Intrathecal delivery of PDGF produces tactile allodynia through its receptors in spinal microglia

Neuropathic pain is a debilitating pain condition that occurs after nerve damage. Such pain is considered to be a reflection of the aberrant excitability of dorsal horn neurons. Emerging lines of evidence indicate that spinal microglia play a crucial role in neuronal excitability and the pathogenesis of neuropathic pain, but the mechanisms underlying neuron-microglia communications in the dorsal horn remain to be fully elucidated. A recent study has demonstrated that platelet-derived growth factor (PDGF) expressed in dorsal horn neurons contributes to neuropathic pain after nerve injury, yet how PDGF produces pain hypersensitivity remains unknown. Here we report an involvement of spinal microglia in PDGF-induced tactile allodynia. A single intrathecal delivery of PDGF B-chain homodimer (PDGF-BB) to naive rats produced a robust and long-lasting decrease in paw withdrawal threshold in a dose-dependent manner. Following PDGF administration, the immunofluorescence for phosphorylated PDGF β-receptor (p-PDGFRβ), an activated form, was markedly increased in the spinal dorsal horn. Interestingly, almost all p-PDGFRβ-positive cells were double-labeled with an antibody for the microglia marker OX-42, but not with antibodies for other markers of neurons, astrocytes and oligodendrocytes. PDGF-stimulated microglia in vivo transformed into a modest activated state in terms of their cell number and morphology. Furthermore, PDGF-BB-induced tactile allodynia was prevented by a daily intrathecal administration of minocycline, which is known to inhibit microglia activation. Moreover, in rats with an injury to the fifth lumbar spinal nerve (an animal model of neuropathic pain), the immunofluorescence for p-PDGFRβ was markedly enhanced exclusively in microglia in the ipsilateral dorsal horn. Together, our findings suggest that spinal microglia critically contribute to PDGF-induced tactile allodynia, and it is also assumed that microglial PDGF signaling may have a role in the pathogenesis of neuropathic pain.

of the spinal cord [4][5][6][7][8]. There is a rapidly growing body of evidence indicating that peripheral nerve damage activates glial cells in the dorsal horn and results in changing expression and activity of various molecules [9,10]. Importantly, pharmacological, molecular and genetic manipulations of the function or expression of glial molecules have been shown to substantially influence nerve injury-induced pain behaviors and hyperexcitability of the dorsal horn pain pathway [11][12][13][14][15]. Therefore, signaling between neurons and glia might critically contribute to the pathologically enhanced pain processing in the dorsal horn that underlies neuropathic pain. However, the mechanisms underlying neuropathic pain caused by neuron-glia communications in the dorsal horn remain to be fully elucidated.
Although PDGF signaling is commonly known to have essential roles during development [23], there is limited evidence for its role in the mature CNS. A recent study has shown that PDGF is expressed in dorsal horn neurons in adult mice, and that intrathecal administration of either a selective inhibitor of PDGFR phosphorylation or an antibody trapping endogenous PDGF suppresses thermal hyperalgesia and tactile allodynia after peripheral nerve injury [24]. Thus, PDGF released from dorsal horn neurons is implicated in neuropathic pain. However, how PDGF produces pain hypersensitivity remains unknown.
To investigate this, we first examined whether the intrathecal delivery of PDGF produces tactile allodynia in adult naive rats. We used the PDGF-BB isoform in all experiments in this study because PDGF-B chain expression is induced after peripheral nerve injury [25], neurons throughout the CNS contain the PDGF-B chain [26], and the PDGF-B chain activates both PDGFRα and PDGFRβ [21,23]. We found that a single intrathecal administration of PDGF-BB (0.1, 1 and 10 pmol) produced marked and long-lasting tactile allodynia: the paw withdrawal threshold in response to mechanical stimulation applied to the hindpaw progressively decreased over the first 3 days, reaching the lowest in the threshold on day 3, and this decrease persisted at least for 14 days after PDGF-BB administration (P < 0.001) (Figure 1). The PDGF-BBinduced tactile allodynia was dose dependent (Figure 1).
To identify the cell types on which intrathecally delivered PDGF-BB acts, we performed immunohistochemical experiments using an anti-phospho-Tyr1021 PDGFRβ antibody (p-PDGFRβ) that recognizes activated receptors [27]. The immunofluorescence for p-PDGFRβ in the dorsal horns of vehicle-treated rats remained at low levels, but was markedly increased 30 min after intrathecal PDGF-BB (10 pmol) administration (P < 0.001) ( Figure  2A, B). By double-staining with cell type-specific markers, we found that almost all p-PDGFRβ-positive cells were double-labeled with OX-42 (a marker of microglia), but not with GFAP (an astrocyte marker), CC1 (an oligodendrocyte marker), MAP2 or NeuN (neuronal markers) ( Figure 2C). These results indicate that activation of PDGFRβ evoked by intrathecally delivered PDGF-BB occurs specifically in microglia. Consistently, both PDG-FRα and PDGFRβ mRNAs were detected in primary cultured microglia and in the spinal cord as well as positive A single intrathecal PDGF-BB administration produces tactile allodynia Figure 1 A single intrathecal PDGF-BB administration produces tactile allodynia. The paw withdrawal thresholds in response to mechanical stimuli were measured in rats intrathecally administered vehicle (n = 5) or PDGF-BB (0.1, 1 and 10 pmol, n = 5) just before administration (day 0) and 1, 2, 3, 5, 7, 10, and 14 days after administration. Data represent the means ± SEM of the thresholds. ***P < 0.001, **P < 0.01, *P < 0.05 vs vehicle group by repeated measures twoway ANOVA with Bonferroni post-hoc tests. control tissues (cerebral cortex and spleen) ( Figure 2D). Furthermore, applying PDGF-BB (50 ng/ml) to primary cultured microglial cells enhanced the immunofluorescence for p-PDGFRβ ( Figure 2E).

PDGF-BB phosphorylates its receptors in spinalmicroglia
Because the half life of PDGF in vivo is extremely short [28], it is predicted that PDGF-induced long-lasting tactile allodynia might be due to plastic changes in the spinal cord, especially in microglia. To investigate the status of microglia in the dorsal horn after PDGF stimulation, we performed immunohistochemical analysis using the microglia marker Iba1. After PDGF-BB (10 pmol) administration, the number of Iba1-positive cells in the dorsal horn was significantly increased on day 3 and day 7 compared with vehicle-treated controls (P < 0.01) ( Figure 3A, B). Iba1-positive microglia in the dorsal horns of PDGF-BB-administered rats also showed an increase in Iba1 labeling and a tendency toward a hypertrophied morphology ( Figure 3C). We also observed an increase in the level of expression of the proinflammatory cytokine interleukin-1β (IL-1β) (P < 0.05) ( Figure 3D). Because these changes are consistent with the criteria for activated microglia in vivo, it is suggested that spinal microglia are activated by PDGF-BB.
To examine whether microglia are involved in PDGF-BBinduced tactile allodynia, we tested the effect of minocycline, which inhibits microglia activation [29,30], on the decrease in the paw withdrawal threshold after PDGF-BB administration. Daily intrathecal administration of minocycline (100 μg) from one day before PDGF-BB (10 pmol) administration significantly suppressed the decrease in paw withdrawal threshold (P < 0.05, day 3; P < 0.01, other testing days) ( Figure 3E). This finding suggests that spinal microglia are involved in PDGF-BB-induced tactile allodynia. The mechanisms underlying the anti-allodynic effect of minocycline remains unclear, but we found that minocycline did not inhibit PDGF-induced PDGFRβ phosphorylation in the dorsal horn ( Figure 3F), indicating that minocycline does not directly interrupt the PDGF binding to the PDGFRβ and PDGFRβ dimerization and autophosphorylation. Thus, it is conceivable that minocycline may produce its anti-allodynic effect through inhibiting the downstream consequences of PDGFRβ phosphorylation in microglia including p38 mitogen-activated protein kinase that is an important signaling molecule in tactile allodynia [11,15] and is also known as one of targets of minocycline [31,32].
Purinergic receptors expressed in microglia (P2X 4 , P2X 7 and P2Y 12 ) are implicated in neuropathic pain [12,13,33,34]. Thus, we examined the role of these receptors in PDGF-BB-induced allodynia. After intrathecal PDGF-BB (10 pmol) administration, the level of mRNA expression of P2X 4 receptor in the spinal cord was signifi-cantly increased on day 3 (P < 0.05) ( Figure 4A). By contrast, the mRNA levels of P2X 7 and P2Y 12 receptors were not changed. Furthermore, intrathecally administered TNP-ATP (30 nmol), an antagonist of P2X receptor subtypes P2X 1-4 receptors, produced a significant attenuation of the decreased paw withdrawal threshold on day 7 after PDGF-BB (10 pmol) administration (P < 0.05) ( Figure  4B). Considering that the anti-allodynic effect of TNP-ATP was weak, these results suggest that P2X 4 receptors in the spinal cord are involved, at least in part, in the PDGF-BBinduced tactile allodynia.
Activation of PDGFRs in the spinal cord is implicated in tactile allodynia after peripheral nerve injury [24]. Thus, we determined the type of cells in which PDGFRβ activation occurs under a neuropathic pain condition. In contrast to the contralateral dorsal horn, where p-PDGFRβ immunofluorescence was low, we observed strong p-PDGFRβ immunofluorescence in the dorsal horn ipsilateral to the nerve injury; the level of p-PDGFRβ immunofluorescence in individual cells in this region was also much higher than that in individual cells in the dorsal horn contralateral to the nerve injury (P < 0.01) ( Figure  5A, B). Furthermore, almost all p-PDGFRβ-positive cells were also labeled for the microglia marker OX-42 ( Figure  5C). These results indicate that PDGFRβ activation in the dorsal horn occurs exclusively in microglia after nerve injury. How PDGFR activity is enhanced remains unclear, but we examined the time course for changes in the expression levels of PDGFR mRNAs after nerve injury and found no significant change during the period from 1 day to 14 days post-nerve injury ( Figure 5D). It is thus possible that the enhanced PDGFR activity might be due to an increase in the level of endogenous PDGF within the dorsal horn after nerve injury, as suggested by a previous study [24]. Consistently, low levels of PDGFRβ phosphorylation in the dorsal horns of normal rats ( Figure 2A) and in the contralateral dorsal horns of nerve-injured rats (Figure 5A) were observed. In the adult spinal cord, PDGF has been shown to be expressed in dorsal horn neurons [24]. It is thus assumed that PDGF might be a candidate for signaling molecules between neurons and microglia, thereby producing tactile allodynia, although further investigations are needed to determine the pattern and change in the expression of endogenous PDGFRβ ligands in the dorsal horn after nerve injury.
PDGFRs in the CNS have been previously reported to be expressed in O-2A progenitor cells, oligodendrocytes, and neurons [35][36][37][38]. In the present study, by showing that acute PDGF stimulation in vivo in adult rats induced PDG-FRβ phosphorylation specifically in microglia, in addition to our results in in vitro experiments using cultured microglia, we provide the first evidence that microglia are the predominant cell type expressing functional PDGFRβs in the spinal cord. We further revealed that spinal microglia may mediate tactile allodynia caused by intrathecal administration of PDGF. Recently, Narita et al. [24] have shown that inhibiting PDGFR phosphorylation results in suppression of tactile allodynia after peripheral nerve injury, implying a crucial role for PDGF signaling in neuropathic pain. Notably, following peripheral nerve injury, a marked enhancement of PDGFRβ phosphorylation in dorsal horn microglia also occurred in a cell type-specific manner, indicating that spinal microglia may be crucial for PDGFR-mediated tactile allodynia under neuropathic pain conditions. It remains unknown how PDGF-stimulated microglia modulate pain processing in the dorsal horn, but we found an increase in the expression of IL-1β mRNA in the dorsal horn after PDGF administration. IL- Therefore, IL-1β may be a candidate intermediary molecule between PDGF-stimulated microglia and dorsal horn neurons that contributes to central hypersensitization. Further investigation using microglia-specific IL-1βknockout mice will clarify this issue.

Animals
Male Wistar rats (250-280 g, Japan SLC) were used. Rats were housed at a constant temperature of 23 ± 1°C with a 12 h light-dark cycle (light on 8:00 to 20:00) and fed food and water ad libitum. All of the animals used in the present study were obtained, housed, cared for, and used in accordance with the guidelines of Kyushu University.

Microglia culture
Rat primary cultured microglia was prepared according to the method described previously [44]. In brief, the mixed glial culture was prepared from neonatal Wistar rats and maintained for 9-15 days in DMEM with 10% FBS. Microglia were obtained as floating cells over the mixed glial culture. The floating cells were collected by gentle shaking and transferred to culture dishes for each experiment.

Neuropathic pain model and Behavioral tests
The left L5 spinal nerve of rats was tightly ligated with 5-0 silk suture and cut just distal to the ligature under 2% iso- The paw withdrawal thresholds before (pre) and 7 days (day7) after intrathecal PDGF-BB (10 pmol) administration (n = 8). Then TNP-ATP (30 nmol) and vehicle (PBS) was intrathecally administered on day 7 and the changes in the paw withdrawal thresholds were measured (n = 4). Data are means ± SEM of the thresholds. ### P < 0.001 vs pre, *P < 0.05 vs vehicle group by Student's t-test.

ATP receptors relation to PDGF-BB-induced allodynia
Immunofluorescence of phosphorylated PDGF β-receptors and expression of PDGF receptor mRNAs in rats after nerve injury flurane anesthesia [12,45]. To assess the level of tactile allodynia, rats were placed individually in a wire mesh cage and habituated for 30-60 min to allow acclimatization to the new environment. From below the mesh floor, calibrated von Frey filaments (0.4-15 g; North Coast Medical, Morgan Hill, California, USA) were applied to the mid-plantar surface of the hindpaw. The 50% paw withdrawal threshold was determined using the up-down method [46].

Immunocytochemistry
Primary microglial cells were seeded on aminopropyltriethoxysilane-coated glass (Matsunami, Osaka, Japan) at 5 × 10 4 cells/well and incubated for 1 h. After the culture media were replaced with serum-free media, cells were incubated for 2 h and subsequently treated with PBS as a control or 50 ng/ml PDGF-BB for 10 min [47], and then fixed in 3.7% formaldehyde in PBS for 30 min at 25°C. The cells were permeabilized and blocked by incubating them with blocking solution (3% normal goat serum and 0.3% Triton X-100 in PBS) for 15 min at 25°C, and then incubated overnight at 4°C with the primary antibodies against phospho-PDGF β-receptor (1:400) and OX-42 (1:1000). After washing, the cells were incubated for 1 h with appropriate fluorescent-conjugated secondary antibodies (goat anti-rabbit IgG-conjugated Alexa Fluor 488 or goat anti-mouse IgG-conjugated Alexa Fluor 546, 1:1000) and coverslipped in Vectashield containing 4',6diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA). Fluorescent images were obtained and analyzed as mentioned above.