SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Objective

It is known that noxious stimuli from inflamed tissue may increase the excitability of spinal dorsal horn neurons (a process known as central sensitization), which can signal back and contribute to peripheral inflammation. However, the underlying mechanisms have yet to be fully defined. A number of recent studies have indicated that spinal NF-κB/p65 is involved in central sensitization, as well as pain-related behavior. Thus, the aim of this study was to determine whether NF-κB/p65 can facilitate a peripheral inflammatory response in rat adjuvant-induced arthritis (AIA).

Methods

Lentiviral vectors encoding short hairpin RNAs that target NF-κB/p65 (LV-shNF-κB/p65) were constructed for gene silencing. The spines of rats with AIA were injected with LV-shNF-κB/p65 on day 3 or day 10 after treatment with Freund's complete adjuvant (CFA). During an observation period of 20 days, pain-related behavior, paw swelling, and joint histopathologic changes were evaluated. Moreover, the expression levels of spinal tumor necrosis factor α (TNFα), interleukin-1β (IL-1β), and cyclooxygenase 2 (COX-2) were assessed on day 14 after CFA treatment.

Results

The presence of peripheral inflammation in rats with AIA induced an increase in NF-κB/p65 expression in the spinal cord, mainly in the dorsal horn neurons and astrocytes. Delivery of LV-shNF-κB/p65 to the spinal cord knocked down the expression of NF-κB/p65 and significantly attenuated hyperalgesia, paw edema, and joint destruction. In addition, spinal delivery of LV-shNF-κB/p65 reduced the overexpression of spinal TNFα, IL-1β, and COX-2.

Conclusion

These findings indicate that spinal NF-κB/p65 plays an important role in the initiation and development of both peripheral inflammation and hyperalgesia. Thus, inhibition of spinal NF-κB/p65 expression may provide a potential treatment to manage painful inflammatory disorders.

Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by synovial inflammation and progressive destruction of the articular tissue. It is well known that continuous and intense nociceptive input from the inflamed joint may induce the hyperexcitability of spinal dorsal horn neurons (a process known as central sensitization) and hyperalgesia (an enhanced response to noxious stimuli) ([1-3]). Intriguingly, the increased excitation of the spinal cord can also signal back to the periphery via a variety of neuronal pathways, such as the autonomic nervous system ([4]), dorsal root reflexes ([5, 6]), and hypothalamic–pituitary–adrenal (HPA) axis, to control peripheral inflammation ([7]). There is convincing evidence to support the idea that a variety of spinally administered compounds attenuate peripheral inflammation. These compounds include adenosine ([8]), serotonin ([9]), ketamine and morphine ([10]), thalidomide ([11]), antagonists of microglia and inhibitors of astrocytes ([12]), and blockers of p38 MAPK ([13]).

NF-κB is a transcription factor that plays a pivotal role in the central nervous system (CNS), in processes such as inflammation, neuronal plasticity, synaptic transmission, learning, memory, and pain ([14]). NF-κB contributes to the regulation of these CNS processes by positively regulating the transcription of numerous genes, including cytokines (interleukin-1β [IL-1β], IL-6, and tumor necrosis factor α [TNFα]), proinflammatory enzymes (cyclooxygenase 2 [COX-2] and inducible nitric oxide synthase [iNOS]), chemokines, and adhesion factors ([14]). Five subunits of NF-κB have been identified, namely, gp105/p50 (NF-κB1), p100/p52 (NF-κB2), p65 (RelA), RelB, and c-Rel ([15]). The active form of NF-κB is a dimer formed from 2 of these subunits. The most common and best-characterized form of NF-κB is the p50/p65 heterodimer, which is widely expressed in the CNS and plays an important role in the regulation of gene expression ([16]).

Previous studies have indicated that activation of spinal NF-κB/p65 occurs in peripheral tissue damage or inflammation ([17, 18]). Findings in a recent study from our group showed that intrathecal injection of a lentiviral vector, LV-shNF-κB/p65, an inhibitor of NF-κB/ p65 that encodes short hairpin RNAs (shRNAs) targeting NF-κB/p65, significantly reduced the expression of spinal NF-κB/p65 and also reduced mechanical and thermal hyperalgesia following peripheral nerve injury ([19]). These findings suggest that NF-κB/p65 has a role in central sensitization. However, it is still unknown whether spinal NF-κB/p65 can also facilitate a peripheral inflammatory response.

Therefore, the goal of the present study was to test whether the inhibition of spinal NF-κB/p65 expression could significantly influence the progression and severity of peripheral inflammation and hyperalgesia, using an experimental arthritis model of rat adjuvant-induced arthritis (AIA). We also evaluated the expression of IL-1β, TNFα, and COX-2 in the spinal cord of these rats to gain insight into the mechanisms involved in the contributions of NF-κB/p65 to peripheral inflammation and hyperalgesia in this model system.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Animals

Experiments were performed on male Lewis rats (weight 180–220 gm) (Vital River Laboratories Animal Technology). All animals were kept in a temperature-controlled room at a constant temperature of 24 ± 1°C (mean ± SEM) and with a 12-hour light/dark cycle. Food and water were provided ad libitum. The experiments were conducted on independent groups of animals during the light phase, and all assays were performed in a quiet room by an observer (WCX) who was blinded with regard to the experimental conditions of the animals. All animal studies were approved by the Animal Care and Use Committee at Shandong University and were carried out in accordance with the university's guidelines for the care and use of laboratory animals. The ethics guidelines from the International Association for the Study of Pain were followed ([20]). Intrathecal catheters were implanted using the method described by Yaksh and Rudy ([21]). After a 7-day recovery period, all animals, except those that appeared to have signs of motor weakness or paralysis, were used in the experiments.

AIA model

Arthritis was induced by Freund's complete adjuvant (CFA) inoculation of the rats. Briefly, on day 0, rats were anesthetized with a mixture of ketamine plus xylazine (80:10 mg/kg, intraperitoneally) and then injected intradermally at the base of the tail with 0.1 ml CFA (1 mg/ml of heat-inactivated Mycobacterium tuberculosis in 85% paraffin oil and 15% mannide monooleate; Sigma). Rats in the control groups were injected with an equal volume of saline instead of CFA.

Lentiviral vector production

Lentiviral vectors were used to generate constructs encoding NF-κB/p65 small interfering RNAs (siRNAs) or control siRNA (scrambled sequence), as previously reported ([22]). The sequences of the 2 NF-κB/p65 siRNAs were as follows: 5′-GGACCTACGAGACCTTCAA-3′ (siRNA1) and 5′-GACATTGAGGTGTATTTCA-3′ (siRNA2). An additional scrambled sequence, 5′-TTCTCCGAACGTGTCACGT-3′, was designed as a negative control (NC). The lentiviral vectors LV-shNF-κB/p65 and LV-NC were then generated in a manner as previously described ([19]). The final titer of recombinant virus was 5 × 108 transducing units/ml.

Treatment protocols

On day 0, rats were injected with CFA to produce AIA. Control rats were injected with an equal volume of saline instead of CFA. Arthritis typically began to develop on day 10 and reached a maximum on days 16–20 ([8]). The animals that received 10-μl intrathecal injections of the lentiviral construct encoding either the scrambled shRNA sequence (LV-NC) or the NF-κB/p65 shRNA sequence (LV-shNF-κB/p65) were separated into 2 treatment groups. In one group, the rats were treated on day 3, and in the other group, the rats were treated on day 10. Thus, 128 rats were randomly assigned to 1 of 8 groups, consisting of 12–18 rats per group, as follows: the control group (saline only), the group receiving CFA only, the groups receiving CFA in conjunction with LV-NC, LV-shNF-κB/p65-1, or LV-shNF-κB/p65-2 on day 3 (d3), and the groups receiving CFA in conjunction with LV-NC, LV-shNF-κB/p65-1, or LV-shNF-κB/p65-2 on day 10 (d10).

In order to confirm the knockdown of NF-κB/p65 expression by spinally delivered LV-shNF-κB/p65, the L4–L6 lumbar segments of the spinal cord were removed on day 14, and the expression and activation of NF-κB/p65 were assayed by Western blot analysis, immunofluorescence, and electrophoretic mobility shift assay (EMSA) (n = 4–6/group). Further experiments with these groups of animals included assessment of pain-related behavior, evaluation of arthritis, and measurement of selected cytokine levels (n = 4–10/group).

Assessment of pain-related behavior

Mechanical thresholds (paw withdrawal thresholds) induced with CFA were evaluated with von Frey filaments (Stoelting) using the “up-down” method, as described in detail previously ([23]). Thermal thresholds (paw withdrawal latency) were detected using a BME-410A thermal dolorimeter (Biomedical Engineering Institute of the Chinese Academy of Medical Sciences), according to the methods described previously ([24]). Tests were performed at different times (6, 8, 10, 12, 14, 16, 18, and 20 days) following CFA injection. All groups were evaluated 1 day before the injection of CFA in order to determine the baseline mechanical and thermal thresholds.

Evaluation of arthritis

Paw edema assessment

Paw edema was determined by measurement of the paw volume using a water replacement plethysmometer (YLS-7A; Yiyan Sci). Measurements were obtained at baseline (1 day before CFA injection), and day 0 was the time point of CFA injection. Measurements were also obtained on days 6, 8, 10, 12, 14, 16, 18, and 20 following CFA injection.

Histopathologic assessment

Arthritic paws were collected on day 20, fixed with 10% formalin in phosphate buffered saline (PBS), and decalcified for 3 days. Sections from paraffin-embedded tissue were stained with hematoxylin and eosin. The arthritic ankles were scored with the following scale: 0 = no inflammation, 1 = mild inflammatory infiltrate in skin and overlying tissues, 2 = dense inflammatory infiltrate but no synovitis or arthritis, 3 = synovitis, 4 = hyperplastic synovium and/or inflammatory infiltrate in the joint, and 5 = arthritis with destruction of articular tissues and/or pannus formation ([25]).

Immunostaining of spinal cords

Immunofluorescence staining or double immunostaining was performed as previously described ([13]). Briefly, on day 14, the animals were killed and the spinal cord tissue (L4–L6) was harvested and cryosectioned at 10 μm. Tissue sections were fixed in 4% paraformaldehyde, washed in PBS, and blocked with blocking solution (10% normal goat serum and 0.3% Triton X-100 in 0.1M PBS), followed by incubation with polyclonal rabbit anti–NF-κB/p65 antibody (1:100; Abcam) for 24 hours at 4°C. Thereafter, the sections were incubated with tetramethylrhodamine isothiocyanate (TRITC)–conjugated goat anti-rabbit secondary antibodies (1:50; CoWin Biotech) for 1 hour at 37°C. Double immunofluorescence staining was performed using anti–NF-κB/p65 antibodies and antibodies specific for neurons (neuronal-specific nuclear protein [NeuN]), astrocytes (glial fibrillary acidic protein [GFAP]), or microglia (CD11b [OX-42]). Briefly, the spinal cord tissue sections were incubated with appropriate mixtures of the following antibodies for 24 hours at 4°C: rabbit anti–NF-κB/p65 and mouse anti-NeuN (1:500; Chemicon), mouse anti-GFAP (1:50; Santa Cruz Biotechnology), or mouse anti–OX-42 (1:200; Abcam). This was followed by incubation with a mixture of TRITC-conjugated goat anti-rabbit and fluorescein isothiocyanate–conjugated goat anti-mouse secondary antibodies (1:50; CoWin Biotech) for 1 hour at 37°C.

Western blot analysis

Spinal cord tissues (L4–L6) were harvested on day 14. Cytoplasmic and nuclear extracts from the tissues were obtained using a nuclear/cytoplasmic isolation kit (Pierce Biotechnology) and protein levels in the extracts were determined using a BCA Protein Assay kit (Pierce Biotechnology). Protein samples were electrophoresed on a 10% sodium dodecyl sulfate–polyacrylamide electrophoresis gel and transferred onto a PVDF membrane (Millipore). After blocking with 5% nonfat milk and 0.1% Tween 20 in Tris buffered saline solution for 2 hours at room temperature, membranes were incubated overnight at 4°C with polyclonal rabbit anti–NF-κB/p65 antibody (1:1,000; Abcam) or polyclonal rabbit anti–COX-2 antibody (1:1,000; Abcam). Membranes were then washed and incubated with anti-rabbit secondary antibody (1:5,000; Santa Cruz Biotechnology) for 1 hour at 37°C. After extensive washing, the protein bands were visualized using an enhanced chemiluminescence assay (Millipore), following the manufacturer's instructions. The blots were also probed with an anti–β-actin antibody (1:500; Santa Cruz Biotechnology) or anti–β-tubulin antibody (1:500; Santa Cruz Biotechnology), each of which served as a loading control.

EMSA

EMSA was used to study the effect of LV-shNF-κB/p65 on NF-κB activation. Using a LightShift Chemiluminescent EMSA kit (Pierce Biotechnology), 10 μg of nuclear extract was incubated with 10× binding buffer, 1 μg/μl poly(dI-dC), and 200 fmoles biotin-labeled double-stranded NF-κB binding consensus oligonucleotides (5′-AGTTGAGGGGACTTTCCCAGGC-3′) in a total volume of 15 μl. The binding reaction was performed for 20 minutes at room temperature. The DNA–protein complexes were electrophoresed on 6.0% nondenaturing polyacrylamide gels, electrotransferred, and detected according to the manufacturer's instructions. Specificity of binding was also examined by competition with a 100-fold excess of unlabeled oligonucleotides.

Measurement of cytokine levels

Levels of the proinflammatory cytokines TNFα and IL-1β were determined in the spinal cord segments (L4–L6) using previously described methods ([26]). Briefly, segment samples were pooled and homogenized in ice-cold PBS containing 0.05% Tween 20, 0.1 mM phenylmethylsulfonyl fluoride, 0.1 mM benzethonium chloride, 10 mM EDTA, and 20 IU aprotinin A. After centrifugation at 7,000g at 4°C for 10 minutes, the supernatants were stored at −80°C for future protein quantification. Cytokine levels were evaluated using an enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems), according to the manufacturer's recommendations.

Statistical analysis

All statistical analyses were performed using SPSS statistical software (version 19.0). Pain-related behavior and paw edema measurements were analyzed by two-way analysis of variance (ANOVA) with repeated measures, followed by the Bonferroni post hoc test. The variance factors for the two-way ANOVAs were time and group comparisons. Western blot, EMSA, and ELISA data were analyzed by one-way ANOVA, followed by the Bonferroni post hoc test. The variance factor for the one-way ANOVAs was group comparisons. Histopathologic data were analyzed by Kruskal-Wallis test, followed by the Mann-Whitney U test. P values less than 0.05 were considered significant. Results are expressed as the mean ± SEM.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

Silencing of the NF-κB/p65 gene by spinal delivery of LV-shNF-κB/p65.

On day 14 after CFA injection, Western blotting was performed on tissue sections of the lumbar spine enlargements (L4–L6). As shown in Figures 1A and B, the level of nuclear NF-κB/p65 protein differed significantly between the groups (F[4,20] = 13.958, P < 0.01). After Bonferroni post hoc test, the results revealed that the expression of NF-κB/p65 protein in the CFA group was markedly increased compared with that in the saline control group (P < 0.01), whereas no significant difference in NF-κB/p65 protein expression was observed between the CFA group and the d3 CFA + LV-NC group. Moreover, when compared with the CFA and d3 CFA + LV-NC groups, the CFA + LV-shNF-κB/p65-1 and p65-2 groups treated on day 3 showed a trend toward less expression of NF-κB/p65 protein (P < 0.05 versus d3 CFA + LV-shNF-κB/p65-1 and P < 0.01 versus d3 CFA + LV-shNF-κB/p65-2). Similarly, NF-κB/p65 protein expression in the CFA + LV-shNF-κB/p65-1 and p65-2 groups treated on day 10 was also significantly lower than that in the d10 CFA + LV-NC group (F[2,9] = 11.302, overall P < 0.01; P < 0.05 versus d10 CFA + LV-shNF-κB/p65-1 and P < 0.01 versus d10 CFA + LV-shNF-κB/p65-2).

image

Figure 1. Effects of spinally delivered lentiviral vector LV-shNF-κB/p65 on the expression of NF-κB/p65. Groups of rats were treated with saline only (control), Freund's complete adjuvant (CFA) alone, or CFA in conjunction with negative control lentiviral vector (LV-NC), LV-shNF-κB/p65-1, or LV-shNF-κB/p65-2 on day 3 or day 10 after CFA injection. A, Representative Western blots of NF-κB/p65 expression in nuclear extracts from the rat spinal cord tissue. β-tubulin was used as a loading control. B, Quantitation of the relative densities of NF-κB/p65 protein bands. Results are expressed as the ratio of NF-κB/p65 to β-tubulin, and the control is set at 1.0. Values are the mean ± SEM of 4–5 rats per group. ## = P < 0.01 versus control; ∗ = P < 0.05 and ∗∗ = P < 0.01 versus CFA alone or day 3 CFA + LV-NC; ▵ = P < 0.05 and ▵▵ = P < 0.01 versus day 10 CFA + LV-NC. C, Representative immunofluorescence detection of NF-κB/p65 in the rat spinal dorsal horns. Panels f–j are higher-power magnifications of the boxed areas in panels a–e. Bars = 50 μm.

Download figure to PowerPoint

Immunostaining was performed to detect the expression of NF-κB/p65 in the spinal dorsal horn. Compared with the control group, rats in the CFA group exhibited significantly higher NF-κB/p65 expression and nuclear translocation, as demonstrated by the intense red fluorescence localized in the cell nuclei. Moreover, spinally delivered LV-shNF-κB/p65 (on day 3 or day 10 following CFA injection) significantly decreased the expression of NF-κB/p65 in the spinal dorsal horn (Figure 1C).

To determine which cell types express NF-κB/ p65 in rats with AIA, double staining was performed using anti–NF-κB/p65 antibodies and antibodies specific for neurons (NeuN), astrocytes (GFAP), or microglia (OX-42). NF-κB/p65 was demonstrated to localize in neurons and astrocytes, as it colocalized with the neuronal marker NeuN and the astrocytic marker GFAP, respectively (Figure 2A, parts a–f, j, and k). In contrast, NF-κB/p65 did not colocalize with the microglial marker OX-42 (Figure 2A, parts g–i and l), indicating that NF-κB/p65 was not expressed in microglia.

image

Figure 2. Cellular localization of NF-κB/p65 in the rat spinal dorsal horn, and effects of spinally delivered lentiviral vector LV-shNF-κB/p65 on NF-κB DNA binding activity. A, Cell type–specific immunolabeling of NF-κB/p65 in the spinal dorsal horn tissue shows colocalization of NF-κB/p65 immunoreactivity (in red) with neuronal-specific nuclear protein [NeuN]–immunoreactive neurons (in green; panels a–c) and glial fibrillary acidic protein (GFAP)–immunoreactive astrocytes (in green; panels d–f), but not with OX-42–immunoreactive microglia (in green; panels g–i). Panels j, k, and l are higher-power magnifications of panels c, f, and i, respectively. Arrows indicate colocalization of NF-κB/p65 with the respective cell markers (in yellow). In panels a–i, bars = 50 μm. In panels j–l, bars = 25 μm. B and C, The spinal cord tissue was evaluated by electrophoretic mobility shift assay (EMSA) for NF-κB/p65 expression in the nuclear extracts (B), and levels of NF-κB DNA binding activity, as assessed in the EMSAs, were determined semiquantitatively in each group (C). In C, the control value is set at 1.0. Values are the mean ± SEM of 4 rats per group. ## = P < 0.01 versus control; ∗∗ = P < 0.01 versus Freund's complete adjuvant (CFA) alone or day 3 CFA + negative control lentiviral vector (LV-NC); ▵▵ = P < 0.01 versus day 10 CFA + LV-NC.

Download figure to PowerPoint

EMSA was performed to evaluate NF-κB DNA binding activity (Figures 2B and C). The DNA binding activity of NF-κB differed significantly between the groups (F[4,15] = 14.945, P < 0.01). It was markedly increased in the CFA group compared with the control group (P < 0.01). In addition, the d3 CFA + LV-shNF-κB/p65-1 and p65-2 groups showed a trend toward less DNA binding activity of NF-κB when compared with the d3 CFA + LV-NC group (P < 0.01 for both comparisons). Similarly, the DNA binding activity of NF-κB in the d10 CFA + LV-shNF-κB/p65-1 and p65-2 groups was also obviously down-regulated when compared with that in the d10 CFA + LV-NC group (F[2,9] = 14.480, P < 0.01 overall and for each comparison).

Effects of spinally delivered LV-shNF-κB/p65 on mechanical and thermal hyperalgesia

The paw withdrawal thresholds in response to mechanical stimuli decreased significantly (P < 0.01) from day 10 after CFA treatment to the last time point assessed (day 20), whereas no trend toward a differential response could be found between the CFA group and the d3 CFA + LV-NC group. Moreover, the mechanical hyperalgesia of the animals in the d3 CFA + LV-shNF-κB/p65-1 and p65-2 groups was significantly higher than that of the animals in the d3 CFA + LV-NC group or CFA group (each P < 0.01). Compared with the d10 CFA + LV-NC group, the CFA + LV-shNF-κB/p65-1 and p65-2 groups treated on day 10 showed an obvious increase in paw withdrawal thresholds from day 10 to day 20 (each P < 0.01) (Figure 3A).

image

Figure 3. Analgesic effects of spinally delivered lentiviral vector LV-shNF-κB/p65 in rats with adjuvant-induced arthritis (AIA). The paw withdrawal threshold (A) and paw withdrawal latency (B) were measured in the rats 1 day before AIA induction (baseline [BL]) and on days 6, 8, 10, 12, 14, 16, 18, and 20 after injection of Freund's complete adjuvant (CFA). Day 0 is the time point before CFA injection. Values are the mean ± SEM of 6–10 rats per group. ## = P < 0.01 versus control; ∗∗ = P < 0.01 versus CFA alone or day 3 CFA + negative control lentiviral vector (LV-NC); ▵▵ = P < 0.01 versus day 10 CFA + LV-NC.

Download figure to PowerPoint

Similarly, injection of CFA resulted in a decrease in the paw withdrawal latency on day 10 after injection (P < 0.01 versus control); this decrease in latency reached a peak on day 12. In addition, the paw withdrawal latencies in the d3 CFA + LV-shNF-κB/p65-1 and p65-2 groups were significantly higher than those in the d3 CFA + LV-NC group or CFA group (each P < 0.01). Similar results were obtained in the groups treated with CFA + LV-shNF-κB/p65-1 or p65-2 on day 10 (each P < 0.01 versus d10 CFA + LV-NC) (Figure 3B).

Effects of spinally delivered LV-shNF-κB/p65 on arthritis

Paw edema

The paw volume was significantly increased in the CFA group compared with the control group (P < 0.01) from day 10 to day 20. Compared with the d3 CFA + LV-NC group, the CFA + LV-shNF-κB/p65-1 and p65-2 groups treated on day 3 showed an obvious decrease in paw edema (each P < 0.01). Similarly, the CFA + LV-shNF-κB/p65-1 and p65-2 groups treated on day 10 also showed a trend toward less paw edema when compared with the d10 CFA + LV-NC group (each P < 0.01) (Figure 4A).

image

Figure 4. Effects of spinally delivered lentiviral vector LV-shNF-κB/p65 on the development of arthritis in rats. A, Paw edema in rats with adjuvant-induced arthritis (AIA) was measured as the change in paw volume, with measurements made 1 day before AIA induction (baseline [BL]) and on days 6, 8, 10, 12, 14, 16, 18, and 20 after injection of Freund's complete adjuvant (CFA). Day 0 is the time point before CFA injection. B, Histopathologic scores of arthritis severity in the rat ankle joints were determined in each treatment group. Symbols represent the histopathologic score in individual rats (n = 5–10 animals/group). ## = P < 0.01 versus control; ∗∗ = P < 0.01 versus CFA alone or day 3 CFA + negative control lentiviral vector (LV-NC); ΔΔ = P < 0.01 versus day 10 CFA + LV-NC. C, Representative photomicrographs of hematoxylin and eosin–stained rat ankle joint sections are shown. Bars = 100 μm. ca = cartilage; syn = synovial tissue.

Download figure to PowerPoint

Joint inflammation and destruction

Histopathologic findings were evaluated on day 20 after CFA injection in the paws of rats with AIA. The extent of massive inflammatory cell infiltration and destruction of the cartilage and joint structures differed significantly between the groups (overall P < 0.01) (Figures 4B and C). Results of the Mann-Whitney U test showed that these histologic features of arthritis were more prominent in the CFA-treated rats compared with the control animals (P < 0.01). Moreover, when compared with the group treated with CFA + LV-NC on day 3, the severity of these changes was significantly lower in the group treated with LV-shNF-κB/p65-2 on day 3 (P < 0.01), whereas these changes were modestly, but not significantly, less severe in the group treated with LV-shNF-κB/p65-1 on day 3 (P = 0.056). Furthermore, the histologic features of arthritis were not statistically significantly different between the d10 CFA + LV-NC group and the d10 LV-shNF-κB/p65-1 or p65-2 groups (P = 0.434 versus d10 CFA + LV-shNF-κB/p65-1 and P = 0.116 versus d10 CFA + LV-shNF-κB/p65-2).

Effects of spinally delivered LV-shNF-κB/p65 on the expression of IL-1β, TNFα, and COX-2 in the spinal cord

We observed that the level of TNFα and IL-1β differed markedly between the groups (for TNFα, F[4,25] = 57.497, P < 0.01; for IL-1β, F[4,25] = 51.374, P < 0.01) (Figures 5A and B). Results of the Bonferroni post hoc test indicated that the expression levels of TNFα and IL-1β were both significantly increased in the CFA group compared with the control group (each P < 0.01). In addition, the IL-1β and TNFα protein levels in the d3 CFA + LV-shNF-κB/p65-1 and p65-2 groups were significantly lower than those in the d3 CFA + LV-NC group (each P < 0.01). Moreover, TNFα expression levels in the d10 CFA + LV-shNF-κB/p65-1 and p65-2 groups were also obviously lower than those in the d10 CFA + LV-NC group (F[2,9] = 7.039, overall P < 0.01 and P < 0.05 for each comparison). Furthermore, when compared with the d10 CFA + LV-NC group, the expression of IL-1β was observed to be significantly lower in the d10 LV-shNF-κB/p65-2 group (P < 0.01), whereas the expression of IL-1β was modestly, but not significantly, lower in the d10 LV-shNF-κB/p65-1 group (F[2,9] = 5.882, overall P < 0.05; P = 0.064 versus d10 CFA + LV-shNF-κB/p65-1 and P < 0.05 versus d10 CFA + LV-shNF-κB/p65-2).

image

Figure 5. Effects of spinally delivered lentiviral vector LV-shNF-κB/p65 on the expression of spinal interleukin-1β (IL-1β), tumor necrosis factor α (TNFα), and cyclooxygenase 2 (COX-2). A and B, Expression levels of spinal IL-1β (A) and TNFα (B) were assessed at the protein level by enzyme-linked immunosorbent assay in each treatment group 14 days after the initiation of antigen-induced arthritis (AIA). C and D, Expression levels of spinal COX-2 were assessed at the protein level by Western blotting 14 days after the initiation of AIA (with β-actin used as a loading control) (C), and the relative density of the COX-2 protein bands was assessed (D). Relative density values are expressed as the ratio of COX-2 to β-actin, and the control is set at 1.0. Values are the mean ± SEM of 4–6 rats per group. ## = P < 0.01 versus control; ∗ = P < 0.05 and ∗∗ = P < 0.01 versus Freund's complete adjuvant (CFA) alone or day 3 CFA + negative control lentiviral vector (LV-NC); ▵▵ = P < 0.01 versus day 10 CFA + LV-NC.

Download figure to PowerPoint

We also evaluated the effect of spinally delivered LV-shNF-κB/p65 on spinal COX-2 expression in rats with AIA. As shown in Figures 5C and D, the cytoplasmic levels of spinal COX-2 protein differed significantly between the groups (F[4,20] = 11.202, overall P < 0.01), and the levels were found to be markedly increased in the CFA group (P < 0.01). Moreover, when compared with the CFA group or the d3 LV-NC group, the d3 LV-shNF-κB/p65-2 group showed a decreased expression of COX-2 protein (each P < 0.05), while the d3 LV-shNF-κB/p65-1 group showed a modestly reduced expression of COX-2 protein, but the difference did not reach statistical significance (P = 0.082 versus CFA and P = 0.096 versus d3 LV-NC). In addition, in the LV-shNF-κB/p65-1 and p65-2 groups treated on day 10, the COX-2 protein levels were also not statistically significantly different when compared with those in the d10 CFA + LV-NC group (F[2,9] = 1.765, overall P = 0.226; P = 0.461 versus d10 CFA + LV-shNF-κB/p65-1 and P = 0.377 versus d10 CFA + LV-shNF-κB/p65-2).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

The aim of the current study was to evaluate the role of spinal NF-κB/p65 in the regulation of peripheral inflammation and hyperalgesia resulting from induction of arthritis in rats. To this end, we knocked down the expression of NF-κB/p65 using LV-shNF-κB/p65, which was administered intrathecally into lumbar segments of the spinal cords of rats. We observed that intrathecal pretreatment with LV-shNF-κB/p65 on day 3 following CFA injection caused a significant reduction in the hyperalgesia, paw edema, and joint destruction in rats with AIA. This indicates that spinal NF-κB/p65 has a pivotal role in the generation of peripheral inflammation and hyperalgesia. Moreover, we also found that intrathecal delivery of LV-shNF-κB/p65 10 days post–CFA injection was able to partially reverse the already-established hyperalgesia and paw edema. This indicates that NF-κB/p65 activation was still ongoing, and that inhibiting the NF-κB/p65 pathway could disrupt the development of painful inflammatory disorders.

Peripheral tissue damage or inflammation induces a series of activation events in the spinal cord. During the development of experimental arthritis, the peripheral nociceptors may be sensitized by inflamed synovium and damaged articular tissue ([27, 28]), and continuous and intense nociceptive input from inflamed joints may induce neural–immune interactions ([12, 29]). This leads to the production and secretion of cytokines, excitatory amino acids, COX-2, and prostaglandins, which can increase the excitability of nociceptive neurons at the spinal level, including the central terminals of the primary sensory afferents (i.e., central sensitization) ([30, 31]). In addition, the spinal cord can also signal to the periphery to regulate inflammation.

A variety of neuronal pathways that modulate peripheral inflammation have been implicated, including the sympathetic and the parasympathetic branches of the autonomic system ([4]). The branches of the vagal nerve and sympathetic fibers innervate immune organs, wherein they can influence peripheral immune responses. Another mechanism, the dorsal root reflex, involves antidromic signaling along somatic afferent fibers that influence peripheral inflammation by releasing neuropeptides, such as vasoactive intestinal peptide, substance P, and calcitonin gene–related peptide, from the sensory nerve endings ([5, 6]). Moreover, the adrenal cortex (the HPA axis), which can also provide a feedback mechanism in the processes of inflammation, is often blunted in a wide range of autoimmune and inflammatory diseases such as rheumatoid arthritis ([7]). Previous studies have supported this notion. For instance, spinal cord MAPK and specific cytokines such as IL-1β and TNFα, which can be regulated by NF-κB, are involved in the regulation of peripheral inflammation ([13, 30, 32]).

The NF-κB family of transcription factors controls the expression of genes that are critical for inflammation and immune activation ([33]). Family members of NF-κB in the articular tissue can play a vital role in the initiation and development of arthritis by the regulation of cytokines, such as IL-1β, IL-6, and TNFα, and the regulation of enzymes involved in tissue remodeling, such as COX-2, iNOS, and matrix metalloproteinase 9 ([34, 35]). Interestingly, NF-κB is also activated in the spinal cord in response to peripheral nerve injury or tissue inflammation. In the rat chronic constriction injury (CCI) model, we previously observed extensive colocalization of NF-κB/p65 with TNFα in the spinal dorsal horn ([36]), and down-regulation of spinal NF-κB/p65 expression significantly attenuated sciatic nerve ligation–induced mechanical and thermal hyperalgesia ([19]). In addition, spinal NF-κB is also activated following injection of CFA or formalin into the footpad, and spinal application of an NF-κB inhibitor reduces pain-related behavior ([18, 37, 38]). Similar to the findings in these studies, the present results show that peripheral inflammation resulting from AIA led to the activation of NF-κB/p65 in the spinal cord, and a spinally delivered inhibitor of NF-κB/p65 expression, the lentiviral vector LV-shNF-κB/p65, significantly attenuated the inflammation-induced hyperalgesia in this rat model of arthritis.

Since spinal NF-κB/p65 has been shown to be activated following peripheral nerve tissue injury or inflammation, and since it is involved in the induction of central sensitization and hyperalgesia, we explored the possibility that spinal NF-κB/p65 regulates peripheral inflammation. In the present study, the spinally delivered inhibitor LV-shNF-κB/p65 markedly decreased the severity of paw edema, inflammatory cell infiltration, and destruction of cartilage and joint structures in rats with AIA. In contrast, spinal delivery of LV-NC had no effect on the paws of arthritic rats. In addition, our findings indicated that NF-κB/p65 was expressed in neurons and astrocytes in the spinal dorsal horn, as observed on day 14 after CFA injection.

A recent study demonstrated that the up-regulation of spinal NF-κB in rats with CCI could be abated by systemic treatment with MK-801 ([39]), an N-methyl-d-aspartate (NMDA) receptor antagonist known to improve central sensitization and hyperalgesia in rats ([40]). NMDA receptor activation may activate NF-κB in neurons ([39, 41, 42]), suggesting that there is some involvement of neuronal NF-κB in central sensitization. On the other hand, previous studies have shown that selective inhibition of NF-κB activity in astrocytes in the spinal cord significantly reduced the extent of hyperalgesia and allodynia after formalin injection ([43]), which suggests a role for NF-κB in the modulation of astrocytes following peripheral inflammation. Therefore, NF-κB up-regulation in the spinal cord after peripheral nerve injury or inflammation may reflect changes in both neuron and astrocyte function.

Moreover, we found that the expression levels of IL-1β, TNFα, and COX-2 protein were highly up-regulated in the spinal cord of rats with AIA. In the family of proinflammatory cytokines, TNFα and IL-1β have long been strongly implicated in the modulation of peripheral pain and inflammation of the joints. In addition to their peripheral action, TNFα and IL-1β are up-regulated in the spinal cord following peripheral inflammation, and both are known to be involved in the development and maintenance of experimental arthritis ([11, 13, 30, 32]). COX-2, which is constitutively expressed in the spinal cord, is a major contributor to the induction of spinal prostaglandin E2 ([44]). It has been shown that peripheral inflammation can up-regulate the expression of COX-2 in the spinal cord ([45]), and spinal administration of the COX inhibitor indomethacin attenuated inflammatory edema ([46]). Our results demonstrate that spinally delivered LV-shNF-κB/p65 markedly reduced the overexpression of spinal TNFα, IL-1β, and COX-2 in rats with AIA. These findings are consistent with those reported previously ([19, 37]).

Interestingly, in addition to being regulated by NF-κB, spinal TNFα, IL-1β, and COX-2 can help to propagate the extension of the neuroinflammatory response by activating NF-κB ([47-49]), which forms a positive feedback mechanism to exaggerate the inflammatory process. Thus, NF-κB activation may amplify/perpetuate the neuroinflammatory responses, which may act directly on the central terminals of primary afferent neurons and on dorsal horn neurons and thereby contribute to central sensitization and hyperalgesia. We therefore propose that the analgesic and antiinflammatory effects of LV-shNF-κB/p65 might be mediated, at least in part, through the prevention of neuronal and astrocytic NF-κB/p65 and subsequent suppression of the positive feed-forward loop between NF-κB/p65 and inflammatory mediators such as TNFα, IL-1β, and COX-2.

In conclusion, our findings indicate a pivotal role for spinal NF-κB/p65 in the initiation and development of both peripheral inflammation and inflammation-related hyperalgesia. Reduction in the expression of NF-κB/p65 protein in the spinal neurons and astrocytes markedly reduced both phenomena. Thus, interference with NF-κB/p65 on the spinal level may provide a novel treatment option for painful inflammatory disorders.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Drs. Sun and Fu had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Luo, Sun, Fu.

Acquisition of data. Luo, Xu, Wang, Lin.

Analysis and interpretation of data. Luo, X.-L. Zhao, X.-J. Zhao.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES