafers et al. 2003; Svensson et al. 2003b; Sweitzer et al. 2004a). By either route, the identified behavioral consequences are an absence of effect upon acute nociception but rather a pronounced suppression of nerve- and tissue-injury-evoked hyperalgesia (Watkins et al. 1997; Jin et al. 2003; Milligan et al. 2003; Schafers et al. 2003; Svensson et al. 2003b; Tsuda et al. 2004). The p38α and β isoforms are both found within the CNS and, given their distinct cellular distribution in the dorsal horn, the issue of which isoform(s) is relevant to the observed facilitated nociceptive processing must be considered. Current agents used to target p38 are able to distinguish α and β from the other isoforms but, to our knowledge, no available small molecule inhibitors exist which adequately distinguish between p38α and p38β. The use of transgenic animals to address this issue is precluded at present by the lethality of the p38 knockout (Adams et al. 2000; Mudgett et al. 2000). Accordingly, we considered the utility of intrathecally delivered AS.

The use of IT AS has been well documented as a strategy to transiently knock down the expression of constitutively expressed spinal protein (Stone and Vulchanova 2003). In the present study, we used 2′-methoxyethyl-modified AS. Previous work with this structure has shown it to possess minimal toxicity and to be effective in down-regulating the expression of proteins (Henry et al. 2000; Butler et al. 2002), including spinal kinases (Hua et al. 2002). The efficacy and specificity of the two individual AS employed in the present study was confirmed by the demonstration of (i) dose-dependent inhibition of protein expression; (ii) differential regulation of two closely related proteins, i.e. p38α and p38β; (iii) complete lack of activity of equimolar concentrations of control MS and (iv) lack of sensory or motor dysfunction associated with either IT AS. The observation in the present study, that only treatment with IT p38β AS produced a uniform antihyperalgesic effect in addition to preventing stimulus-induced spinal p38 phosphorylation, is strong evidence supporting the role of the p38β isoform in spinal nociceptive processing.

p38α and p38β isoform in spinal cord

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

Our finding that different spinal cell types express p38α and p38β is especially significant in light of the different roles of these isoforms in induction of hyperalgesia. Here we report that p38α is found predominantly in neurons while p38β is localized to microglia throughout the spinal parenchyma and motor neurons in the ventral horn. The absence of p38α in spinal microglia is somewhat unexpected given its reported expression in human primary cultures of monocytes and macrophages (Hale et al. 1999). While we observed p38β protein expression in spinal microglia, Hale et al. (1999) could not detect p38β in either monocytes or macrophages, pointing to differences in hematopoetic cells potentially based on their location (peripheral vs. central) as well as their activation state (cultured cells vs. naive tissue).

Although the data presented show that down-regulation of p38α does not affect hyperalgesia in our models it does not exclude the possibility that this isoform may be involved in other models of more chronic hyperpathia. For example, p38 is known to regulate long-term adaptive changes (hours to days) in expression of proteins such as cytokines (Kumar et al. 2003), enzymes (e.g. cyclooxygenase-2) (Lasa et al. 2000) and receptors (e.g. vanilloid receptor-1) (Ji et al. 2002) that are important for spinal sensitization. At this point it is unknown which isoform is responsible for this long-term regulation. In addition, it has been reported that p38α is involved in peripheral inflammation (Hale et al. 1999) and this isoform might play a more critical role in nociception at the local site of injury. We noticed that there was an increase in spinal p38α expression in the rats treated with p38β AS, suggesting a compensatory up-regulation of p38α in response to the decreased expression of p38β. As IT delivery of SB203580, a potent inhibitor of both isoforms (Barone et al. 2001), prevented formalin-induced flinching, an effect parallel to that seen in p38β knock-down animals, we believe that the p38α isoform does not have an acute role in spinal nociceptive processing. We have previously shown that IT administration of the p38 inhibitor SD-282 prevents hyperalgesia induced by IT SP and injection of formalin or carrageenan into the paw. It has been reported that SD-282 is a p38α-selective inhibitor, which might seem contradictory to our current data. However, the IC50 for SD-282 has been reported to be 1.61 and 23 nm for p38α versus p38β by Li et al. (2004) and 1.4 and 22.5 nm for p38α versus p38β by Sweitzer et al. (2004b). This gives IC50 ratios of 14 and 16 and we feel that this ratio is not high enough to ensure selectivity. For comparison, SB203580, referred to as a p38α/p38β non-selective inhibitor, has an IC50 ratio of between 5 (Davies et al. 2000) and 10 (Calbiochem, San Diego, CA, USA). The similar antinociceptive profile of these two inhibitors in tandem with the results from our isoform-specific antisense data make us inclined to believe that the antihyperalgesic effect observed in our previous work, using SD-282, is mainly due to p38β inhibition.

Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

In the present study, we show that IT SP produces a thermal hyperalgesia and a concurrent release of PGE2. Previous work has shown that SP-evoked thermal hyperalgesia and spinal PGE2 release are mediated by the activation of NK-1 receptors (Hua et al. 1999) and that this PGE2 release and hyperalgesia are reversed by spinal cyclooxygenase-2 inhibition (Malmberg and Yaksh 1992; Yaksh et al. 2001a). In addition, hyperalgesia induced by IT SP is blocked by IT delivery of the p38 inhibitor SB203580 (Svensson et al. 2003a). In the present work, we extended these observations by showing that the hyperalgesia was mediated by p38β but not p38α. We further show that PGE2 release evoked by IT SP is also blocked by the same dose of SB203580, which reverses the hyperalgesia (Svensson et al. 2003a). The use of IT AS in defining the isoform role in PGE2 release was precluded by the technical necessity of having to treat the animal with IT AS for 5 days and the limited viability of long-term IT dialysis probes (Koetzner et al. 2004). Nevertheless, these data jointly support the hypothesized cascade in which it is envisioned that persistent excitatory input activating a variety of dorsal horn receptors leads to activation of p38 isoforms. This, in turn, activates constitutively expressed dorsal horn phospholipase A2 and cyclooxygenase-2 yielding an increase in extracellular PGE2. Abundant evidence has shown the importance of local PGE (EP) receptors in facilitating dorsal horn processing through an enhanced afferent transmitter release (Nicol et al. 1992; Hingtgen et al. 1995; Vasko 1995; Southall and Vasko 2001) and a direct sensitization of the post-synaptic neuron (Baba et al. 2001).

Microglia in nociceptive processing

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References

The present demonstration of an essential role for p38β in facilitation of spinal pain processing provides direct support for an equally essential role of the local dorsal horn microglia. Our findings that p38β contributes to hyperalgesia and that this isoform is located in microglia, provide an important corollary to previous observations that p38 is activated in spinal cord microglia in several models of tissue and nerve injury (Ji et al. 2002; Svensson et al. 2003b; Tsuda et al. 2004). While glia have typically been viewed as playing a supportive role in neuronal function, it has become increasingly clear that these cells are crucial in the regulation of spinal plasticity (Watkins et al. 2001).

An important question is the nature of the communication between sensory neurons and microglia. Microglia express receptors for many neurotransmitters and neuromodulators (Inagaki et al. 1991; Palma et al. 1997; Kommers et al. 1998) and can synthesize and release neuroactive factors upon activation, including prostanoids and cytokines (Watkins and Maier 2000). The present study, together with previous work (Svensson et al. 2003a), indicates that SP is one of the mediators initiating signaling from primary afferents to microglia. Previously, we have demonstrated by immunocytochemistry that phosphorylated p38, observed 10 min after IT SP, is exclusively located in microglia (Svensson et al. 2003a) and this effect is mediated via activation of NK-1 receptors (C.I. Svensson, unpublished observation). Expression of functional NK-1 receptors on cultured murine microglia has been reported (Rasley et al. 2002) although it has not been confirmed in rat spinal cord in vivo. A number of other linkages are also likely. ATP is released from afferents and interneurons and astrocytes trigger microglial responses through activation of the ATP receptor P2X4 (Inoue 2002; Tsuda et al. 2003). The chemokine fractalkine is released from neuronal membranes (Chapman et al. 2000) and can act upon the fractalkine receptor CX3CR1 found predominantly on microglia in brain (Harrison et al. 1998) and spinal cord (Verge et al. 2004). Recent work indicates that fractalkines are involved in spinal sensitization (Milligan et al. 2004).

In conclusion, our results show that p38β in the spinal cord plays an acute and essential role in the cascades initiated by tissue injury and inflammation. The functional association of p38β and its presence in microglia provides strong support for the hypothesis that these cells play a substantial role in spinal nociceptive processing. Development of a centrally active p38β-selective inhibitor would contribute to the understanding of the function of this isoform and could constitute a potential therapeutic target for pain relief.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals
  5. Drugs: preparation and administration
  6. Antisense oligonucleotides
  7. Western blot
  8. Immunohistochemistry
  9. Nociceptive models
  10. Intrathecal dialysis and prostaglandin E (PGE)2 assay
  11. Statistics
  12. Results
  13. p38α and p38β are expressed in rat spinal cord
  14. p38α is expressed in neurons while p38β is primarily expressed in microglia
  15. Intrathecal p38α and p38β antisense oligonucleotides suppress protein expression of respective p38 isoform
  16. Reduction of spinal p38β, but not p38α, protein expression attenuates formalin-induced flinching
  17. Down-regulation of p38β, but not p38α, blocks formalin-induced phosphorylation of spinal p38
  18. Down-regulation of p38β, but not p38α, prevents hyperalgesia and p38 phosphorylation induced by intrathecal substance P
  19. Intrathecal substance P-induced PGE2 release is mediated by activation of p38
  20. Discussion
  21. Intrathecal antisense oligonucleotides to define spinal p38 isoform function
  22. p38α and p38β isoform in spinal cord
  23. Spinal p38 MAPK/phospholipase A2/cyclooxygenase cascade
  24. Microglia in nociceptive processing
  25. Acknowledgements
  26. References