- Email: [email protected]
Serial Determination of Tumor Necrosis Factor-Alpha Content in Rat Sciatic Nerve after Chronic Constriction Injury
Experimental Neurology 160, 124–132 (1999) Article ID exnr.1999.7193, available online at http://www.idealibrary.com on
Serial Determination of Tumor Necrosis Factor-Alpha Content in Rat Sciatic Nerve after Chronic Constriction Injury Annette George, Christine Schmidt, Andreas Weishaupt, Klaus V. Toyka, and Claudia Sommer Neurologische Universita¨tsklinik Wu¨rzburg, Germany Received December 11, 1997; accepted July 20, 1999
Wallerian degeneration, induced after injury to a peripheral nerve, is associated with upregulation of proinflammatory cytokines, which are suggested to contribute to the development of lesion-induced neuropathic pain. In chronic constrictive injury (CCI), an animal model of injury-induced painful mononeuropathy, inhibition of synthesis, release, or function of the cytokine tumor necrosis factor-␣ (TNF) results in reduced pain-associated behavior. Here, changes of TNF content in rat sciatic nerves after CCI (days 0, 0.5, 1, 3, 7 and 14) were investigated by enzyme-linked-immunoassay. Low levels of TNF were already detectable in control nerves. Concentrations increased rapidly after CCI, with a maximum (2.7-fold) at 12 h, and remained elevated on a lower level until day 3. Baseline levels were reached again at day 14. These results indicate that TNF is produced at an early time point in the cascade of events resulting in Wallerian degeneration and hyperalgesia following peripheral nerve injury. Given that only prophylactic treatment with TNF inhibitors efficiently reduces hyperalgesia in CCI, TNF seems to contribute to the initiation of neuropathic pain in this model. r 1999 Academic Press Key Words: Wallerian degeneration; tumor necrosis factor-␣ (TNF); enzyme-linked immunoassay (ELISA); tissue homogenization; sedimental fraction; membrane TNF.
INTRODUCTION
The recent years of cytokine research in the central nervous system (CNS) have shown that the actions of the ‘‘proinflammatory’’ cytokine tumor necrosis factor-␣ (TNF, syn. TNF-␣) are not confined only to typical inflammatory or autoimmune diseases, but that TNF is also involved in degenerative, metabolic, ischemic, or traumatic lesions. This has led to new therapeutic approaches (22). Similarly, in the peripheral nervous system (PNS), TNF is not only known as a mediator in inflammatory neuropathies like the Guillain-Barre´Syndrom (29), leprosy (16), HIV neuropathy (13), neuroborreliosis (10), or in the animal model of experimen0014-4886/99 $30.00 Copyright r 1999 by Academic Press All rights of reproduction in any form reserved.
tal autoimmune neuritis (EAN) (37, 46). TNF is also involved in Wallerian degeneration and in regeneration processes after peripheral nerve injury. This has been shown in animal models of complete (transection) (37) or incomplete sciatic nerve lesion (crush, chronic constriction injury, CCI) (17, 42, 43). Injury to a peripheral nerve is often associated with the development of chronic neuropathic pain, refractory to standard analgesic drugs. It has been hypothesized that local cytokine actions, produced by invading immune cells and/or local resident cells which proliferate and become immunologically active after injury (‘‘periaxonal inflammatory processes’’) (5) may contribute to the generation and maintanance of such pain states (39). The key role of TNF and interleukin (IL)-1 as mediators in inflammatory pain is well documented (6, 9). A better understanding of the action of these cytokines after peripheral nerve injury might offer a new tool for managing neuropathic pain symptoms after nerve lesion (7, 33). Wells et al. described an increase of circulating TNF levels after unilateral crush lesion in the rat already in 1992 (43). An induction of TNFmRNA in the nerve after crush was recently described by La Fleur et al. using semiquantitative RT-PCR (17) and in CCI by in situ hybridization (ISH) (42). However, the expression of TNF is regulated not only transcriptionally but also on a translational and posttranslational level (1). Therefore, knowledge of actual local tissue protein concentrations is essential in order to further characterize a possible involvement of TNF in Wallerian degeneration. Furthermore, TNF can produce even contrasting effects depending on its site of action (21, 44). Using immunohistochemistry (IHC), an increase of TNFimmunoreactive (IR) protein was observed after transection (37), and CCI (42). Rotshenker and coworkers were able to detect various cytokines (IL-1, IL-6, GM-CSF) by ELISA in the cultured media of segments of lesioned nerve in vitro (24, 25, 26), but could not detect TNF. However, failure of TNF protein detection in lesioned peripheral nerve may perhaps be rather due to methodological difficulties than to absence of the protein. Therefore we decided to investigate endoneurial
124
TNF IN PERIPHERAL NERVE INJURY
changes of TNF in a behaviorally and morphologically well characterized animal model of painful mononeuropathy, the chronic constriction injury (CCI) (2, 30, 32), using an adapted extraction procedure and a more sensitive enzyme-linked immunoassay (ELISA) system. MATERIALS AND METHODS
Animals and Surgery All experiments were approved by the Bavarian State animal experimentation committee and carried out in accordance with its regulations. Female Sprague– Dawley rats (200–250 g), obtained from Charles-River, Germany, were housed in plastic cages under natural lighting conditions with access to food and water ad libitum. Chronic constrictive injury was performed as described by Bennett and Xie (2) with minor modifications. Under barbiturate anesthesia, the sciatic nerve was exposed unilaterally at the midthigh level. Three ligatures (chromic gut 4.O) were placed around the nerve with 1 mm spacing, and tied until they just slightly constricted the diameter of the nerve and a short twitch was seen in the respective hindlimb. Sham operations, where the nerve was exposed and mobilized, were performed contralaterally. Animals were sacrified on days 0.5, 1, 3, 7, and 14 post surgery (n ⫽ 12 per time point). Sciatic nerves were removed by cutting the nerve shortly above the site of the constrictive ligatures and 1.5 cm distally. Nerves from the sham-operated sides were removed accordingly. Length and wet weight of the nerve segments were determined. Nerves were immediately frozen on dry ice and stored at ⫺80°C. Control nerves (day 0) were removed from healthy animals.
125
well as membrane-bound TNF (Biosource International, pers. commun.). Recovery assays of spiked tissue samples were performed (n ⫽ 12). Standard curves in the standard buffer were compared to standard curves in nerve homogenates diluted in standard buffer (n ⫽ 3). The mean absorbance of the ‘‘zero standard’’ (buffer or nerve homogenate, respectively) was substracted from all data prior to plotting. Positive and negative controls were included in each assay. TNF concentration was expressed as pg/mg protein. Protein content was determined by the bicinchoninic acid protein assay reagent (Pierce, KMF Laborchemie, St. Augustin, Germany). The relative distribution of endogenous TNF between supernatant (S) and pellet (P) was expressed in percentages (pg TNF/ml found in pellet referred to the sum of pg TNF/ml found in the supernatant and pellet fraction ⫻ 100) (Fig. 2a). This form of presentation was used because it was independent of the absolute concentration of TNF shown in Fig. 3. Recovery of exogenous TNF (Biosource International, CA), added during the homogenization procedure, was investigated in normal (day 0) and injured (day 7) tissue: Rat sciatic nerves were pooled and minced under liquid nitrogen. The powder was dissolved in homogenization buffer. TNF standard (1 ng/ml) was added to one aliquot of this homogenous material (containing equal amounts of endogenously contained TNF levels) and vortexed thoroughly. Subsequently, samples (containing or lacking exogenous TNF standard) were treated as described above. Recovery of exogenous TNF was expressed as the percentage of the known amount of TNF that had been added (1 ng/ml) after subtraction of the concentration found for endogenous TNF in an aliquot treated accordingly in the absence of TNF standard.
Homogenization Procedure and ELISA Determination Sciatic nerves were pooled (n ⫽ 4) and homogenized in ice-cold phosphate-buffered saline (PBS), pH 7.4, containing protease inhibitors (aprotinin, leupeptin, pepstatin, PMSF; Boehringer Mannheim, Germany). After centrifugation at 10,000g at ⫹4°C for 10 min, supernatant (S) was removed. The remaining pellet (P) was rehomogenized in the original volume of homogenization buffer. Triton X-100 (Boehringer Mannheim, Germany) was added to S and P at a final concentration of 0.01%. The samples were vortexed thoroughly and centrifuged again. Supernatants were aliquoted and assayed in duplicate (after dilution in the standard buffer supplied) by the Cytoscreen Rat TNF-␣ Ultrasensitive ELISA kit (Biosource International, CA) according to the manufacturer’s instructions. This assay system detects rat TNF with a sensitivity of 0.7 pg/ml and also in the presence of soluble TNF-receptor-1. The monoclonal antibody for rat TNF recognizes soluble as
Western Blot Analysis Tissue homogenates (supernatant, pellet) of injured rat sciatic nerves (days 0.5 and 14 post-CCI) were separated on a 15% SDS–polyacrylamide gel and transferred to nitrocellulose membrane (Schleicher & Schu¨ll, Dassel, Germany). The Western blot was blocked for 2 h at 4°C and then incubated overnight with polyclonal sheep antiserum to mouse TNF. Goat anti-sheep IgG horseradish peroxidase conjugate (Dianova, Hamburg, Germany) was used as second antibody. The blots were developed with a 3,38-diaminobenzidine/H2O2 solution. Histology One additional animal with CCI was sacrificed at each time point used for ELISA in order to verify the lesion (plastic embedding, 1-µm sections, toluidine blue) (30) and to visualize TNF immunoreactivity (fro-
126
GEORGE ET AL.
zen sections). Immunohistochemistry was performed with a polyclonal antibody to TNF-␣ (1:1000, Genzyme, Ru¨sselsheim, Germany). Indirect immunogold silver staining method (Biotrend, Ko¨ln, Germany) was used for detection. Statistical Analysis Data represent mean values (from at least three independent experiments) ⫾ standard deviation. Significance of differences between groups was analyzed by two-tailed Student’s t test (Fig. 2). Analysis of variance (ANOVA) was used to test for differences between groups and subsequent Scheffe´ test for comparison of individual means throughout the time course (Fig. 3). Statistical significance was assumed with P ⬍ 0.05. RESULTS
Adapting an ELISA for TNF Determination in Rat Sciatic Nerve Homogenates The standard curve for rat TNF standard serially diluted in nerve homogenate lacking endogenous TNF was comparable to the standard curve in the standard buffer supplied (Fig. 1). Recovery of rat TNF standard added to nerve homogenate samples was 103 ⫾ 6% (n ⫽ 12). Relative Distribution of TNF between Supernatant and Pellet Fraction in Control Versus Injured Rat Sciatic Nerve Homogenates (a) Endogenously contained TNF. In control sciatic nerve homogenates, 88 ⫾ 1.7% of the TNF found was
detected in the supernatant, 12% in the supernatant of the rehomogenized pellet (Fig. 2a). Detection of TNF in the pellet was not due to residues of an uncompletely removed supernatant, because washing of the remaining sedimental fraction (instead of rehomogenization) failed to reach measurable TNF concentrations (data not shown). After lesion of the sciatic nerve by CCI, an increase of the amount of TNF found in the pellet fraction could be observed (Fig. 2a), which became statistically significant already on day 1 (81 ⫾ 1.6% in supernatant and 19% in pellet, P ⱕ 0.01) with further increase until day 7 (70 ⫾ 1.7 and 30%, P ⱕ 0.01). This change in the relative distribution of endogenous TNF between supernatant and pellet was independent of the absolute TNF levels found (see Fig. 3). In sham-operated sciatic nerves no significant changes in the relative amount of endogenous TNF found in the pellet fraction were observed (with 10 ⫾ 2.3 and 12 ⫾ 2.8% on days 0.5 and 7, respectively). (b) Exogenously added TNF standard. In control sciatic nerve homogenates (n ⫽ 5), 90.1 ⫾ 7.5% of exogenously added TNF standard (1 ng/ml homogenate) could be recovered: 84.4 ⫾ 9.5% in the supernatant, 5.7% in the supernatant of the rehomogenized pellet fraction (Fig. 2b). In sciatic nerve homogenates after lesion by CCI (day 7; n ⫽ 4), the recovery of exogenous TNF standard decreased to 43.3 ⫾ 5.2% (P ⱕ 0.01) in the supernatant; no TNF standard was detectable in the pellet fraction.
FIG. 1. Representative standard curve for rat TNF standard serially diluted in standard buffer (diamonds) and nerve homogenate lacking endogenous TNF (squares). Homogenization procedure as described under Material and Methods, corrected for background by substracting mean absorbance of zero standard (buffer or homogenate, respectively) from all data; range, 2.3–150 pg/ml.
TNF IN PERIPHERAL NERVE INJURY
127
FIG. 2. (a) Relative distribution of endogenous TNF in the supernatant (S) vs pellet (P) fraction after centrifugation of control (day 0) and CCI operated rat sciatic nerve homogenates (days 0.5, 1, 3, 7, 14). Extraction procedure as described under Materials and Methods. Data represent mean values of at least three experiments, expressed in % ⫾ SD; **P ⱕ 0.01. Note that the relative amount of endogenous TNF found in the pellet fraction increases postlesion independent of the absolute concentration of TNF shown in Fig. 3. (b) Recovery of exogenously added TNF (1 ng/ml) in the supernatant (S) vs pellet (P) fraction after centrifugation of control (day 0) and CCI-operated rat sciatic nerve homogenates (day 7). Extraction procedure as described under Materials and Methods. Data represent mean values of at least three experiments, expressed in % ⫾ SD; **P ⱕ 0.01.
Time Course of TNF Content in Rat Sciatic Nerve after CCI The TNF content in control rat sciatic nerve was 40.6 ⫾ 2 pg/mg protein. After CCI (Fig. 3), a rapid increase of TNF levels was observed, reaching its maximum (110.2 ⫾ 8.9 vs 53.7 ⫾ 13.7 pg/mg protein in CCI vs sham-operated side, respectively; P ⱕ 0.01)
already after 12 h. TNF levels declined afterward (62.8 ⫾ 10.4 and 52.3 ⫾ 9.7 pg/mg protein on day 1; P ⬎ 0.05) but still remained elevated significantly until day 3 (51.5 ⫾ 5.2 and 35.4 ⫾ 3.3 pg/mg protein CCI vs sham-operated side, respectively; P ⱕ 0.05). Baseline levels were reached again on day 14 (41.1 ⫾ 7.2 vs 44.1 ⫾ 7.2 pg/mg protein). On days 0.5 and 1, a slight,
128
GEORGE ET AL.
FIG. 3. Time course of TNF content in rat sciatic nerve after CCI. Endogenous TNF levels in rat sciatic nerve homogenates of control (day 0), CCI-operated (triangles), and sham-operated (squares) nerves (days 0.5, 1, 3, 7, 14); n ⫽ 12 animals per time point. Extraction procedure as described under Materials and Methods. TNF in supernatant and pellet fraction was determined by ELISA and data presented are the sum of TNF levels in the supernatant and the pellet fraction. Levels are expressed as mean values (pg/mg protein) ⫾ SD; n ⫽ 3; *P ⱕ 0.05, **P ⱕ 0.01 (CCI vs sham-operated side).
but statistically not significant increase of TNF levels was found in sham-operated nerves (53.7 ⫾ 13.7 and 52.3 ⫾ 9.7 pg/mg TNF on days 0.5 and 1, respectively, in sham-operated vs 40.6 ⫾ 2 pg/mg TNF in control sciatic nerves; P ⬎ 0.05). Histology Analysis of semithin sections confirmed Wallerian degeneration on the injured side (Fig. 4a). Immunohistochemistry for TNF revealed increased TNF-IR in nerve tissue on the side of the lesion (Figs. 4b and 4c). Molecular Forms of TNF Western blot analysis confirmed the presence of proteins with molecular weights of 17, 26, 34, and 51 kDa in the rat sciatic nerve homogenates (day 0.5 post-CCI) investigated by ELISA (Fig. 5). This is consistent with published molecular weights for the soluble, transmembrane, dimeric, or trimeric form of the TNF molecule, respectively (1). The transmembrane molecule was more prominent in the sedimental fraction compared to the supernatant; however, we did not detect a definite increase of the transmembrane form in the pellet fraction of homogenates of injured nerves harvested on day 14 post-CCI (data not shown). DISCUSSION
Our sensitive assay system allowed us to detect low (physiologic) TNF concentrations in rat sciatic nerve
and to follow its upregulation under pathological conditions such as chronic constrictive injury with subsequent Wallerian degeneration. Quantitative analysis of tissue TNF is critically dependent on the extraction procedure used (11). Here, we observed that it may also depend on the normal or pathologic state of the tissue investigated. Endogenous TNF may be lost to the pellet phase separated after centrifugation. Indeed, when nerve growth factor (NGF) in homogenized nervous tissue was analyzed by ELISA, a several-fold loss of NGF to the pellet fraction was found (14). Since this enormous loss of NGF may be due to the high centrifugation speed (100,000g for 1 h) (45), we used a lower centrifugation speed (10,000g), resulting in only low levels of endogenous as well as exogenous TNF detected in the pellet fraction. However, the relative amount of endogenous TNF recaptured from the pellet was increased after CCI. According to our Western blot analysis this seems probably not (or only to a minor degree) due to a shift in relation from the soluble (17 kDa) to the membrane-associated (26 kDa) form of TNF molecule in situ, as it has been described in a study on TNF in abnormal vessel walls (20). Other factors contributing to this observation may be (1) increased volume of the sedimental fraction, e.g., due to invading cells (but volume of homogenization buffer was always adapted to the wet weight of tissue investigated such that equal concentrations of homogenized tissue were achieved); (2) masking or increased degradation of TNF by mediators in pathologic tissue.
TNF IN PERIPHERAL NERVE INJURY
129
FIG. 4. (a) Semithin section, toluidine-blue stained, of CCI nerve, 10 mm distal from lesion. At 14 days postsurgery only few myelinated fibers remain (arrows). (b, c) Immunohistochemistry on frozen sections of (b) sham-operated sciatic nerve, (c) CCI nerve 7 days postsurgery. TNF immunoreactivity is weakly positive in occasional endoneurial cells (presumably Schwann cells) in sham-operated nerve (arrow) and abundant in endoneural cells (arrows) of the injured nerve. Bar ⫽ 20 µm for all photomicrographs.
Recovery of exogenously added TNF was reduced in injured nerve tissue. This may reflect degradation or complexing of soluble TNF during the homogenization procedure. A higher turnover and degradation rate of TNF at the site of local inflammatory tissue reaction has indeed been assumed by Khanolkar-Young comparing localization of TNF protein by IHC and TNFmRNA by ISH in nerve biopsies of leprosy patients (16). Therefore, actual concentrations of TNF in damaged tissue may be underestimated.
regions after sham operation (1.6–10 pg/mg for hippocampus, approx. 25–50 pg/mg protein for striatum and thalamus) (28). By other authors, lower (4, 20, 36, 40) or even undetectable (12, 18, 38) levels of TNF have been described in various nonnervous tissues in human or animal studies. Differences in species, type of organ investigated, procedure of tissue disruption, homogenization and extraction, and the sensitivity of the ELISA system may account for the different levels found.
Physiological Levels of TNF in the Rat Sciatic Nerve
Time Course of TNF Content in Rat Sciatic Nerve after Chronic Constriction Injury
TNF tissue concentrations found in normal rat sciatic nerve (40.6 pg/mg protein) are in the range described for other normal tissues in mice (between 17 and 38 pg/mg protein for muscle and lung, respectively) (15) and are also comparable to TNF levels found in human brain (18 pg/mg protein) (19) and in gerbil brain
After CCI, a rapid and early increase of endoneurial TNF was followed by a plateau phase returning to baseline levels by day 14. The fact that the maximum of the TNF increase after CCI occurs already 12 h postlesion or even earlier, suggests that resident cells become activated after
130
GEORGE ET AL.
FIG. 5. Western blot analysis confirms the presence of the soluble (17 kDa), transmembrane (26 kDa), dimeric (34 kDa), and trimeric (51 kDa) form of TNF molecule in the rat sciatic nerve homogenates (day 0.5 post-CCI; lanes 1 and 2).
injury and may be the main source of secreted TNF rather than hematogenous macrophages, which transmigrate not before day 3 in CCI (31). Hitherto available ISH studies have localized TNFmRNA in Schwann cells, fibroblasts, endothelial cells, and macrophages on day 7 post-CCI (42). La Fleur et al. (17) identified macrophages and Schwann cells as producers of TNFmRNA on day 4 after a crush lesion in the distal segment of the mouse sciatic nerve. Earlier time points regarding the cellular source are not available; but these results suggest that proliferating Schwann cells and resident macrophages synthesize and secrete TNF at this early time point after peripheral nerve injury. The time course of TNF protein determined by ELISA corresponds to the elevation of TNF-IR found in the CCI model using IHC (Marziniak and Sommer, unpublished observations) as well as to TNF mRNA changes reported recently after crush lesion of the mouse sciatic nerve (17). Interestingly, an early, sharp increase of TNF mRNA on day 1 was observed in the segment of crush lesion while a delayed elevation of longer duration was found in the segment distal to lesion. Elevation of endoneurial TNF levels parallels the initial structural changes during Wallerian degenera-
tion after CCI (30, 31): appearance of endoneurial edema, proliferation of Schwann- and endothelial cells, and axonal degeneration with almost complete loss of myelinated axons by day 7. First signs of regeneration (axonal sprouting, remyelination) are known to occur at time points of decreasing TNF levels. Whether there is a causal link between the rise in TNF and axonal damage in CCI, remains to be investigated. After endoneurial application of TNF to the rat sciatic nerve edema, axonal degeneration, demyelination, and vascular changes have been observed (23, 27). Reduction of endoneurial TNF production was associated with attenuated CCI-induced vascular changes of endoneurial vessels (32). On the other hand, reduced axonal damage and an improved recovery of motor function after crush lesion of rat sciatic nerve were reported when TNF was applied intraperitoneally prior to nerve lesion (3). Experimental administration of TNF into normal rat sciatic nerve induces pain-associated behavior (41). In CCI, the maximum of endoneurial TNF increase precedes the development of pain-associated behavior (30): thermal hyperalgesia and mechanical allodynia develop by days 1 to 3, with the maximum of thermal hyperalgesia between days 6 and 14. TNF levels declined subsequently and baseline levels were reached again on day 14, when hyperalgesia was still present. Administration of inhibitors of TNF synthesis, release, or function results in a reduction of pain-associated behavior in CCI only, when the inhibitor is given prophylactically but not when treatment is started 1 week after CCI (33–35). The early maximum of TNF elevation found may explain these results: TNF may contribute to the initiation of pain-associated behavior after nerve lesion, which is consistent with the role of TNF in initiation of the cytokine cascade involving sequentially IL-1, IL-6, and IL-8 (8). Pain-associated behavior may be maintained by the cytokine cascade induced by TNF involving IL-1, IL-6, and IL-8. Selective immunopharmacological blockade of these cytokines may help to elucidate their role in the maintanance of neuropathic pain states. ACKNOWLEDGMENTS Supported by Deutsche Forschungsgemeinschaft So 328/2-1. Presented in part at the Annual Meeting of the Society for Neuroscience, New Orleans, LA, October 25, 1997.
REFERENCES 1. Aggarwal, B. B., and K. Natarajan. 1996. Tumor necrosis factors: developments during the last decade. Eur. Cytokine Netw. 7: 93–124. 2. Bennett, G. J., and Y. K. Xie. 1988. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain. 33: 87–107. 3. Chen, L. E., A. V. Seaber, G. H. Wong, and J. R. Urbaniak. 1996.
TNF IN PERIPHERAL NERVE INJURY
4.
5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Tumor necrosis factor promotes motor functional recovery in crushed peripheral nerve. Neurochem. Int. 29: 197–203. Clancy, K. D., K. Lorenz, E. Hahn, B. Christiansen, C. Hofmann, and R. L. Gamelli. 1997. Down-regulation of tissue specific tumor necrosis factor-alpha in the liver and lung after burn injury and endotoxemia. J. Trauma. 42: 169–176. Clatworthy, A. L., P. A. Illich, G. A. Castro, and E. T. Walters. 1995. Role of peri-axonal inflammation in the development of thermal hyperalgesia and guarding behavior in a rat model of neuropathic pain. Neurosci. Lett. 184: 5–8. Cunha, F. Q., S. Poole, B. B. Lorenzetti, and S. H. Ferreira. 1992. The pivotal role of tumour necrosis factor alpha in the development of inflammatory hyperalgesia. Br. J. Pharmacol. 107: 660–664. DeLeo, J., R. Colburn, and A. Rickman. 1997. Cytokine and growth factor immunohistochemical spinal profiles in two animal models of mononeuropathy. Brain Res. 759: 50–57. Eugui, E., B. DeLustro, S. Rouhafza, R. Wilhelm, and A. Allison. 1997. Coordinate inhibition by some antioxidants of TNF, IL-1, and IL-6 production by human peripheral blood mononuclear cell. Ann. NY Acad. Sci. 171–184. Ferreira, S. H. 1993. The role of interleukins and nitric oxide in the mediation of inflammatory pain and its control by peripheral analgesics. Drugs. 46: 1–9. Garcia-Monco, J. C., and J. L. Benach. 1997. Mechanisms of injury in Lyme neuroborreliosis. Semin. Neurol. 17: 57–62. George, A., C. Schmidt, and C. Sommer. 1998. Quantification of TNF protein in homogenized rat sciatic nerve by ELISA. Soc. Neurosci. Abs. 24: 1539. Gionchetti, P., M. Campieri, A. Belluzzi, E. Bertinelli, M. Ferretti, C. Brignola, G. Poggioli, M. Miglioli, and L. Barbara. 1994. Mucosal concentrations of interleukin-1 beta, interleukin-6, interleukin-8, and tumor necrosis factor-alpha in pelvic ileal pouches. Dig. Dis. Sci. 39: 1525–1531. Griffin, J. W., S. L. Wesselingh, D. E. Griffin, J. D. Glass, and J. C. McArthur. 1994. In HIV, AIDS and the Brain (Q. W. Price and S. W. Perry, Eds.), pp. 159–182. Raven Press, New York. Hoener, M., E. Hewitt, J. Conner, J. Costello, and S. Varon. 1996. Nerve growth factor content in adult rat brain tissues is several-fold higher than generally reported and is largely associated with sedimentable fractions. Brain Res. 728: 47–56. Hotchkiss, R., D. Osborne, G. Lappas, and I. Karl. 1995. Calcium antagonists decrease plasma and tissue concentrations of tumor necrosis factor, interleukin-1 alpha and beta in amouse model of endotoxin. Shock 3: 337–342. Khanolkar-Young, S., N. Rayment, P. M. Brickell, D. R. Katz, S. Vinayakumar, M. J. Colston, and D. N. Lockwood. 1995. Tumour necrosis factor-alpha synthesis is associated with the skin and peripheral nerve pathology of leprosy reversal reactions. Clin. Exp. Immunol. 99: 196–202. La Fleur, M., J. L. Underwood, D. A. Rappolee, and Z. Werb. 1996. Basement membrane and repair of injury to peripheral nerve: Defining a potential role for macrophages, matrix metalloproteinases, and tissue inhibitor of metalloproteinases-1. J. Exp. Med. 184: 2311–2326. MacMaster, M., R. Newton, K. Sudhansu, and G. Andrews. 1992. Activation and distribution of inflammatory cells in the mouse uterus during the preimplantation period. J. Immunol. 148: 1699–1705. Mogi, M., M. Harada, P. Riederer, H. Narabayashi, K. Fujita, and T. Nagatsu. 1994. TNF alpha increases both in the brain and in the cerebrospinal fluid from parkinsonian patients. Neurosci. Lett. 165: 208–219. Newman, K. M., J. Jean-Claude, H. Li, W. G. Ramey, and M. D.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32. 33.
34.
35.
36.
37.
38.
131
Tilson. 1994. Cytokines that activate proteolysis are increased in abdominal aortic aneurysms. Circulation. 90: II224–227. Paya, C., P. Leibson, A. Patick, and M. Rodriguez. 1990. Inhibition of Theiler’s virus-induced demyelination in vivo by TNF. Int. Immunol. 2: 909–913. Probert, L. K. Selmaj. 1997. TNF and related molecules: Trends in neuroscience and clinical applications. J. Neuroimmunol. 72: 113–117. Redford, E. J., S. M. Hall, and K. J. Smith. 1995. Vascular changes and demyelination induced by the intraneural injection of tumour necrosis factor. Brain 118: 869–878. Reichert, F., R. Levitzky, and S. Rotshenker. 1996. Interleukin-6 in intact and injured mouse peripheral nerves. Eur. J. Neurosci. 8: 530–535. Rotshenker, S., S. Aamar, and V. Barak. 1992. Interleukin-1 activity in lesioned peripheral nerve. J. Neuroimmunol. 39: 75–80. Saada, A., F. Reichert, and S. Rotshenker. 1996. Granulocyte macrophage colony stimulating factor produced in lesioned peripheral nerves induces the up-regulation of cell surface expression of MAC-2 by macrophages and schwann cells. J. Cell Biol. 133: 159–167. Said, G. M., and M. Hontebeyrie-Joskowicz. 1992. Nerve lesions induced by macrophage activation. Res. Immunol. 143: 589– 599. Saito, K., K. Suyama, K. Nishida, Y. Sei, and A. S. Basile. 1996. Early increases in TNF-alpha, IL-6 and IL-1 beta levels following transient cerebral ischemia in gerbil brain. Neurosci. Lett. 206: 149–152. Sharief, M., B. McLean, and E. Thompson. 1993. Elevated serum levels of TNF in Guillain-Barre-Syndrome. Ann. Neurol. 33: 591–596. Sommer, C., J. A. Galbraith, H. M. Heckman, and R. R. Myers. 1993. Pathology of experimental compression neuropathy producing hyperesthesia. J. Neuropathol. Exp. Neurol. 52: 223– 233. Sommer, C., A. Lalonde, H. M. Heckman, M. Rodriguez, and R. R. Myers. 1995. Quantitative neuropathology of a focal nerve injury causing hyperalgesia. J. Neuropathol. Exp. Neurol. 54: 635–643. Sommer, C., and R. R. Myers. 1996. Vascular changes in a model of painful neuropathy. Exp. Neurol. 141: 113–119. Sommer, C., C. Schmidt, and A. George. 1998. Hyperalgesia in experimental neuropathy is dependent on the TNF receptor 1. Exp. Neurol. 151: 138–142. Sommer, C., M. Marziniak, and R. R. Myers. 1998. The effect of thalidomide treatment on vascular pathology and hyperalgesia caused by chronic constrictive injury of rat sciatic nerve. Pain 74: 83–91. Sommer, C., C. Schmidt, A. George, and K. Toyka. 1997. A metalloprotease-inhibitor reduces pain associated behavior in mice with experimental neuropathy. Neurosci. Lett. 237: 45–48. Stashenko, P., J. J. Jandinski, P. Fujiyoshi, J. Rynar, and S. S. Socransky. 1991. Tissue levels of bone resorptive cytokines in periodontal disease. J. Periodontol. 62: 504–509. Stoll, G., S. Jung, S. Jander, P. van der Meide, and H. P. Hartung. 1993. Tumor necrosis factor-alpha in immunemediated demyelination and Wallerian degeneration of the rat peripheral nervous system. J. Neuroimmunol. 45: 175–182. Takematsu, H., H. Ohta, and H. Tagami. 1989. Absence of tumor necrosis factor-alpha in suction blister fluids and stratum corneum from patients with psoriasis. Arch. Dermatol. Res. 281: 398–400.
132 39. 40.
41.
42.
43.
GEORGE ET AL. Tracey, D. J., and J. S. Walker. 1995. Pain due to nerve damage: Are inflammatory mediators involved? Inflamm. Res. 44: 407–411. von Bieberstein, S., J. Spiro, R. Lindquist, and D. Kreutzer. 1995. Enhanced tumor cell expression of TNF receptors in head and neck squamous cell carcinoma. Am. J. Surgery. 170: 416–422. Wagner, R., and R. R. Myers. 1996. Endoneurial injection of TNF-alpha produces neuropathic pain behaviors. Neuroreport. 7: 2897–2901. Wagner, R., and R. R. Myers. 1996. Schwann cells produce tumor necrosis factor alpha: Expression in injured and noninjured nerves. Neuroscience. 73: 625–629. Wells, M. R., S. P. Racis, Jr., and U. Vaidya. 1992. Changes in plasma cytokines associated with peripheral nerve injury. J. Neuroimmunol. 39: 261–268.
44. Willenborg, D., S. Fordham, W. Cowden, and I. Ramshaw. 1995. Cytokines and murine autoimmune encephalomyelitis: Inhibition or enhancement of disease with antibodies to select cytokines, or by delivery of exogenous cytokines using a recombinant vaccinia virus system. Scand. J. Immunol. 41: 31–41. 45. Zettler, C., D. McLeod Bridges, X. Zhou, and R. Rush. 1996. Detection of increased tissue concentrations of nerve growth factor with an improved extraction procedure. J. Neurosci. Res. 46: 581–594. 46. Zhu, J., X. F. Bai, E. Mix, and H. Link. 1997. Cytokine dichotomy in peripheral nervous system influences the outcome of experimental allergic neuritis: Dynamics of mRNA expression for IL-1 beta, IL-6, IL-10, IL-12, TNF-alpha, TNF-beta, and cytolysin. Clin. Immunol. Immunopathol. 84: 85–94.
Serial Determination of Tumor Necrosis Factor-Alpha Content in Rat Sciatic Nerve after Chronic Constriction Injury Annette George, Christine Schmidt, Andreas Weishaupt, Klaus V. Toyka, and Claudia Sommer Neurologische Universita¨tsklinik Wu¨rzburg, Germany Received December 11, 1997; accepted July 20, 1999
Wallerian degeneration, induced after injury to a peripheral nerve, is associated with upregulation of proinflammatory cytokines, which are suggested to contribute to the development of lesion-induced neuropathic pain. In chronic constrictive injury (CCI), an animal model of injury-induced painful mononeuropathy, inhibition of synthesis, release, or function of the cytokine tumor necrosis factor-␣ (TNF) results in reduced pain-associated behavior. Here, changes of TNF content in rat sciatic nerves after CCI (days 0, 0.5, 1, 3, 7 and 14) were investigated by enzyme-linked-immunoassay. Low levels of TNF were already detectable in control nerves. Concentrations increased rapidly after CCI, with a maximum (2.7-fold) at 12 h, and remained elevated on a lower level until day 3. Baseline levels were reached again at day 14. These results indicate that TNF is produced at an early time point in the cascade of events resulting in Wallerian degeneration and hyperalgesia following peripheral nerve injury. Given that only prophylactic treatment with TNF inhibitors efficiently reduces hyperalgesia in CCI, TNF seems to contribute to the initiation of neuropathic pain in this model. r 1999 Academic Press Key Words: Wallerian degeneration; tumor necrosis factor-␣ (TNF); enzyme-linked immunoassay (ELISA); tissue homogenization; sedimental fraction; membrane TNF.
INTRODUCTION
The recent years of cytokine research in the central nervous system (CNS) have shown that the actions of the ‘‘proinflammatory’’ cytokine tumor necrosis factor-␣ (TNF, syn. TNF-␣) are not confined only to typical inflammatory or autoimmune diseases, but that TNF is also involved in degenerative, metabolic, ischemic, or traumatic lesions. This has led to new therapeutic approaches (22). Similarly, in the peripheral nervous system (PNS), TNF is not only known as a mediator in inflammatory neuropathies like the Guillain-Barre´Syndrom (29), leprosy (16), HIV neuropathy (13), neuroborreliosis (10), or in the animal model of experimen0014-4886/99 $30.00 Copyright r 1999 by Academic Press All rights of reproduction in any form reserved.
tal autoimmune neuritis (EAN) (37, 46). TNF is also involved in Wallerian degeneration and in regeneration processes after peripheral nerve injury. This has been shown in animal models of complete (transection) (37) or incomplete sciatic nerve lesion (crush, chronic constriction injury, CCI) (17, 42, 43). Injury to a peripheral nerve is often associated with the development of chronic neuropathic pain, refractory to standard analgesic drugs. It has been hypothesized that local cytokine actions, produced by invading immune cells and/or local resident cells which proliferate and become immunologically active after injury (‘‘periaxonal inflammatory processes’’) (5) may contribute to the generation and maintanance of such pain states (39). The key role of TNF and interleukin (IL)-1 as mediators in inflammatory pain is well documented (6, 9). A better understanding of the action of these cytokines after peripheral nerve injury might offer a new tool for managing neuropathic pain symptoms after nerve lesion (7, 33). Wells et al. described an increase of circulating TNF levels after unilateral crush lesion in the rat already in 1992 (43). An induction of TNFmRNA in the nerve after crush was recently described by La Fleur et al. using semiquantitative RT-PCR (17) and in CCI by in situ hybridization (ISH) (42). However, the expression of TNF is regulated not only transcriptionally but also on a translational and posttranslational level (1). Therefore, knowledge of actual local tissue protein concentrations is essential in order to further characterize a possible involvement of TNF in Wallerian degeneration. Furthermore, TNF can produce even contrasting effects depending on its site of action (21, 44). Using immunohistochemistry (IHC), an increase of TNFimmunoreactive (IR) protein was observed after transection (37), and CCI (42). Rotshenker and coworkers were able to detect various cytokines (IL-1, IL-6, GM-CSF) by ELISA in the cultured media of segments of lesioned nerve in vitro (24, 25, 26), but could not detect TNF. However, failure of TNF protein detection in lesioned peripheral nerve may perhaps be rather due to methodological difficulties than to absence of the protein. Therefore we decided to investigate endoneurial
124
TNF IN PERIPHERAL NERVE INJURY
changes of TNF in a behaviorally and morphologically well characterized animal model of painful mononeuropathy, the chronic constriction injury (CCI) (2, 30, 32), using an adapted extraction procedure and a more sensitive enzyme-linked immunoassay (ELISA) system. MATERIALS AND METHODS
Animals and Surgery All experiments were approved by the Bavarian State animal experimentation committee and carried out in accordance with its regulations. Female Sprague– Dawley rats (200–250 g), obtained from Charles-River, Germany, were housed in plastic cages under natural lighting conditions with access to food and water ad libitum. Chronic constrictive injury was performed as described by Bennett and Xie (2) with minor modifications. Under barbiturate anesthesia, the sciatic nerve was exposed unilaterally at the midthigh level. Three ligatures (chromic gut 4.O) were placed around the nerve with 1 mm spacing, and tied until they just slightly constricted the diameter of the nerve and a short twitch was seen in the respective hindlimb. Sham operations, where the nerve was exposed and mobilized, were performed contralaterally. Animals were sacrified on days 0.5, 1, 3, 7, and 14 post surgery (n ⫽ 12 per time point). Sciatic nerves were removed by cutting the nerve shortly above the site of the constrictive ligatures and 1.5 cm distally. Nerves from the sham-operated sides were removed accordingly. Length and wet weight of the nerve segments were determined. Nerves were immediately frozen on dry ice and stored at ⫺80°C. Control nerves (day 0) were removed from healthy animals.
125
well as membrane-bound TNF (Biosource International, pers. commun.). Recovery assays of spiked tissue samples were performed (n ⫽ 12). Standard curves in the standard buffer were compared to standard curves in nerve homogenates diluted in standard buffer (n ⫽ 3). The mean absorbance of the ‘‘zero standard’’ (buffer or nerve homogenate, respectively) was substracted from all data prior to plotting. Positive and negative controls were included in each assay. TNF concentration was expressed as pg/mg protein. Protein content was determined by the bicinchoninic acid protein assay reagent (Pierce, KMF Laborchemie, St. Augustin, Germany). The relative distribution of endogenous TNF between supernatant (S) and pellet (P) was expressed in percentages (pg TNF/ml found in pellet referred to the sum of pg TNF/ml found in the supernatant and pellet fraction ⫻ 100) (Fig. 2a). This form of presentation was used because it was independent of the absolute concentration of TNF shown in Fig. 3. Recovery of exogenous TNF (Biosource International, CA), added during the homogenization procedure, was investigated in normal (day 0) and injured (day 7) tissue: Rat sciatic nerves were pooled and minced under liquid nitrogen. The powder was dissolved in homogenization buffer. TNF standard (1 ng/ml) was added to one aliquot of this homogenous material (containing equal amounts of endogenously contained TNF levels) and vortexed thoroughly. Subsequently, samples (containing or lacking exogenous TNF standard) were treated as described above. Recovery of exogenous TNF was expressed as the percentage of the known amount of TNF that had been added (1 ng/ml) after subtraction of the concentration found for endogenous TNF in an aliquot treated accordingly in the absence of TNF standard.
Homogenization Procedure and ELISA Determination Sciatic nerves were pooled (n ⫽ 4) and homogenized in ice-cold phosphate-buffered saline (PBS), pH 7.4, containing protease inhibitors (aprotinin, leupeptin, pepstatin, PMSF; Boehringer Mannheim, Germany). After centrifugation at 10,000g at ⫹4°C for 10 min, supernatant (S) was removed. The remaining pellet (P) was rehomogenized in the original volume of homogenization buffer. Triton X-100 (Boehringer Mannheim, Germany) was added to S and P at a final concentration of 0.01%. The samples were vortexed thoroughly and centrifuged again. Supernatants were aliquoted and assayed in duplicate (after dilution in the standard buffer supplied) by the Cytoscreen Rat TNF-␣ Ultrasensitive ELISA kit (Biosource International, CA) according to the manufacturer’s instructions. This assay system detects rat TNF with a sensitivity of 0.7 pg/ml and also in the presence of soluble TNF-receptor-1. The monoclonal antibody for rat TNF recognizes soluble as
Western Blot Analysis Tissue homogenates (supernatant, pellet) of injured rat sciatic nerves (days 0.5 and 14 post-CCI) were separated on a 15% SDS–polyacrylamide gel and transferred to nitrocellulose membrane (Schleicher & Schu¨ll, Dassel, Germany). The Western blot was blocked for 2 h at 4°C and then incubated overnight with polyclonal sheep antiserum to mouse TNF. Goat anti-sheep IgG horseradish peroxidase conjugate (Dianova, Hamburg, Germany) was used as second antibody. The blots were developed with a 3,38-diaminobenzidine/H2O2 solution. Histology One additional animal with CCI was sacrificed at each time point used for ELISA in order to verify the lesion (plastic embedding, 1-µm sections, toluidine blue) (30) and to visualize TNF immunoreactivity (fro-
126
GEORGE ET AL.
zen sections). Immunohistochemistry was performed with a polyclonal antibody to TNF-␣ (1:1000, Genzyme, Ru¨sselsheim, Germany). Indirect immunogold silver staining method (Biotrend, Ko¨ln, Germany) was used for detection. Statistical Analysis Data represent mean values (from at least three independent experiments) ⫾ standard deviation. Significance of differences between groups was analyzed by two-tailed Student’s t test (Fig. 2). Analysis of variance (ANOVA) was used to test for differences between groups and subsequent Scheffe´ test for comparison of individual means throughout the time course (Fig. 3). Statistical significance was assumed with P ⬍ 0.05. RESULTS
Adapting an ELISA for TNF Determination in Rat Sciatic Nerve Homogenates The standard curve for rat TNF standard serially diluted in nerve homogenate lacking endogenous TNF was comparable to the standard curve in the standard buffer supplied (Fig. 1). Recovery of rat TNF standard added to nerve homogenate samples was 103 ⫾ 6% (n ⫽ 12). Relative Distribution of TNF between Supernatant and Pellet Fraction in Control Versus Injured Rat Sciatic Nerve Homogenates (a) Endogenously contained TNF. In control sciatic nerve homogenates, 88 ⫾ 1.7% of the TNF found was
detected in the supernatant, 12% in the supernatant of the rehomogenized pellet (Fig. 2a). Detection of TNF in the pellet was not due to residues of an uncompletely removed supernatant, because washing of the remaining sedimental fraction (instead of rehomogenization) failed to reach measurable TNF concentrations (data not shown). After lesion of the sciatic nerve by CCI, an increase of the amount of TNF found in the pellet fraction could be observed (Fig. 2a), which became statistically significant already on day 1 (81 ⫾ 1.6% in supernatant and 19% in pellet, P ⱕ 0.01) with further increase until day 7 (70 ⫾ 1.7 and 30%, P ⱕ 0.01). This change in the relative distribution of endogenous TNF between supernatant and pellet was independent of the absolute TNF levels found (see Fig. 3). In sham-operated sciatic nerves no significant changes in the relative amount of endogenous TNF found in the pellet fraction were observed (with 10 ⫾ 2.3 and 12 ⫾ 2.8% on days 0.5 and 7, respectively). (b) Exogenously added TNF standard. In control sciatic nerve homogenates (n ⫽ 5), 90.1 ⫾ 7.5% of exogenously added TNF standard (1 ng/ml homogenate) could be recovered: 84.4 ⫾ 9.5% in the supernatant, 5.7% in the supernatant of the rehomogenized pellet fraction (Fig. 2b). In sciatic nerve homogenates after lesion by CCI (day 7; n ⫽ 4), the recovery of exogenous TNF standard decreased to 43.3 ⫾ 5.2% (P ⱕ 0.01) in the supernatant; no TNF standard was detectable in the pellet fraction.
FIG. 1. Representative standard curve for rat TNF standard serially diluted in standard buffer (diamonds) and nerve homogenate lacking endogenous TNF (squares). Homogenization procedure as described under Material and Methods, corrected for background by substracting mean absorbance of zero standard (buffer or homogenate, respectively) from all data; range, 2.3–150 pg/ml.
TNF IN PERIPHERAL NERVE INJURY
127
FIG. 2. (a) Relative distribution of endogenous TNF in the supernatant (S) vs pellet (P) fraction after centrifugation of control (day 0) and CCI operated rat sciatic nerve homogenates (days 0.5, 1, 3, 7, 14). Extraction procedure as described under Materials and Methods. Data represent mean values of at least three experiments, expressed in % ⫾ SD; **P ⱕ 0.01. Note that the relative amount of endogenous TNF found in the pellet fraction increases postlesion independent of the absolute concentration of TNF shown in Fig. 3. (b) Recovery of exogenously added TNF (1 ng/ml) in the supernatant (S) vs pellet (P) fraction after centrifugation of control (day 0) and CCI-operated rat sciatic nerve homogenates (day 7). Extraction procedure as described under Materials and Methods. Data represent mean values of at least three experiments, expressed in % ⫾ SD; **P ⱕ 0.01.
Time Course of TNF Content in Rat Sciatic Nerve after CCI The TNF content in control rat sciatic nerve was 40.6 ⫾ 2 pg/mg protein. After CCI (Fig. 3), a rapid increase of TNF levels was observed, reaching its maximum (110.2 ⫾ 8.9 vs 53.7 ⫾ 13.7 pg/mg protein in CCI vs sham-operated side, respectively; P ⱕ 0.01)
already after 12 h. TNF levels declined afterward (62.8 ⫾ 10.4 and 52.3 ⫾ 9.7 pg/mg protein on day 1; P ⬎ 0.05) but still remained elevated significantly until day 3 (51.5 ⫾ 5.2 and 35.4 ⫾ 3.3 pg/mg protein CCI vs sham-operated side, respectively; P ⱕ 0.05). Baseline levels were reached again on day 14 (41.1 ⫾ 7.2 vs 44.1 ⫾ 7.2 pg/mg protein). On days 0.5 and 1, a slight,
128
GEORGE ET AL.
FIG. 3. Time course of TNF content in rat sciatic nerve after CCI. Endogenous TNF levels in rat sciatic nerve homogenates of control (day 0), CCI-operated (triangles), and sham-operated (squares) nerves (days 0.5, 1, 3, 7, 14); n ⫽ 12 animals per time point. Extraction procedure as described under Materials and Methods. TNF in supernatant and pellet fraction was determined by ELISA and data presented are the sum of TNF levels in the supernatant and the pellet fraction. Levels are expressed as mean values (pg/mg protein) ⫾ SD; n ⫽ 3; *P ⱕ 0.05, **P ⱕ 0.01 (CCI vs sham-operated side).
but statistically not significant increase of TNF levels was found in sham-operated nerves (53.7 ⫾ 13.7 and 52.3 ⫾ 9.7 pg/mg TNF on days 0.5 and 1, respectively, in sham-operated vs 40.6 ⫾ 2 pg/mg TNF in control sciatic nerves; P ⬎ 0.05). Histology Analysis of semithin sections confirmed Wallerian degeneration on the injured side (Fig. 4a). Immunohistochemistry for TNF revealed increased TNF-IR in nerve tissue on the side of the lesion (Figs. 4b and 4c). Molecular Forms of TNF Western blot analysis confirmed the presence of proteins with molecular weights of 17, 26, 34, and 51 kDa in the rat sciatic nerve homogenates (day 0.5 post-CCI) investigated by ELISA (Fig. 5). This is consistent with published molecular weights for the soluble, transmembrane, dimeric, or trimeric form of the TNF molecule, respectively (1). The transmembrane molecule was more prominent in the sedimental fraction compared to the supernatant; however, we did not detect a definite increase of the transmembrane form in the pellet fraction of homogenates of injured nerves harvested on day 14 post-CCI (data not shown). DISCUSSION
Our sensitive assay system allowed us to detect low (physiologic) TNF concentrations in rat sciatic nerve
and to follow its upregulation under pathological conditions such as chronic constrictive injury with subsequent Wallerian degeneration. Quantitative analysis of tissue TNF is critically dependent on the extraction procedure used (11). Here, we observed that it may also depend on the normal or pathologic state of the tissue investigated. Endogenous TNF may be lost to the pellet phase separated after centrifugation. Indeed, when nerve growth factor (NGF) in homogenized nervous tissue was analyzed by ELISA, a several-fold loss of NGF to the pellet fraction was found (14). Since this enormous loss of NGF may be due to the high centrifugation speed (100,000g for 1 h) (45), we used a lower centrifugation speed (10,000g), resulting in only low levels of endogenous as well as exogenous TNF detected in the pellet fraction. However, the relative amount of endogenous TNF recaptured from the pellet was increased after CCI. According to our Western blot analysis this seems probably not (or only to a minor degree) due to a shift in relation from the soluble (17 kDa) to the membrane-associated (26 kDa) form of TNF molecule in situ, as it has been described in a study on TNF in abnormal vessel walls (20). Other factors contributing to this observation may be (1) increased volume of the sedimental fraction, e.g., due to invading cells (but volume of homogenization buffer was always adapted to the wet weight of tissue investigated such that equal concentrations of homogenized tissue were achieved); (2) masking or increased degradation of TNF by mediators in pathologic tissue.
TNF IN PERIPHERAL NERVE INJURY
129
FIG. 4. (a) Semithin section, toluidine-blue stained, of CCI nerve, 10 mm distal from lesion. At 14 days postsurgery only few myelinated fibers remain (arrows). (b, c) Immunohistochemistry on frozen sections of (b) sham-operated sciatic nerve, (c) CCI nerve 7 days postsurgery. TNF immunoreactivity is weakly positive in occasional endoneurial cells (presumably Schwann cells) in sham-operated nerve (arrow) and abundant in endoneural cells (arrows) of the injured nerve. Bar ⫽ 20 µm for all photomicrographs.
Recovery of exogenously added TNF was reduced in injured nerve tissue. This may reflect degradation or complexing of soluble TNF during the homogenization procedure. A higher turnover and degradation rate of TNF at the site of local inflammatory tissue reaction has indeed been assumed by Khanolkar-Young comparing localization of TNF protein by IHC and TNFmRNA by ISH in nerve biopsies of leprosy patients (16). Therefore, actual concentrations of TNF in damaged tissue may be underestimated.
regions after sham operation (1.6–10 pg/mg for hippocampus, approx. 25–50 pg/mg protein for striatum and thalamus) (28). By other authors, lower (4, 20, 36, 40) or even undetectable (12, 18, 38) levels of TNF have been described in various nonnervous tissues in human or animal studies. Differences in species, type of organ investigated, procedure of tissue disruption, homogenization and extraction, and the sensitivity of the ELISA system may account for the different levels found.
Physiological Levels of TNF in the Rat Sciatic Nerve
Time Course of TNF Content in Rat Sciatic Nerve after Chronic Constriction Injury
TNF tissue concentrations found in normal rat sciatic nerve (40.6 pg/mg protein) are in the range described for other normal tissues in mice (between 17 and 38 pg/mg protein for muscle and lung, respectively) (15) and are also comparable to TNF levels found in human brain (18 pg/mg protein) (19) and in gerbil brain
After CCI, a rapid and early increase of endoneurial TNF was followed by a plateau phase returning to baseline levels by day 14. The fact that the maximum of the TNF increase after CCI occurs already 12 h postlesion or even earlier, suggests that resident cells become activated after
130
GEORGE ET AL.
FIG. 5. Western blot analysis confirms the presence of the soluble (17 kDa), transmembrane (26 kDa), dimeric (34 kDa), and trimeric (51 kDa) form of TNF molecule in the rat sciatic nerve homogenates (day 0.5 post-CCI; lanes 1 and 2).
injury and may be the main source of secreted TNF rather than hematogenous macrophages, which transmigrate not before day 3 in CCI (31). Hitherto available ISH studies have localized TNFmRNA in Schwann cells, fibroblasts, endothelial cells, and macrophages on day 7 post-CCI (42). La Fleur et al. (17) identified macrophages and Schwann cells as producers of TNFmRNA on day 4 after a crush lesion in the distal segment of the mouse sciatic nerve. Earlier time points regarding the cellular source are not available; but these results suggest that proliferating Schwann cells and resident macrophages synthesize and secrete TNF at this early time point after peripheral nerve injury. The time course of TNF protein determined by ELISA corresponds to the elevation of TNF-IR found in the CCI model using IHC (Marziniak and Sommer, unpublished observations) as well as to TNF mRNA changes reported recently after crush lesion of the mouse sciatic nerve (17). Interestingly, an early, sharp increase of TNF mRNA on day 1 was observed in the segment of crush lesion while a delayed elevation of longer duration was found in the segment distal to lesion. Elevation of endoneurial TNF levels parallels the initial structural changes during Wallerian degenera-
tion after CCI (30, 31): appearance of endoneurial edema, proliferation of Schwann- and endothelial cells, and axonal degeneration with almost complete loss of myelinated axons by day 7. First signs of regeneration (axonal sprouting, remyelination) are known to occur at time points of decreasing TNF levels. Whether there is a causal link between the rise in TNF and axonal damage in CCI, remains to be investigated. After endoneurial application of TNF to the rat sciatic nerve edema, axonal degeneration, demyelination, and vascular changes have been observed (23, 27). Reduction of endoneurial TNF production was associated with attenuated CCI-induced vascular changes of endoneurial vessels (32). On the other hand, reduced axonal damage and an improved recovery of motor function after crush lesion of rat sciatic nerve were reported when TNF was applied intraperitoneally prior to nerve lesion (3). Experimental administration of TNF into normal rat sciatic nerve induces pain-associated behavior (41). In CCI, the maximum of endoneurial TNF increase precedes the development of pain-associated behavior (30): thermal hyperalgesia and mechanical allodynia develop by days 1 to 3, with the maximum of thermal hyperalgesia between days 6 and 14. TNF levels declined subsequently and baseline levels were reached again on day 14, when hyperalgesia was still present. Administration of inhibitors of TNF synthesis, release, or function results in a reduction of pain-associated behavior in CCI only, when the inhibitor is given prophylactically but not when treatment is started 1 week after CCI (33–35). The early maximum of TNF elevation found may explain these results: TNF may contribute to the initiation of pain-associated behavior after nerve lesion, which is consistent with the role of TNF in initiation of the cytokine cascade involving sequentially IL-1, IL-6, and IL-8 (8). Pain-associated behavior may be maintained by the cytokine cascade induced by TNF involving IL-1, IL-6, and IL-8. Selective immunopharmacological blockade of these cytokines may help to elucidate their role in the maintanance of neuropathic pain states. ACKNOWLEDGMENTS Supported by Deutsche Forschungsgemeinschaft So 328/2-1. Presented in part at the Annual Meeting of the Society for Neuroscience, New Orleans, LA, October 25, 1997.
REFERENCES 1. Aggarwal, B. B., and K. Natarajan. 1996. Tumor necrosis factors: developments during the last decade. Eur. Cytokine Netw. 7: 93–124. 2. Bennett, G. J., and Y. K. Xie. 1988. A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain. 33: 87–107. 3. Chen, L. E., A. V. Seaber, G. H. Wong, and J. R. Urbaniak. 1996.
TNF IN PERIPHERAL NERVE INJURY
4.
5.
6.
7.
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Tumor necrosis factor promotes motor functional recovery in crushed peripheral nerve. Neurochem. Int. 29: 197–203. Clancy, K. D., K. Lorenz, E. Hahn, B. Christiansen, C. Hofmann, and R. L. Gamelli. 1997. Down-regulation of tissue specific tumor necrosis factor-alpha in the liver and lung after burn injury and endotoxemia. J. Trauma. 42: 169–176. Clatworthy, A. L., P. A. Illich, G. A. Castro, and E. T. Walters. 1995. Role of peri-axonal inflammation in the development of thermal hyperalgesia and guarding behavior in a rat model of neuropathic pain. Neurosci. Lett. 184: 5–8. Cunha, F. Q., S. Poole, B. B. Lorenzetti, and S. H. Ferreira. 1992. The pivotal role of tumour necrosis factor alpha in the development of inflammatory hyperalgesia. Br. J. Pharmacol. 107: 660–664. DeLeo, J., R. Colburn, and A. Rickman. 1997. Cytokine and growth factor immunohistochemical spinal profiles in two animal models of mononeuropathy. Brain Res. 759: 50–57. Eugui, E., B. DeLustro, S. Rouhafza, R. Wilhelm, and A. Allison. 1997. Coordinate inhibition by some antioxidants of TNF, IL-1, and IL-6 production by human peripheral blood mononuclear cell. Ann. NY Acad. Sci. 171–184. Ferreira, S. H. 1993. The role of interleukins and nitric oxide in the mediation of inflammatory pain and its control by peripheral analgesics. Drugs. 46: 1–9. Garcia-Monco, J. C., and J. L. Benach. 1997. Mechanisms of injury in Lyme neuroborreliosis. Semin. Neurol. 17: 57–62. George, A., C. Schmidt, and C. Sommer. 1998. Quantification of TNF protein in homogenized rat sciatic nerve by ELISA. Soc. Neurosci. Abs. 24: 1539. Gionchetti, P., M. Campieri, A. Belluzzi, E. Bertinelli, M. Ferretti, C. Brignola, G. Poggioli, M. Miglioli, and L. Barbara. 1994. Mucosal concentrations of interleukin-1 beta, interleukin-6, interleukin-8, and tumor necrosis factor-alpha in pelvic ileal pouches. Dig. Dis. Sci. 39: 1525–1531. Griffin, J. W., S. L. Wesselingh, D. E. Griffin, J. D. Glass, and J. C. McArthur. 1994. In HIV, AIDS and the Brain (Q. W. Price and S. W. Perry, Eds.), pp. 159–182. Raven Press, New York. Hoener, M., E. Hewitt, J. Conner, J. Costello, and S. Varon. 1996. Nerve growth factor content in adult rat brain tissues is several-fold higher than generally reported and is largely associated with sedimentable fractions. Brain Res. 728: 47–56. Hotchkiss, R., D. Osborne, G. Lappas, and I. Karl. 1995. Calcium antagonists decrease plasma and tissue concentrations of tumor necrosis factor, interleukin-1 alpha and beta in amouse model of endotoxin. Shock 3: 337–342. Khanolkar-Young, S., N. Rayment, P. M. Brickell, D. R. Katz, S. Vinayakumar, M. J. Colston, and D. N. Lockwood. 1995. Tumour necrosis factor-alpha synthesis is associated with the skin and peripheral nerve pathology of leprosy reversal reactions. Clin. Exp. Immunol. 99: 196–202. La Fleur, M., J. L. Underwood, D. A. Rappolee, and Z. Werb. 1996. Basement membrane and repair of injury to peripheral nerve: Defining a potential role for macrophages, matrix metalloproteinases, and tissue inhibitor of metalloproteinases-1. J. Exp. Med. 184: 2311–2326. MacMaster, M., R. Newton, K. Sudhansu, and G. Andrews. 1992. Activation and distribution of inflammatory cells in the mouse uterus during the preimplantation period. J. Immunol. 148: 1699–1705. Mogi, M., M. Harada, P. Riederer, H. Narabayashi, K. Fujita, and T. Nagatsu. 1994. TNF alpha increases both in the brain and in the cerebrospinal fluid from parkinsonian patients. Neurosci. Lett. 165: 208–219. Newman, K. M., J. Jean-Claude, H. Li, W. G. Ramey, and M. D.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32. 33.
34.
35.
36.
37.
38.
131
Tilson. 1994. Cytokines that activate proteolysis are increased in abdominal aortic aneurysms. Circulation. 90: II224–227. Paya, C., P. Leibson, A. Patick, and M. Rodriguez. 1990. Inhibition of Theiler’s virus-induced demyelination in vivo by TNF. Int. Immunol. 2: 909–913. Probert, L. K. Selmaj. 1997. TNF and related molecules: Trends in neuroscience and clinical applications. J. Neuroimmunol. 72: 113–117. Redford, E. J., S. M. Hall, and K. J. Smith. 1995. Vascular changes and demyelination induced by the intraneural injection of tumour necrosis factor. Brain 118: 869–878. Reichert, F., R. Levitzky, and S. Rotshenker. 1996. Interleukin-6 in intact and injured mouse peripheral nerves. Eur. J. Neurosci. 8: 530–535. Rotshenker, S., S. Aamar, and V. Barak. 1992. Interleukin-1 activity in lesioned peripheral nerve. J. Neuroimmunol. 39: 75–80. Saada, A., F. Reichert, and S. Rotshenker. 1996. Granulocyte macrophage colony stimulating factor produced in lesioned peripheral nerves induces the up-regulation of cell surface expression of MAC-2 by macrophages and schwann cells. J. Cell Biol. 133: 159–167. Said, G. M., and M. Hontebeyrie-Joskowicz. 1992. Nerve lesions induced by macrophage activation. Res. Immunol. 143: 589– 599. Saito, K., K. Suyama, K. Nishida, Y. Sei, and A. S. Basile. 1996. Early increases in TNF-alpha, IL-6 and IL-1 beta levels following transient cerebral ischemia in gerbil brain. Neurosci. Lett. 206: 149–152. Sharief, M., B. McLean, and E. Thompson. 1993. Elevated serum levels of TNF in Guillain-Barre-Syndrome. Ann. Neurol. 33: 591–596. Sommer, C., J. A. Galbraith, H. M. Heckman, and R. R. Myers. 1993. Pathology of experimental compression neuropathy producing hyperesthesia. J. Neuropathol. Exp. Neurol. 52: 223– 233. Sommer, C., A. Lalonde, H. M. Heckman, M. Rodriguez, and R. R. Myers. 1995. Quantitative neuropathology of a focal nerve injury causing hyperalgesia. J. Neuropathol. Exp. Neurol. 54: 635–643. Sommer, C., and R. R. Myers. 1996. Vascular changes in a model of painful neuropathy. Exp. Neurol. 141: 113–119. Sommer, C., C. Schmidt, and A. George. 1998. Hyperalgesia in experimental neuropathy is dependent on the TNF receptor 1. Exp. Neurol. 151: 138–142. Sommer, C., M. Marziniak, and R. R. Myers. 1998. The effect of thalidomide treatment on vascular pathology and hyperalgesia caused by chronic constrictive injury of rat sciatic nerve. Pain 74: 83–91. Sommer, C., C. Schmidt, A. George, and K. Toyka. 1997. A metalloprotease-inhibitor reduces pain associated behavior in mice with experimental neuropathy. Neurosci. Lett. 237: 45–48. Stashenko, P., J. J. Jandinski, P. Fujiyoshi, J. Rynar, and S. S. Socransky. 1991. Tissue levels of bone resorptive cytokines in periodontal disease. J. Periodontol. 62: 504–509. Stoll, G., S. Jung, S. Jander, P. van der Meide, and H. P. Hartung. 1993. Tumor necrosis factor-alpha in immunemediated demyelination and Wallerian degeneration of the rat peripheral nervous system. J. Neuroimmunol. 45: 175–182. Takematsu, H., H. Ohta, and H. Tagami. 1989. Absence of tumor necrosis factor-alpha in suction blister fluids and stratum corneum from patients with psoriasis. Arch. Dermatol. Res. 281: 398–400.
132 39. 40.
41.
42.
43.
GEORGE ET AL. Tracey, D. J., and J. S. Walker. 1995. Pain due to nerve damage: Are inflammatory mediators involved? Inflamm. Res. 44: 407–411. von Bieberstein, S., J. Spiro, R. Lindquist, and D. Kreutzer. 1995. Enhanced tumor cell expression of TNF receptors in head and neck squamous cell carcinoma. Am. J. Surgery. 170: 416–422. Wagner, R., and R. R. Myers. 1996. Endoneurial injection of TNF-alpha produces neuropathic pain behaviors. Neuroreport. 7: 2897–2901. Wagner, R., and R. R. Myers. 1996. Schwann cells produce tumor necrosis factor alpha: Expression in injured and noninjured nerves. Neuroscience. 73: 625–629. Wells, M. R., S. P. Racis, Jr., and U. Vaidya. 1992. Changes in plasma cytokines associated with peripheral nerve injury. J. Neuroimmunol. 39: 261–268.
44. Willenborg, D., S. Fordham, W. Cowden, and I. Ramshaw. 1995. Cytokines and murine autoimmune encephalomyelitis: Inhibition or enhancement of disease with antibodies to select cytokines, or by delivery of exogenous cytokines using a recombinant vaccinia virus system. Scand. J. Immunol. 41: 31–41. 45. Zettler, C., D. McLeod Bridges, X. Zhou, and R. Rush. 1996. Detection of increased tissue concentrations of nerve growth factor with an improved extraction procedure. J. Neurosci. Res. 46: 581–594. 46. Zhu, J., X. F. Bai, E. Mix, and H. Link. 1997. Cytokine dichotomy in peripheral nervous system influences the outcome of experimental allergic neuritis: Dynamics of mRNA expression for IL-1 beta, IL-6, IL-10, IL-12, TNF-alpha, TNF-beta, and cytolysin. Clin. Immunol. Immunopathol. 84: 85–94.