Leptin derived from adipocytes in injured peripheral nerves facilitates development of neuropathic pain via macrophage stimulation Takehiko Maedaa,1, Norikazu Kiguchia, Yuka Kobayashia, Toshihiko Ikutaa, Masanobu Ozakib, and Shiroh Kishiokaa aDepartment of Pharmacology, Wakayama Medical University, Wakayama 641-0012, Japan; and bDepartment of Toxicology, Niigata University of Pharmacy and Applied Life Science, Niigata 950-2028, Japan
Edited by David Julius, University of California, San Francisco, CA, and approved June 12, 2009 (received for review April 1, 2009)
Nerve injury may result in neuropathic pain, characterized by allodynia and hyperalgesia. Accumulating evidence suggests the existence of a molecular substrate for neuropathic pain produced by neurons, glia, and immune cells. Here, we show that leptin, an adipokine exclusively produced by adipocytes, is critical for the development of tactile allodynia through macrophage activation in mice with partial sciatic nerve ligation (PSL). PSL increased leptin expression in adipocytes distributed at the epineurium of the injured sciatic nerve (SCN). Leptin-deficient animals, ob/ob mice, showed an absence of PSL-induced tactile allodynia, which was reversed by the administration of leptin to the injured SCN. Perineural injection of a neutralizing antibody against leptin reproduced this attenuation. Macrophages recruited to the perineurium of the SCN expressed the leptin receptor and phosphorylated signal transducer and activator of transcription 3 (pSTAT3), a transcription factor downstream of leptin. PSL also up-regulated the accepted mediators of neuropathic pain—namely, cyclooxygenase-2, inducible nitric oxide synthase, and matrix metalloprotease-9 —in the injured SCN, with transcriptional activation of their gene promoters by pSTAT3. This up-regulation was partly reproduced in a macrophage cell line treated with leptin. Administration of peritoneal macrophages treated with leptin to the injured SCN reversed the failure of ob/ob mice to develop PSLinduced tactile allodynia. We suggest that leptin induces recruited macrophages to produce pronociceptive mediators for the development of tactile allodynia. This report shows that adipocytes associated with primary afferent neurons may be involved in the development of neuropathic pain through adipokine secretion. adipokine 兩 allodynia 兩 C/EBP 兩 fat 兩 STAT
N
europathic pain is elicited by a lesion or inflammation of the nervous system and is often severely debilitating and largely resistant to treatment. The symptoms of neuropathic pain may include allodynia, hyperalgesia, and spontaneous pain. A great deal has been reported on neuropathic pain and its possible causes from preclinical studies involving the development of animal models of pain. It has become clear that the neuroinflammatory mechanism in the periphery nervous system (PNS) plays an important role in neuropathic pain (1). Infiltration of inflammatory cells, as well as the activation of resident immune cells in response to nervous system damage, leads to the subsequent production and secretion of various inflammatory mediators. For example, inflammatory cells, such as mast cells, neutrophils, and macrophages, produce and secrete prostanoids, nitric oxide, cytokines, chemokines, and matrix metalloproteases (2). These mediators promote neuroimmune activation and can sensitize primary afferent neurons and contribute to the development of neuropathic pain. Although the number of molecules involved in neuropathic pain is increasing, their mechanisms of action are still unclear. Adipose tissue distributed around internal organs, visceral fat, is well known to serve as an endocrine organ involved in obesity and metabolism. Adipocytes are also localized in the PNS (3), 13076 –13081 兩 PNAS 兩 August 4, 2009 兩 vol. 106 兩 no. 31
but little attention has been paid to their functional importance. Adipokines secreted from adipose tissue have come to be recognized as molecular substrates contributing to obesity and related metabolic disorders (4). Leptin, an adipokine, is a 16-kDa, nonglycosylated peptide hormone encoded by the obese gene in mice (5). Stimulation of its receptor leads to activation of the Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling pathway (6). Leptin also has a role in the regulation of energy homeostasis (7). Recently, leptin has emerged as a modulator of inflammation and immune responses in the nervous system. The underlying mechanism is likely to be leptin-stimulated production of inflammatory mediators (8–10), which are also accepted molecular substrates for neuropathic pain (11–13). The present study focused on the role of leptin secreted from adipocytes in the PNS in neuropathic pain. Here, we show that tactile allodynia induced by partial sciatic nerve ligation is affected in its developmental phase by leptin possibly secreted from adipocytes in the epineurium of the sciatic nerve (SCN). Furthermore, the leptin signaling molecules involved in this regulation include some conventional molecular substrates for allodynia (11–13) produced by macrophages through activation of the JAK-STAT pathway. Results Role of Leptin in Allodynia. We examined the effect of leptin
deficiency on tactile allodynia and thermal hyperalgesia induced by PSL. PSL in lean mice elicited a significantly higher ratio of the withdrawal response to tactile stimuli after day 3 following PSL than did sham-operated lean mice, indicating that tactile allodynia occurred (Fig. 1A). By contrast, ob/ob mice did not show a significant increase in the ratio of responses during the examined days after PSL; however, PSL did induce thermal hyperalgesia in ob/ob mice (Fig. 1B). There was no significant difference in the ratio of tactile responses between shamoperated ob/ob and lean mice during the study period after PSL. These results suggest that tactile allodynia was not developed in ob/ob mice. Consistent with these results, perineural injection of a neutralizing antibody for leptin into the ligated SCN reduced the magnitude of PSL-induced tactile allodynia in ICR mice (Fig. 1C). PSL induced an increase in tyrosine phosphorylation of Ob-Rb (P-Ob-R), the long isoform of the leptin receptor, in the SCN, indicating that PSL stimulated leptin signaling (Fig. S1B). The behavioral effect of the neutralizing antibody was associated with the attenuation of the increase in PSL-induced expression of P-Ob-R (Fig. S1B). To test whether leptin replaceAuthor contributions: T.M. and M.O. designed research; T.M., N.K., and T.I. performed research; T.M. and Y.K. analyzed data; and T.M. and S.K. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. 1To
whom correspondence should be addressed. E-mail:
[email protected].
This article contains supporting information online at www.pnas.org/cgi/content/full/ 0903524106/DCSupplemental.
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0903524106
0.5
10 ***
***
** pre-PSL 3 5 7 Day after PSL
Ratio of responses
*** ICR
***
0.5
0
IgG Nab IgG Nab sham PSL
1.0
Leptin at early phase ns *** ob/ob
0.5
0
** sham/lean sham/ob PSL/lean PSL/ob
* ##
**
** #
###
pre-PSL 3 5 7 Day after PSL
14
D Ratio of responses
Neutralizing antibody
1.0
8 6 4 2 0
0
C
Thermal hyperalgesia
veh lep veh lep sham PSL
E1.0 Ratio of responses
sham/lean sham/ob PSL/lean PSL/ob ***
14
Leptin at late phase ob/ob
0.5
0
ns
ns
veh lep veh lep sham PSL
Fig. 1. Leptin was involved in the development of tactile allodynia. (A and B) Time course of tactile allodynia and thermal hyperalgesia. *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 vs. sham in lean mice. #, P ⬍ 0.05; ##, P ⬍ 0.01; ###, P ⬍ 0.001 vs. sham-operated ob/ob mice. (C) Leptin-neutralizing antibody (Nab; 1 ng) or normal goat IgG (10 ng) as a control was injected into the perineural SCN of ICR mice every other day from day 0 (immediately after PSL) to day 4 after PSL. A behavioral test was conducted on day 5 after PSL. ***, P ⬍ 0.001. (D and E) Recombinant mouse leptin (lep; 1 g) or vehicle (veh) was perineurally administered to the SCN with surgery in ob/ob mice, once daily in the early phase [from day 0 (immediately after PSL) to day 4 after PSL] or in the late phase (from day 7 to day 11 after PSL). A behavioral test was conducted on day 5 after PSL in the early-phase group, or on day 12 after PSL in the late-phase group (n ⫽ 8). ***, P ⬍ 0.001. ns, not significant.
ment could prevent the failure of ob/ob mice to develop tactile allodynia, the effect of leptin injection into the SCN in mice that had undergone PSL was examined. Daily perineural injection of leptin (1 g) into the ligatured SCN after PSL reversed the failure of ob/ob mice to develop tactile allodynia on day 5 after PSL (Fig. 1D). By contrast, treatment of ob/ob mice with leptin on days 7–11 after PSL did not affect the failure of these mice to develop PSL-induced tactile allodynia (Fig. 1E). To test whether the effect of perineurally applied leptin involved a systemic action, leptin was administered s.c. Systemic administration at the same dose as used in the perineural injection did not lead to tactile allodynia in ob/ob mice with PSL (Fig. S1 A). Leptin injections on either days 0–4 or days 7–11 produced no significant effect on tactile allodynia in lean mice (Fig. S1 C and D). These results indicate that endogenous leptin works in the early phase of PSL-induced tactile allodynia. PSL-Induced Up-Regulation of Leptin in Perineural Adipocytes. Western blotting revealed that PSL induced rapid expression of leptin protein in the SCN, the level of which declined after day 7 following PSL, with no significant change being observed in sham-operated mice (Fig. 2A). PSL also increased the level of leptin mRNA in the SCN at 3 h after PSL (Fig. 2B). However, PSL did not significantly affect the serum level of leptin at 3 h after operation (3.11 ⫾ 0.58 ng/mL in sham and 3.97 ⫾ 0.70 ng/mL in PSL, n ⫽ 7; P ⫽ 0.181). The expression of leptin mRNA is transcriptionally enhanced by CCAAT/enhancer-binding protein-␣ (C/EBP-␣), a transcription factor that is also known to be an acute-phase protein and a key regulator of adipocyte differentiation (14). We addressed whether PSL-induced expression of leptin mRNA was mediated transcriptionally by C/EBP-␣. A ChIP assay revealed that markedly increased binding of C/EBP-␣ to the leptin promoter occurred in the injured SCN 3 h after PSL Maeda et al.
(Fig. 2C). Adipose tissue in the epineurium of the SCN has been noted previously, and more recently it was shown to function as an endocrine organ through the production and secretion of adipokines, such as leptin (3). Immunohistochemistry of the SCN was performed to determine the distribution of leptin and C/EBP-␣, molecules expressed specifically in adipocytes. Adipocytes, stained with BODIPY, were distributed in the epineurium of the SCN in sham-operated (Fig. 2K) and PSL (Fig. 2G) mice. The adipocytes in the SCN of sham-operated mice expressed C/EBP-␣ and leptin peripheral to BODIPYstained neutral lipids (Fig. 2 I and J). Increased staining for leptin was observed in the perineural adipocytes of the ligatured SCN (Fig. 2F). These data suggest that leptin is up-regulated in the epineurial adipocytes of the injured SCN, at least in part through transactivation of the leptin gene by C/EBP-␣. Leptin Stimulates Macrophages Through Activation of STAT3. It has been found that phosphorylation of both STAT3 and JAK2 follows activation of the leptin receptor (15); therefore, to test whether the JAK-STAT pathway is activated by leptin during the development of tactile allodynia, we evaluated the levels of phospho-STAT3 (pSTAT3), an active form of STAT3, in the SCN of PSL mice. A rapid increase in the level of pSTAT3 occurred in the SCN of PSL mice (Fig. 3A). Macrophages contribute to pain resulting from peripheral nerve injury (16). Furthermore, leptin stimulates tyrosine phosphorylation of JAK2 and STAT3 in macrophages (17). We examined whether PSL stimulates the JAK-STAT pathway in macrophages recruited to the injured SCN. Macrophages were recruited to the perineurium of the SCN at 3 h after PSL (Fig. 3K), whereas there were far fewer in the SCN of sham-operated mice (Fig. 3F). The number of macrophages at 3 h after PSL was significantly different between sham-operated (27.6 ⫾ 6.0 cells, n ⫽ 7) and PSL (64.7 ⫾ 9.3 cells, n ⫽ 7; P ⫽ 0.006 vs. sham) mice. The leptin receptor and pSTAT3 were coexpressed in macrophages in the injured SCN (Fig. 3H). Additionally, the possible function of the JAK-STAT pathway in tactile allodynia was assessed. A single administration of AG490 (40 mg/kg i.p.), a JAK2 inhibitor, prevented the development of tactile allodynia (Fig. S2 A) and inhibited the increase in the levels of phospho-JAK2 (pJAK2), an active form (Fig. S2B), and pSTAT3 (Fig. S2C) in the injured SCN. These results indicate a significant role of JAK-STAT pathway activation, downstream of leptin, in recruited macrophages in the development of tactile allodynia. Transcriptional Activation of Molecules Involved in Allodynia Development Through pSTAT3 in Macrophages. Many studies have re-
ported key molecules related to neuropathic pain established by peripheral nerve injury. Most of these studies focused on expression of these molecules in the spinal cord and DRG, but there have been few reports of the molecules involved in the injured SCN. Furthermore, the mechanisms underlying the expression of these molecules remain to be determined. To investigate which allodynia-related molecules are driven by activation of the JAK-STAT pathway in the injured SCN, we focused on 3 allodynia-related molecules reported previously: cyclooxygenase-2 (COX-2), inducible nitric oxide synthase (iNOS), and matrix metalloprotease-9 (MMP-9). A ChIP assay tk;2elucidated a marked increase in pSTAT3 binding on the promoter sequence in the injured SCN from 3 h after PSL (Fig. 4 A–C). Transcription levels of all 3 molecules also increased coincidentally with enhanced binding of pSTAT3 (Fig. 4 D–F). We used a macrophage cell line, J774A.1, to examine whether transcription of these 3 molecules was enhanced in macrophages. In line with the findings in the injured SCN, J774A.1 cells treated with leptin showed a significant increase in the mRNA levels for iNOS and MMP-9 (Fig. 4 H and I). On the other hand, the level of COX-2 mRNA was not significantly changed, although LPS PNAS 兩 August 4, 2009 兩 vol. 106 兩 no. 31 兩 13077
PHARMACOLOGY
B 12
Mechanical allodynia 1.0
Latency (s)
Ratio of responses
A
A
C
B
D
E
F
G
H
I
J
K
treatment induced a marked increase in the level of this mRNA (Fig. 4G). Leptin-Stimulated Macrophages Facilitate Allodynia Development.
We addressed whether leptin-stimulated macrophages in the injured SCN may be responsible for the development of tactile allodynia. Peritoneal macrophages (PMs) from naive ob/ob mice showed pSTAT3 expression (Fig. 5A) when treated with leptin (1 g/mL for 3 h). PMs were distributed in the epineural SCN in PSL (Fig. 5C) and sham-operated (Fig. 5D) ob/ob mice 1 day after they were percutaneously administered to the perineural SCN, which was performed immediately after PSL. These results suggest the possibility that leptin treatment evoked JAK-STAT pathway activation in PMs, and that injected PMs are functional in the SCN. Administration of leptin-treated PMs evoked tactile allodynia in ob/ob mice subjected to PSL, but administration of vehicle-treated PMs did not have a significant effect in these mice (Fig. 5E). The ratio of withdrawal responses was not significantly different between sham-operated mice injected with leptin-treated or vehicle-treated PMs. Discussion We provide evidence that leptin, possibly derived from adipocytes in the epineurium of the SCN, contributes to the development of tactile allodynia through the production of the molecular substrates for neuropathic pain by stimulation of the JAK-STAT pathway in macrophages, as supported by the following results: (i) ob/ob mice, which are leptin-deficient, failed to show PSL-induced tactile allodynia; (ii) PSL enhanced transcription of the leptin gene in the injured SCN; (iii) PSL resulted in recruitment of pSTAT3-positive macrophages to the epineurium of the SCN; (iv) PSL induced expression of mRNAs for MMP-9, COX-2, and iNOS, molecules known to underlie allodynia development, through pSTAT3 transactivation, in the injured SCN; (v) MMP-9 and iNOS mRNA levels were also increased in a macrophage cell line treated with leptin; and (vi) 13078 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0903524106
Fig. 2. Expression of leptin was increased in the adipocytes in the epineurium of the injured SCN. (A) Time course of leptin expression in the SCN. Leptin expression was corrected based on the level of -tubulin. (B) RT-PCR for leptin mRNA in the SCN 3 h after PSL. Leptin expression was corrected based on the level of GAPDH. (C) ChIP analysis of binding of C/EBP-␣ to the leptin promoter. Immunoprecipitation (IP) of complexes of C/EBP-␣ binding and the gene promoter was corrected for using the gene without IP (input) (n ⫽ 4). *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 vs. sham-operated mice. (D–K) Sections of the SCN from shamoperated (H–K) and PSL (D–G) mice at 3 h of PSL were stained with BODIPY (G and K), which binds neutral lipids in adipocytes, and were immunostained for C/EBP-␣ (E and I) and leptin (F and J). (D and H) Merged images. A control section of the SCN of a PSL mouse at 3 h, for which normal IgG was substituted for the primary antibody tested, showed no immunostaining for leptin or C/EBP-␣ in the ligatured SCN (Fig. S3 Upper). (Scale bars: 100 m.)
Administration of leptin-treated PMs reversed the failure of ob/ob mice to show PSL-induced tactile allodynia. C/EBP-␣ has been identified as a transcription factor that binds to the leptin promoter, working through a consensus
A
B
C
D
E
F
G
H
I
J
K
Fig. 3. PSL-induced stimulation of intracellular signaling downstream of the leptin receptor in macrophages. (A) Time course of STAT3 activation. The pSTAT3 expression was corrected based on the level of STAT3 expression (n ⫽ 4). *, P ⬍ 0.05; ***, P ⬍ 0.001 vs. sham-operated mice. (B–K) The sections of SCN from sham-operated (B–F) and PSL (G–K) mice were immunostained for pSTAT3 (B, D, G, I), leptin receptor (E and J), and F4/80, a marker for macrophages (F and K). (C–F, H–K) High-power magnification of the outlined areas in B and G, respectively. (C and H) The colocalization of pSTAT3, leptin receptor and F4/80 in sham-operated and PSL mice, respectively. A control section, for which normal IgG was substituted for the primary antibody used, showed no immunostaining for pSTAT3 or leptin receptor in the injured SCN (Fig. S3 Lower). (Scale bars: C–F and H–K, 10 m; B and G, 150 m.)
Maeda et al.
B
C
D
E
F
G
H
I
Fig. 4. PSL up-regulates molecules underlying allodynia development in macrophages through pSTAT3. (A–C) ChIP analysis of pSTAT3 binding on the promoter sequences of COX-2 (A), iNOS (B), and MMP-9 (C) in the SCN (n ⫽ 5 in A and B; n ⫽ 4 in C). (D–F) Time course of mRNA expression for COX-2 (D), iNOS (E), and MMP-9 (F) in the SCN (n ⫽ 4). Sham and PSL are abbreviated as S and P, respectively. White and black columns indicate sham and PSL, respectively. (G–I) Effect of leptin on the mRNA expression of COX-2 (G), iNOS (H), and MMP-9 (I) in J774A.1 cells, a macrophage cell line. The cell line was cultured in serum-free culture medium including leptin (16 g/mL) or LPS (0.1 g/mL) for the indicated number of hours. The cell line was cultured in drug-free medium for 24 h as a control group, indicated by C (n ⫽ 6 –9). *, P ⬍ 0.05; **, P ⬍ 0.01; ***, P ⬍ 0.001 vs. sham treatment (A–F) or control (G–I).
C/EBP-binding site in the proximal leptin promoter (18, 19); it plays a critical role in the transcription of acute-phase proteins (14). Transactivation of the leptin promoter by C/EBP-␣ was rapidly increased (Fig. 2C). Although it remains unclear how nerve damage induces leptin up-regulation through the transactivation of C/EBP-␣, the prompt induction of transcription indicates that leptin may work in the induction phase of tactile allodynia. This hypothesis was robustly confirmed by the result that leptin administration immediately after PSL was more effective in reversing the absence of tactile allodynia in ob/ob mice (Fig. 1D) than administration commencing 7 days after PSL (Fig. 1E). By day 7 after PSL, leptin expression returned to the basal level (Fig. 2 A), whereas binding of C/EBP-␣ to the leptin promoter remained elevated until day 14 after PSL (Fig. 2C). The results of the present study cannot explain this different temporal profile, but the discrepancy might be resolved by further studies on the time course of the transcriptional activity of a potential repressor of leptin (20) in adipocytes in the injured SCN. The present evidence indicates that adipocytes in the epineurium of the SCN are an important source of leptin. First, Maeda et al.
LPS and cytokines, which are possible inducers of neuropathic pain, reportedly both induce leptin expression (21). Second, increased levels of leptin protein were observed in adipocytes in the epineurium of the SCN 3 h after PSL (Fig. 2F), when the level of leptin expression in the SCN was significantly increased, as revealed by Western blot analysis and RT-PCR (Fig. 2 A and B). Secreted leptin may diffuse into the perineurium and act on macrophages recruited by cytokines and chemokines (1) and/or leptin itself (22). In addition, leptin might be delivered to the endoneurium and involved in allodynia development: although the perineurium forms a diffusion barrier between the epineurium and the endoneurium, these 2 compartments are nonetheless connected by the vascular network of the peripheral nerve (23). Third, at 3 h after PSL, we observed no significant difference in the serum levels of leptin between sham-operated and PSL mice. This result suggests that leptin locally up-regulated in the injured SCN contributes more to the development of tactile allodynia than circulating leptin: the significance of its role is less in systemically distributed adipocytes. PNAS 兩 August 4, 2009 兩 vol. 106 兩 no. 31 兩 13079
PHARMACOLOGY
A
vehicle-treated
B
E 1.0
Cultured PM (pSTAT3/F4/80)
PSL
sham
C Delivered PM (Hoechst)
D SCN
Ratio of responses
Leptin-treated
A
ns
**
0.5
SCN
0 shamPSL sham PSL vehicle PM leptin PM
Fig. 5. Administration of leptin-treated macrophages reverses the failure of allodynia development in ob/ob mice. PMs from ob/ob mice were cultured for 3 h in culture medium, including Hoechst 33342 (for nuclear staining) and leptin (1 g/mL) or vehicle (PBS). (A and B) Some cultured PMs were immunostained for pSTAT3 and F4/80. Cultured PMs were washed with drug-free PBS and then perineurally injected into the SCN of a different ob/ob mouse immediately after PSL and on days 2 and 4 after PSL. (C and D) The SCNs isolated from PM-administered sham-operated and PSL mice were sectioned on day 1 after PSL. Arrowheads, double-stained (A) or nuclearstained (C and D) cells. (Scale bars: A and B, 10 m; C and D, 100 m.) (E) Tactile allodynia was evaluated on day 5 of PSL (n ⫽ 7– 8). **, P ⬍ 0.01. ns, not significant.
Studies of rodents with genetic abnormalities of leptin or leptin receptors have revealed deficits in macrophage phagocytosis and the expression of proinflammatory cytokines (24). Macrophages in the PNS contribute to pain resulting from peripheral nerve injury (16). Thus, our hypothesis was that macrophages are target cells for leptin in the perineural SCN of PSL mice. This hypothesis is supported by the following findings. First, macrophages expressing leptin receptors and pSTAT3, a primary leptin receptor signaling molecule, were seen in the injured SCN 3 h after PSL (Fig. 3H), when the expression level of pSTAT3 was maximal in the injured SCN (Fig. 3A). Second, treatment with leptin led to the production of molecules responsible for allodynia development in a macrophage cell line, with a peak level at 3 h after PSL (Fig. 4 H and I). The molecular substrates for allodynia development are beginning to be unraveled by accumulated evidence; among these, we focused on molecules with pSTAT3-binding consensus sequences in the promoter or enhancer, including MMP-9 (11), iNOS (12), and COX-2 (13). These were all up-regulated via pSTAT3 in the injured SCN 3 h after PSL (Fig. 4 A–F). The expression levels of mRNAs for MMP-9 and iNOS but not COX-2 were enhanced in the macrophage cell line after leptin treatment (Fig. 4 G–I). The reason why the change in COX-2 expression in the macrophage cell line was different from that in the SCN was not elucidated by the present study. A possible explanation is that in the early phase after peripheral nerve injury, the increase in COX-2 expression is noted in Schwann cells rather than in macrophages (25). Another is that COX-2 expression is increased by synergy with IFN-␥ treatment in the cell line (26). Further studies are needed to clarify the mechanism underlying the PSL-induced increase in COX-2 expression in the SCN. We applied in vitro leptin-treated PMs to the SCN of ob/ob mice that had received PSL to determine the critical role of leptin-stimulated macrophages in the development of tactile allodynia. Administration of leptin-treated PMs reversed the failure of ob/ob mice to develop PSL-induced tactile allodynia (Fig. 5E). These results strongly suggest that the distribution of leptin-stimulated macrophages in the injured SCN is critical for the development of PSL-induced tactile allodynia. The effects of leptin-treated PMs are consistent with the effect of leptin replacement on the absence of tactile allodynia in ob/ob mice (Fig. 1D). Moreover, these results raise the possibility that macrophages are the key target of leptin during the development of 13080 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0903524106
tactile allodynia. On the other hand, the administration of leptintreated PMs did not induce tactile allodynia in sham-operated ob/ob mice (Fig. 5E), which was also consistent with the effect of perineural injection of leptin (Fig. 1D). These results suggest that leptin-stimulated macrophages play a critical role in the early phase of PSL-induced tactile allodynia but not in the late phase. The behavioral data presented here suggest that leptin is more important in mediating tactile allodynia than thermal hyperalgesia. Different pain modalities have distinct mechanisms (27). Some studies have reported that tactile allodynia and thermal hyperalgesia are not both always elicited by peripheral nerve injuries. IL-6 knockout mice show an absence of tactile allodynia, but normal thermal hyperalgesia in these mice can be induced by nerve injury (28). IL-6 stimulates the glycoprotein 130 (gp130)-JAK-STAT pathway, as does leptin (6); thus, the development of specific modalities of abnormal pain might depend on the type of intracellular signaling being recruited. This possibility needs to be studied in the future. In summary, this report demonstrates that leptin in the PNS is a novel mediator of tactile allodynia induced by peripheral nerve injury. Our results established that adipocytes in the epineurium of SCN may have a crucial role in synthesizing and secreting leptin into the perineurium and endoneurium. The present study shows that blocking locally elevated signaling of leptin constitutes a therapeutic approach for the treatment of chronic pain. It is significant to bear in mind that the pathophysiology associated with the overexpression and hypersecretion of leptin—namely, obesity and diabetes mellitus—might induce subjects to have increased susceptibility to neuropathic pain: a higher prevalence of peripheral neuropathy has been reported in subjects with obesity and impaired glucose tolerance (29). Obese patients with non-insulin-dependent diabetes mellitus have shown poor results in clinical neuropathy tests (30). These reports suggest that body mass control, which could lower the levels of leptin in adipocytes, may be a therapeutic approach to reduce diabetic neuropathy. Materials and Methods Subjects. The procedures used in these studies were approved by the Animal Research Committee of Wakayama Medical University in accordance with Japanese Government Animal Protection and Management Law, Japanese Government Notification on Feeding and Safekeeping of Animals, and The Guidelines for Animal Experiments in Wakayama Medical University (approval nos. 271 and 322). For the behavioral tests and PM samples, 5-week-old male, leptin-deficient, obese ob/ob mice (B6.V-Lepob/J) and their lean littermates (12 weeks old) were obtained from Charles River. We found that the body weight in ob/ob mice at 5 weeks of age was similar to that in lean mice at 12 weeks of age (29.5 ⫾ 1.5 g in lean mice and 30.1 ⫾ 2.5 g in ob/ob mice, n ⫽ 8; P ⫽ 0.875). Male ICR mice (5 weeks old, 27.1 ⫾ 2.2 g in body weight, n ⫽ 8) were used for other experiments. Mice were subjected to PSL according to the method described by Maeda et al. (31). Behavioral studies were conducted by using the modified procedures described by Maeda et al. (31), as shown in the SI Methods. Drug Administration. AG490 (Sigma–Aldrich) at 40 mg/kg was administered i.p. to ICR mice 30 min before PSL. For perineural injection, mice were injected percutaneously at the SCN under brief anesthesia with ether to study the role of leptin in allodynia. A single injection of 10 L of recombinant mouse leptin (1 g) or a neutralizing antibody against mouse leptin (1 ng; R & D Systems) was given to mouse SCN immediately after PSL and 3 h before behavioral testing, once daily. Vehicle (PBS) or normal goat IgG (10 ng; R & D Systems) was given to the control groups for leptin treatment or the neutralizing antibody, respectively. Western Blotting, Immunohistochemistry, ELISA, RT-PCR, and ChIP. These methods were performed according to the modified methods described by Maeda et al. (31). Brief methods are described in the SI Methods. Macrophage Cell Line in Culture. The murine macrophage J774A.1 cell line (JCRB9108; Health Science Research Resources Bank, Osaka, Japan) was incubated in serum-free medium including vehicle (PBS), LPS (0.1 g/mL), or leptin (16 g/mL).
Maeda et al.
ACKNOWLEDGMENTS. This research was supported by Grant-in-Aid for Scientific Research 18613014 and Grant-in-Aid for Exploratory Research 19659404 from the Japan Society for the Promotion of Science.
1. Scholz J, Woolf CJ (2007) The neuropathic pain triad: Neurons, immune cells and glia. Nat Neurosci 10:1361–1368. 2. Moalem G, Tracey DJ (2006) Immune and inflammatory mechanisms in neuropathic pain. Brain Res Rev 51:240 –264. 3. Verheijen MH, Chrast R, Burrola P, Lemke G (2003) Local regulation of fat metabolism in peripheral nerves. Genes Dev 17:2450 –2464. 4. Tilg H, Moschen AR (2006) Adipocytokines: Mediators linking adipose tissue, inflammation and immunity. Nat Rev Immunol 6:772–783. 5. Zhang Y, et al. (1994) Positional cloning of the mouse obese gene and its human homologue. Nature 372:425– 432. 6. Bjorbaek C, Uotani S, da Silva B, Flier JS (1997) Divergent signaling capacities of the long and short isoforms of the leptin receptor. J Biol Chem 272:32686 –32695. 7. Friedman JM, Halaas JL (1998) Leptin and the regulation of body weight in mammals. Nature 395:763–770. 8. Inoue W, Poole S, Bristow AF, Luheshi GN (2006) Leptin induces cyclooxygenase-2 via an interaction with interleukin-1beta in the rat brain. Eur J Neurosci 24:2233–2245. 9. Castellucci M, et al. (2000) Leptin modulates extracellular matrix molecules and metalloproteinases: Possible implications for trophoblast invasion. Mol Hum Reprod 6:951–958. 10. Konturek PC, et al. (2001) Role of leptin in ulcer healing. Eur J Pharmacol 414:87–97. 11. Chattopadhyay S, Myers RR, Janes J, Shubayev V (2007) Cytokine regulation of MMP-9 in peripheral glia: Implications for pathological processes and pain in injured nerve. Brain Behav Immun 21:561–568. 12. Levy D, Zochodne DW (2004) NO pain: Potential roles of nitric oxide in neuropathic pain. Pain Pract 4:11–18. 13. Ma W, Quirion R (2008) Does COX2-dependent PGE2 play a role in neuropathic pain? Neurosci Lett 437:165–169. 14. Ramji DP, Foka P (2002) CCAAT/enhancer-binding proteins: Structure, function and regulation. Biochem J 365:561–575. 15. Banks AS, Davis SM, Bates SH, Myers MG, Jr. (2000) Activation of downstream signals by the long form of the leptin receptor. J Biol Chem 275:14563–14572. 16. Myers RR, Heckman HM, Rodriguez M (1996) Reduced hyperalgesia in nerve-injured WLD mice: Relationship to nerve fiber phagocytosis, axonal degeneration, and regeneration in normal mice. Exp Neurol 141:94 –101. 17. O’Rourke L, Yeaman SJ, Shepherd PR (2001) Insulin and leptin acutely regulate cholesterol ester metabolism in macrophages by novel signaling pathways. Diabetes 50:955–961.
18. He Y, Chen H, Quon MJ, Reitman M (1995) The mouse obese gene. Genomic organization, promoter activity, and activation by CCAAT/enhancer-binding protein alpha. J Biol Chem 270:28887–28891. 19. Hwang CS, et al. (1996) Transcriptional activation of the mouse obese (ob) gene by CCAAT/enhancer binding protein alpha. Proc Natl Acad Sci USA 93:873– 877. 20. Dong J, et al. (2008) Role of modulator recognition factor 2 in adipogenesis and leptin expression in 3T3–L1 cells. Biochem Biophys Res Commun 366:551–555. 21. Grunfeld C, et al. (1996) Endotoxin and cytokines induce expression of leptin, the ob gene product, in hamsters. J Clin Invest 97:2152–2157. 22. Gruen ML, Hao M, Piston DW, Hasty AH (2007) Leptin requires canonical migratory signaling pathways for induction of monocyte and macrophage chemotaxis. Am J Physiol 293:C1481–C1488. 23. Rechthand E, Rapoport SI (1987) Regulation of the microenvironment of peripheral nerve: Role of the blood-nerve barrier. Prog Neurobiol 28:303–343. 24. Loffreda S, et al. (1998) Leptin regulates proinflammatory immune responses. FASEB J 12:57– 65. 25. Takahashi M, et al. (2004) Cyclooxygenase-2 expression in Schwann cells and macrophages in the sciatic nerve after single spinal nerve injury in rats. Neurosci Lett 363:203–206. 26. Raso GM, et al. (2002) Leptin potentiates IFN-gamma-induced expression of nitric oxide synthase andcyclo-oxygenase-2 in murine macrophage J774A. 1. Br J Pharmacol 137:799 – 804. 27. Julius D, Basbaum AI (2001) Molecular mechanisms of nociception. Nature 413:203– 210. 28. Ramer MS, Murphy PG, Richardson PM, Bisby MA (1998) Spinal nerve lesion-induced mechanoallodynia and adrenergic sprouting in sensory ganglia are attenuated in interleukin-6 knockout mice. Pain 78:115–121. 29. Smith AG, Singleton JR (2008) Impaired glucose tolerance and neuropathy. Neurologist 14:23–29. 30. Straub RH, et al. (1994) Impact of obesity on neuropathic late complications in NIDDM. Diabetes Care 17:1290 –1294. 31. Maeda T, et al. (2008) Pioglitazone attenuates tactile allodynia and thermal hyperalgesia in mice subjected to peripheral nerve injury. J Pharmacol Sci 108:341–347. 32. de Jonge WJ, et al. (2005) Stimulation of the vagus nerve attenuates macrophage activation by activating the Jak2-STAT3 signaling pathway. Nat Immunol 6:844 – 851.
PHARMACOLOGY
PM Injection. PMs were obtained by using a modification of the technique described by de Jonge et al. (32) and were injected as shown in the SI Methods.
Maeda et al.
PNAS 兩 August 4, 2009 兩 vol. 106 兩 no. 31 兩 13081
