Article Text

GFAP palmitoylcation mediated by ZDHHC23 in spinal astrocytes contributes to the development of neuropathic pain
  1. Xiaoqing Fan1,2,3,
  2. Siyu Zhang1,2,
  3. Suling Sun1,2,
  4. Wenxu Bi1,2,
  5. Shuyang Li1,4,
  6. Wei Wang1,4,
  7. Xueran Chen1,2 and
  8. Zhiyou Fang1,2
  1. 1Hefei Cancer Hospital of CAS; Institute of Health and Medical Technology, Hefei Institutes of Physical Science, Chinese Academy of Sciences (CAS), Hefei, Anhui, People's Republic of China
  2. 2Science Island Branch, Graduate School of University of Science and Technology of China, Hefei, Anhui, People's Republic of China
  3. 3Department of Anesthesiology, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China (USTC), Hefei, Anhui, People's Republic of China
  4. 4School of Basic Medicine, Anhui Medical University, Hefei, Anhui, People's Republic of China
  1. Correspondence to Dr Xueran Chen, Hefei Cancer Hospital of CAS; Institute of Health and Medical Technology, Hefei Institutes of Physical Science Chinese Academy of Sciences, Hefei, Anhui, China; xueranchen{at}cmpt.ac.cn; Dr Zhiyou Fang, Hefei Cancer Hospital of CAS; Institute of Health and Medical Technology, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei, Anhui, People's Republic of China; z.fang{at}cmpt.ac.cn

Abstract

Background Cancer pain has a significant impact on patient’s quality of life. Astrocytes play an important role in cancer pain signaling. The direct targeting of astrocytes can effectively suppress cancer pain, however, they can cause many side effects. Therefore, there is an urgent need to identify the specific signaling pathways or proteins involved within astrocytes in cancer pain as targets for treating pain.

Methods A neuropathic cancer pain (NCP) model was established by inoculating mouse S-180 sarcoma cells around the right sciatic nerve in C57BL/6 mice. Spontaneous persistent pain and paw withdrawal thresholds were measured using von Frey filaments. The NCP spinal cord dorsal horn (L4–L6) and mouse astrocyte cell line MA-C were used to study protein palmitoylation using acyl-biotin exchange, real-time polymerase chain reaction, ELISA, western blotting, and immunofluorescent staining.

Results In a cancer pain model, along with tumor growth, peripheral nerve tissue invasion, and cancer pain onset, astrocytes in the dorsal horn of the spinal cord were activated and palmitoyltransferase ZDHHC23 expression was upregulated, leading to increased palmitoylation levels of GFAP and increased secretion of inflammatory factors, such as (C–X–C motif) ligand (CXCL)10 (CXCL-10), interleukin 6, and granulocyte-macrophage colony-stimulating factor. These factors in turn activate astrocytes by activating the signal transducer and activator of transcription 3 (STAT3) signaling pathway. A competitive peptide targeting GFAP palmitoylations was designed to effectively alleviate morphine tolerance in cancer pain treatment as well as cancer pain signaling and inflammatory factor secretion.

Conclusions In a rodent model, targeting GFAP palmitoylation appears to be an effective strategy in relieving cancer pain and morphine tolerance. Human translational research is warranted.

  • analgesics, opioid
  • animal experimentation
  • pain management
  • cancer pain

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Cancer pain has a major impact on patients' quality of life, and astrocytes play an important role in cancer pain signaling. Dissecting specific signaling pathways or proteins involved in cancer pain in astrocytes could be used as targets for pain treatment.

WHAT THIS STUDY ADDS

  • GFAP palmitoylation modification is strongly associated with the development of cancer pain and morphine tolerance. A competitive peptides targeting GFAP palmitoyl modifications not only effectively inhibited GFAP modifications but also significantly alleviated morphine tolerance and cancer pain signaling.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • Our results justify consideration of translational research in humans.

Introduction

Cancer pain, that is, pain caused by a malignant tumor, is one of the most common chronic pathological pains in clinical practice and seriously affects patients' quality of life.1 2 However, given the limitations of current therapies, approximately 45% of patients with cancer pain still do not have effective pain control.3 Therefore, there is an urgent need to analyze the molecular mechanisms underlying the onset and development of cancer pain and identify therapeutic targets to effectively intervene and relieve cancer pain in oncology patients.

In the central nervous system (CNS), over half of the cells are glial cells, including astrocytes, microglia, and oligodendrocytes.4 Astrocytes, which account for 20%–40% of glial cells and are the most numerous glial cells, are involved in various physiological and pathological processes, in addition to providing structural and nutritional support to neurons.5 In response to tissue injury or disease, astrocytes are activated, release glial mediators, and participate in regulating neurological disorders, including neurodegenerative diseases and chronic pain.6 For example, astrocytes communicate directly with each other by forming gap junction protein complexes, a complex structure that allows neighboring cells to freely exchange ions and small cellular solutes.7 8 When neurons in the CNS are activated, astrocytes regulate blood flow through extensive contact with the cerebral vasculature, which can also regulate injurious synaptic transmission and network function through glial cell–neuron or interglial cell interactions. In chronic pain states, astrocytes amplify pain signaling at the spinal cord level by modulating microglial activation and neuronal synaptic transmission.9 In brain, astrocytes are also involved in regulating chronic pain-related aversion and anxiety through mechanisms such as modulation of synapse formation.

Activation of the cytokines TNF-a, interleukin-1β (IL-1β) and the chemokines CCL1, CCl2, CCL3, CXCL10, and so on expressed by both astrocytes and microglia are involved in regulating neuropathic pain.10 The interaction between astrocytes, microglia and neurons is maintained by the secretion of mediators. Primary afferent terminals or neurons release CCL2 and CX3CL1 to induce activation of microglia in the dorsal horn of the spinal cord; microglia induce activation of astrocytes by releasing inflammatory factors such as TNF-a; activated astrocytes release chemokines such as CCL2, CXCL10 to act on receptors on neurons, leading to increased excitatory synaptic transmission; neurons also promote astrocyte activation by releasing CXCL13 or Wnt5.11 12 It is also worth noting that JNK kinase, JAK/signal transducer and activator of transcription 3 (STAT3), NF-kB and Smad in astrocytes play an important role in the generation of glial cell activity.13 14

As astrocytes play an important role in susceptibility to chronic pain, targeting astrocytes may provide the desired effect in treating chronic pain. Astrocytes play an important supportive and protective role in the CNS, and analgesics that inhibit astrocyte function may have other side effects. Therefore, targeting specific signaling pathways and proteins within astrocytes may provide effective analgesia without interfering with normal physiological functions.

Protein palmitoylation is a reversible post-translational modification of proteins, and the modification process is carried out by a family of proteins containing asp–his–his–cys (DHHC) structural domains.15 16 Activation and proliferation of glial cells in the nervous system are closely associated with the occurrence of protein palmitoylation. For example, the hyperpalmitoylation of glial cytoskeletal protein GFAP in astrocytes promotes astrocyte proliferation, and PPT1, a depalmitoylase, regulates GFAP depalmitoylation.17 Palmitoylacyltransferases ZDHHC5 facilitate oligodendrocyte development by palmitoylating and activating STAT3, and STAT3 may be an effector of insufficient myelin formation following ZDHHC5 deletion.18 Activation of astrocytes is an important event in the maintenance of cancer pain, and in this process, GFAP palmitoyl modification is necessary for glial cell activation. However, GFAP palmitoyltransferase has not been revealed yet. Elucidating the mechanism of this modification, finding the palmitoyltransferase and intervening in it would make it possible to alleviate cancer pain.

Here, using a mouse model of neuropathic cancer pain (NCP), we sought to characterize astrocyte activity and GFAP palmitoylation, and its association with pain behaviors. A competitive peptide targeting the palmitoylation of GFAP was designed, which can alleviate the onset of morphine tolerance during cancer pain treatment and effectively block cancer pain transmission and inflammatory factor secretion. This offers some potential ways to intervene in cancer pain and slow morphine tolerance.

Methods

Animals

C57BL/6 mice (half male and half female) weighing 25–30 g were purchased from Chengdu Yaokang Biotechnology Co (Sichuan, China) and maintained under laboratory conditions of temperature, humidity, and light (12:12 dark–light cycle). The rearing procedures were approved by the Committee on Ethics in Science and Technology of the Hefei Institutes of Physical Science, Chinese Academy of Science.

NCP Model

The NCP model was established as described previously.19 Well-grown mouse sarcoma S-180 cells were inoculated into the peritoneal cavities of mice. After 7 days of culture, the mice inoculated with tumor cells were anesthetized with pentobarbital (50 mg/kg), and the peritoneal fluid was extruded. The tumor source was adjusted to 1×107 cells/mL with a serum-free medium, and 0.2 mL (2×106 cells) was inoculated into the muscle tissue of the biceps femoris semitendinosus muscle distal to the rotor of the common branch of the sciatic nerve in the right leg, in the immediate vicinity of the nerve. Sham surgery was performed on additional mice. During sham surgery, aspirated ascites were centrifuged at 10 000 rpm for 10 min, and the supernatant was collected to separate the fluid component from the tumor cells. Similarly, the same volume of supernatant was injected in place of the ascites tumor cells. MRI was performed to confirm the presence of a tumor around the sciatic nerve by anatomic examination.

Intrathecal injection procedure

Mice were injected intrathecally twice daily for six consecutive days with saline (10 µL) or morphine (10 µg/µL, twice daily for 6 days) with or without the GFAP-S1-linked competitive peptide inhibitor transmembrane peptide (50 µg/10 µL R(9)-QLQSLTCDLESLRGT).

Intrathecal injections were administered as previously described.20 Briefly, the mice were placed in the supine position, and the midpoint between the tips of the iliac crest was determined. A Hamilton syringe with a 30-gage needle was inserted into the subarachnoid space of the spinal cord between the L5 and L6 spinous processes. The correct intrathecal injection was confirmed by observing tail flicking. The intrathecal injection did not affect the initial response compared with the latency recorded before the injection.

Assessment of continuous spontaneous pain

Spontaneous pain behavior in mice was assessed as previously described with some modifications.21 Lifting of the affected paws was a characteristic spontaneous pain event, which was assessed before and after inoculation in the sham and NCP mouse models. After 15 min of acclimation the test chamber, the animals were individually placed in a 19×29×13 cm clear-bottom cage for 30 min and observed by two independent scorers. If there were differences in pain behaviors observed by scorers, a third scorer would be scheduled to independently score pain and reassess pain behavior. The position of each hind paw was classified as follows: 0, normal loading; 1, light rolling; 2, only the edge of the paw was touching the ground; 3, paw almost lifted from the ground; 4, paw fully raised; and 5, lick raised paw, and the time spent in each position was recorded. For each minute of observation, a weighted pain score was calculated (t1+2t2 +…+5t5, where tx is the time spent in category x) and averaged over 6 min of observation.

Mechanical hypersensitivity measurement using von Frey filaments

The animals were acclimatized to the test environment for 2 days. One hour after morphine administration, mechanical thresholds were measured using von Frey filaments. Briefly, the mice were placed in a transparent box and allowed to adaptate for 30 min. Two independent experimenters with no prior knowledge of the characteristics of the mice evaluated the experiments. Mechanical pain threshold was assessed using the ‘up-down method’.22 Briefly, von Frey filaments of increasing stiffness (bending force 0.01–2.0 g) were applied to the mouse hind paw for 5 s. The first filament presented was a 0.1 g filament; if the animal withdrew its paw, a smaller filament was presented. The cut-off for the longer filament was 2 g. Measurements were taken before inoculation with S-180 (baseline), on the dates indicated after inoculation, and after drug treatment. Paw withdrawal thresholds (PWTs) were calculated after the completion of a sequence of six consecutive responses. Data are expressed in grams and represent mean thresholds. The antinociceptive response was calculated as the percentage of the maximum possible effect (%MPE) as follows: 100% × ([PWT after administration − PWT before administration]/[2 g PWT before administration]) = %MPE.

GFAP palmitoylation test

Protein palmitoylation was analyzed using acylbiotinyl exchange (ABE) as described previously.23 Cells were lysed using radio immunoprecipitation assay (RIPA) buffer, and then incubated overnight at 4°C with 1 µg of anti-GFAP antibody (#3670, Cell Signaling Technology) and a final concentration of 50 mM N-ethylmaleimide (alkylated free cysteine residues). The next day, the complex was bound to protein A/G beads for 1 hour at room temperature. Then, 500 µL of 1 M hydroxylamine HCl (pH 7.4) was added to the beads for 1 hour at room temperature to break the thioester bond. The beads were resuspended in 30 µL of sodium dodecyl sulfate (SDS) sample buffer without reducing agents (eg, β-mercaptoethanol), boiled at 100°C for 10 min, and then analyzed by western blotting with Horseradish Peroxidase (HRP)-conjugated Streptavidin antibody.

Statistical analysis

All subgroup data are presented as mean±SD GraphPad Prism V.6 software (GraphPad Software, San Diego, California, USA) was used for statistical analyses. Differences between the two groups were assessed using Student’s t-test. Data from more than two groups were evaluated using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test or two-way ANOVA followed by the Bonferroni post hoc test. All experiments were replicated for each sample with at least three biological replicates. The significance criteria (p values) were defined as described in the figure. Statistical significance was defined as *p<0.05, **p<0.01, and ***p<0.001.

Results

Morphine for cancer pain may lead to the development of drug tolerance

As shown in figure 1A, the S-180 cells grew rapidly and integrated around the sciatic nerve, as confirmed by MRI. Inoculation of S-180 cells, but not the vehicle (aspirated ascites supernatant), in C57BL/6 mice-induced pain on day 3, which became more pronounced after day 5 (figure 1B). Consistent with this finding, the inoculation of S-180 cells induced mechanical hyperalgesia (figure 1C). As the tumor grew, the PWT decreased from 1.34±0.201 g before sarcoma inoculation to 0.139±0.172 g around day 22 after inoculation.

Figure 1

Spontaneous pain occurred, and prolonged morphine injection could lead to drug tolerance in the NCP model. (A) MRI scans of the S-180 tumor mass around the sciatic nerve (indicated by red circle). After 21 days inoculation of S-180 tumor cells around the sciatic nerve, MRI was performed. In contrast, the sham mouse at day 21 had no tumor growth. (B) Time course of spontaneous pain after inoculation in sham and NCP models (left). The spontaneous pain score was calculated from observations of paw favoring, lifting, and licking before and after inoculation (right panel), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (C) The withdrawal latency of the right hind paws and tumor volume was measured at the indicated time points after inoculation, and analyzed by Student’s t-test. (D) On days 3, 5, and 7 after the drug injection, spontaneous pain was examined to determine the chronic analgesic effect (left panel). The spontaneous pain score was calculated from observations of paw favoring, lifting, and licking before and 2, 3, and 5 days after the drug injection (right), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (E) Long-term morphine injections for cancer-related pain can result in drug tolerance. Morphine (10 μg) was intrathecally injected twice daily, and the maximal possible effect was measured 1 hour after the first injection each day (n=15). The antinociceptive response was calculated as the percentage of the maximum possible effect, and analyzed by Student’s t-test. ANOVA, analysis of variance; NCP, neuropathic cancer pain.

Morphine is the main clinical drug used to treat cancer pain. Next, we studied the effects of morphine on NCP. On day 5 of morphine treatment (twice daily for 2–6 days), morphine did not effectively inhibit the occurrence of pain (figure 1D) and produced a tolerance response (figure 1E).

STAT3 activation in astrocytes for the NCP model

The occurrence of cancer pain was associated with increased astrocyte activation from dorsal horn of the spinal cord (figure 2A). In contrast, the ventral areas of the spinal cord did not show significant astrocyte activation (online supplemental figure 1). GFAP expression gradually increases with tumor growth and cancer pain development. This may be related to JAK–STAT pathway activation. Isolation of astrocytes from the L4–L6 dorsal horn of the NCP model and RNA-seq revealed significant activation of JAK–STAT (figure 2B). Immunofluorescence results also showed significant colocalization of p-STAT3 (Y705) with GFAP in the L4–L6 dorsal horn of the NCP model (figure 2C). Consistent with this, an increase in p-STAT3 (Y705) expression was observed with the development of cancer pain, particularly on days 3–9 after model establishment (figure 2D, and online supplemental figure 2).

Supplemental material

Figure 2

STAT3 activation in the spinal astrocytes of NCP. (A) Immunofluorescence analysis of GFAP expression in the dorsal L4–L6 spinal cord horn of the sham and NCP groups. Spinal samples were collected at 0, 7, and 15 days after inoculation. The percentage of GFAP-positive cells was calculated (n=3), and analyzed by using two-way analysis of variance followed by the Bonferroni post hoc test. Scale bar, 100 µm. (B) GSEA pathway enrichment analysis of signal transduction pathway changes in dorsal horn astrocytes of the spinal cord in a cancer pain model (15 days after inoculation). (C) Immunofluorescence analysis of GFAP (green) and p-STAT3 (Y705; red) expression in the L4–L6 spinal cord horn in the sham and NCP groups. Spinal samples were collected 15 days after inoculation. The percentage of p-STAT3-positive cells among GFAP-positive cells was determined (n=3), and analyzed by Student’s t-test. Scale bar, 100 µm. (D) Western blot analysis of STAT3 and p-STAT3 in the L4–L6 spinal cord horn of the sham and NCP groups (3, 5, 7, 9, 12, and 15 days after inoculation). (E) Western blot analysis of STAT3 and p-STAT3 in microglia BV2, and MA-C astrocytes treated with extracts from the L4–L6 spinal dorsal horn of the NCP model for 16 hours. (F) CCK-8 analysis of cell viability of morphine (200 µM) or Static (20 µM)-treated MA-C astrocytes for 16 hours, and analyzed by Student’s t-test. (G) Western blot analysis of STAT3 and p-STAT3 in the L4–L6 spinal cord horn of the NCP group. Morphine (10 μg) was injected into the mice twice daily, and spinal samples were collected before or 1, 2, 3, 4, and 5 days after the final administration. CCK-8, Cell Counting Kit-8; NCP, neuropathic cancer pain; STAT3, signal transducer and activator of transcription 3.

We cultured mouse microglia (BV2) and astrocytes (MA-C) in a medium supplemented with the extract of the L4–L6 dorsal horn from the NCP model, in which STAT3 activation was more strongly reflected in MA-C cells than that in BV2 cells (figure 2E, and online supplemental figure 3). Notably, similar to the results for the STAT3 inhibitor, Static treatment, morphine treatments significantly inhibited astrocyte viability (figure 2F). However, with prolonged morphine treatment, STAT3 was first inhibited and then activated, which was accompanied by morphine tolerance (figure 2G). Taken together, these results suggest that STAT3 activation plays an important regulatory role in astrocyte-mediated cancer pain in carcinogenesis.

ZDHHC23-mediated upregulation of GFAP palmitoylation in the NCP model

GFAP palmitoylation was upregulated in MA-C cells after treatment with L4–L6 spinal cord dorsal horn extracts from the NCP model compared with that in microglial BV2 cells (figure 3A). In addition, the ZDHHC5, ZDHHC16, ZDHHC18, and ZDHHC23 mRNA levels increased in the extract-treated MA-C cells (figure 3B). To identify the protein S-acylases responsible for GFAP modification, 23 HA-tagged ZDHHC protein S-acylases were separately transfected into MA-C cells, and immunoprecipitation results showed that GFAP palmitoylation was detected after the expression of ZDHHC5, 12, 21, and 23 (figure 3C). In cells expressing the GFAP mutant Cys291 Ala, GFAP palmitoylation was significantly reduced and attenuated by ZDHHC23 inactivation (figure 3D). Consistent with these results, the protein levels of ZDHHC23 and the GFAP palmitoylation levels were higher in the cancer pain model group than in the sham group (figure 3E). Notably, GFAP protein and palmitoylation levels showed an increasing trend with the growth of the transplanted tumors and sciatic nerve compression (figure 3F). These results suggest that DHHC family-mediated GFAP palmitoylation in astrocytes plays an important role in NCP development and progression.

Figure 3

GFAP palmitoylation mediated by ZDHHC23 occurred in astrocytes. (A) GFAP palmitoylation was upregulated in MA-C cells 16 hours after treatment with the extract from the L4–L6 spinal dorsal horn in the NCP model (n=3), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (B) Real-time polymerase chain reaction analysis of DHHC mRNA levels in astrocytic MA-C cells 16 hours after treatment with the cancer pain model extract, and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (C) The immunoprecipitation results demonstrated that GFAP interacted with ZDHHC5, 12, 21, and 23. MA-C cells were transfected with constructs encoding HA-DHHCs in a six-well plate. Cell lysates were harvested for immunoprecipitation using an anti-HA antibody and western blot analysis with GFAP antibody. (D) GFAP palmitoylation was mediated by ZDHHC23. MA-C cells were infected with ZDHHC5, 12, 21, or 23 siRNAs, or transfected with GFAP C291A mutant constructs for 48 hours, and GFAP palmitoylation was analyzed using the ABE method. The percentage of GFAP palmitoylation in treated MA-C cells was determined (n=3), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (E) ZDHHC23 was upregulated in the NCP model (12 days postinoculation) compared with the sham group. (F) The incidence of GFAP palmitoylation in the NCP model was positively correlated with tumor growth. GFAP expression and palmitoylation levels were analyzed using the ABE method and immunoblotting at the indicated tumor volumes. ABE, acyl-biotin exchange; ANOVA, analysis of variance; CCK-8, Cell Counting Kit-8DHHC, asp–his–his–cys; NCP, neuropathic cancer pain.

GFAP palmitoylation promoted astrocyte activation

ZDHHC23-mediated GFAP palmitoylation promoted cell viability, astrocyte growth, and branching (figure 4A,B). ZDHHC23 knockdown and GFAP palmitoylation inhibition suppress astrocyte activation.

Figure 4

GFAP palmitoyl modifications facilitate astrocyte activation and secretion of inflammatory factors. (A) Inhibition of GFAP palmitoylation suppresses astrocytic MA-C cell viability. The viability of wild-type MA-C cells transfected with ZDHHC23 siRNA, treated with the palmitoylation inhibitor 2 BP for 48 hours, or stably transfected with GFAP (C291A) was analyzed using the CCK-8 assay, and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (B) Extracts from the cancer pain model resulted in astrocyte activation, and the inhibition of GFAP palmitoylation suppressed this process. MA-C cells transfected with ZDHHC23 siRNA, treated with the palmitoylation inhibitor 2 BP for 48 hours, or stably transfected with GFAP (C291A) were incubated with or without the NCP model extract and analyzed by immunofluorescence staining. Scale bar, 200 µm. The number of branches of each astrocyte in the treated MA-C cells was determined (n=3), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (C) Inhibition of GFAP palmitoylation suppressed the release of inflammatory cytokines in MA-C astrocytes treated with the NCP extract of the NCP model. MA-C cells transfected with ZDHHC23 siRNA, treated with the palmitoylation inhibitor 2 BP for 48 hours, or stably transfected with GFAP (C291A) were incubated with or without the extract from the NCP model, and the inflammatory factors IL-6, CXCL-10, and GM-CSF were analyzed using ELISA. (D) Western blot analysis of STAT3 and p-STAT3 in MA-C cells transfected with ZDHHC23 siRNA, treated with palmitoylation inhibitor 2 BP for 48 hours, or stably transfected with GFAP (C291A). p-STAT3 expression levels were determined (n=3), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. ANOVA, analysis of variance; CCK-8, Cell Counting Kit-8GM-CSF, granulocyte-macrophage colony-stimulating factor; IL-6, interleukin 6; NCP, neuropathic cancer pain; PBS, phosphate buffered solution.

Protein palmitoylation promoted astrocyte activation and the release of inflammatory factors from astrocytes, IL-6, CXCL-10, and granulocyte-macrophage colony-stimulating factor (GM-CSF) in MA-C cells (figure 4C). Furthermore, ZDHHC23 inactivation or GFAP palmitoylation inhibition inhibited the release of IL-6, CXCL-10, and GM-CSF.

In MA-C cells treated with the extract, the STAT3 (Y705) phosphorylation level increased along with GFAP palmitoylation upregulation (figure 4D). Consistent with this, the inactivation of ZDHHC23 or inhibition of GFAP palmitoylation significantly reduced STAT3 activation. Thus, we identified a positive feedback pathway through which activated astrocytes upregulate the expression of palmitoyltransferase ZDHHC23, which in turn modifies GFAP, promotes the activation and secretion of IL-6 by astrocytes, and activates the STAT3 signaling pathway, which in turn promotes cell proliferation.

Targeting GFAP palmitoyl modifications effectively attenuates morphine tolerance in cancer pain

After determining the GFAP palmitoylation pattern, we introduced a competitive inhibitor of GFAP palmitoylation into the NCP model. We synthesized a cell-penetrating peptide (GFAP-S1) containing a GFAP palmitoylation sequence (285–299) and a control peptide (GFAP-S0) with a Cys 291 Ala mutation in vitro. Furthermore, ABE results showed that GFAP-S1 reduced GFAP palmitoylation and expression (figure 5A). Importantly, GFAP expression and STAT3 activation were attenuated (figure 5A,B). GFAP-S1 effectively inhibited astrocyte growth and branching (figure 5C).

Figure 5

Targeting GFAP palmitoylations could effectively alleviate cancer pain. (A) Acyl-biotin exchange analysis of GFAP palmitoylation in astrocyte MA-C cells treated with GFAP-S0 (25 µg/mL) or GFAP-S1 (25 µg/mL). GFAP palmitoylation levels were determined (n=3), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (B) Western blot analysis of STAT3 and p-STAT3 expression in astrocyte MA-C cells treated with GFAP-S0 (25 µg/mL) or GFAP-S1 (25 µg/mL), and analyzed by one-way ANOVA followed by Tukey’s multiple comparison test. (C) Immunofluorescence analysis of GFAP (green) expression in MA-C astrocytes treated with GFAP-S0 (25 µg/mL) or GFAP-S1 (25 µg/mL). The percentage of GFAP-positive cells was calculated (n=3), and analyzed by Student’s t-test. Scale bar, 200 µm. (D) Immunofluorescence analysis of GFAP (green) expression in the L4–L6 spinal cord horn in the NCP group. NCP groups (15 days after inoculation) were injected with GFAP-S0 (50 µg) or GFAP-S1 (50 µg) twice daily and examined on day 7 after the final administration. The percentage of GFAP-positive cells was calculated (n=3), and analyzed by Student’s t-test. Scale bar, 200 µm. (E) ELISA analysis of inflammatory cytokines, TNFα, IL-1β, and IL-17, in the L4–L6 spinal cord horn of the NCP and sham mouse groups. Morphine (10 μg) with or without GFAP-S0 (50 µg) or GFAP-S1 (50 µg) was injected into the mice twice daily, and spinal samples were collected 3 and 7 days after the final administration for the experiment on chronic morphine tolerance. (F) Time course of spontaneous pain in sham and NCP models (left). The spontaneous pain score was calculated from observations of paw favoring, lifting, and licking before and after administration (right). Morphine (10 μg) with or without GFAP-S0 (50 µg) or GFAP-S1 (50 µg) was injected into mice twice daily, and spontaneous pain was examined 3, 5, and 7 days after the final administration for chronic morphine tolerance experiments. (G) The withdrawal latency of the right hind paw and tumor volume was measured at the indicated time points after administration. ANOVA, analysis of variance; DMSO, dimethyl sulfoxide; IL, interleukin; NCP, neuropathic cancer pain; PBS, phosphate buffered solution; STAT3, signal transducer and activator of transcription 3.

Next, we tested the effect of the competing peptide, GFAP-S1, on cancer pain. On day 7 of continuous administration, GFAP-S1 significantly inhibited astrocyte activation compared with GFAP-S0 administration (figure 5D). Similarly, on the third day of continuous administration, morphine or the combination of morphine and GFAP-S0 significantly inhibited the release of the inflammatory factors CXCL-10, IL-6, and GM-CSF, as well as the onset of pain; however, on the seventh day, morphine or the combination of morphine and GFAP-S0 developed resistance, and the treatment failed (figure 5E,F). However, the combination of morphine and GFAP-S1 showed better inhibition of inflammatory factor release and pain on the third, fifth and seventh days after continuous administration. Consistent with this, the MPE results showed that GFAP-S1 significantly attenuated chronic morphine tolerance (figure 5G).

This increased the palmitoylation of GFAP while activating astrocytes in the NCP model, promoting the secretion of inflammatory factors, and enhancing pain signaling (figure 6). Long-term morphine administration leads to drug tolerance and is ineffective in alleviating cancer pain. Competitive peptides targeting GFAP palmitoylation may effectively attenuate morphine tolerance, inhibit the upregulation of GFAP palmitoylation and inflammatory factor release from astrocytes, and reduce cancer pain.

Figure 6

Illustration of the crosstalk between GFAP palmitoylation and NCP. During the occurrence and development of NCP, the expression of palmityltransferase ZDHHC23 is increased, and palmitylated GFAP promotes glial cell activation and IL-6 secretion; IL-6 further increases astrocyte proliferation and activation via the STAT3 signaling pathway. Competitive peptides targeting GFAP palmitoylation may effectively attenuate morphine tolerance, inhibit GFAP palmitoyl changes and inflammatory factor release from astrocytes, and reduce cancer pain. GM-CSF, granulocyte-macrophage colony-stimulating factor; IL-6, interleukin 6; NCP, neuropathic cancer pain; STAT3, signal transducer and activator of transcription 3.

Discussion

Astrocytes, the most abundant cell class in the mammalian brain, are the largest glial cell type by volume. Such numerous astrocytes play an important role in CNS injury and central degenerative lesions, regulating astrocyte morphology, migratory ability, maintenance of internal environmental homeostasis, and the ability to synthesize and release neurotrophic factors, providing a new therapeutic approach for treating pain and controlling chronic pain formation.

Recent studies have shown that a common phenomenon in the pathogenesis of pathological pain, particularly cancer pain, is central sensitization, that is, cellular remodeling and altered synaptic plasticity of neurons in the CNS.24 Previous studies have shown evidence of astrocyte activation at the spinal cord level in models of inflammatory, neuropathic, and bone cancer pain. The classical physiological view is that sensory signals, such as pain, are transmitted only between nerve cells, and glial cells in the CNS do not function to transmit nociceptive signals, but only play a nutritional and support role for neurons.10 As the mechanisms of pain continue to be elucidated, increasing evidence suggests that glial cells play an indispensable role in neuroimmunity and neuronal signaling regulation. Glial cells play an important regulatory role in neuronal synapse formation, synaptic plasticity, and transmission efficiency and are closely related to the generation of central sensitization and maintenance of pain.25 At the spinal cord level, the primary center of nociceptive transmission, glial cells can directly regulate the excitability and function of neurons, as well as the activity of neurons through the receptors on their surface and the release of active substances. There is bidirectional signaling between neurons and glial cells, and this transmission is highly plastic. At the spinal cord level, activated astrocytes release various pro-nociceptive agents, such as NO, IL-6, TNF-α, reactive oxygen species and neurotrophic factors, which act on the presynaptic membrane to promote the release of more neurotransmitters and on the postsynaptic membrane to increase the excitability of postsynaptic neurons and facilitate the transmission of pain signals to the center.26 The pro-nociceptive substances mentioned above are particularly important in pain signaling because IL-6 is an activating ligand of the JAK–STAT pathway, and its binding to cytokine receptors on the cell membrane activates the JAK2–STAT3 pathway. The cytoplasmic nuclear transporter protein α transports active STAT3 into the nucleus, where p-STAT3 recognizes specific nucleic acid sequences, initiating transcription of target genes.27 Gene expression is followed by glial cell activation and proliferation, creating a positive feedback loop that leads to central sensitization. In this study, NCP model mice showed nociceptive sensitization accompanied by increased expression of p-STAT3 and activation of astrocytes in the dorsal horn of the spinal cord. These studies suggest that the IL-6-JAK-STAT pathway in the spinal cord may play a critical role in regulating nociceptive transmission by astrocytes in neurons.

Astrocytes produce GFAP, their signature protein, whose expression reflects astrocyte activation and proliferation. It should be emphasized that glial cell activation is a common pathological feature of several neurological disorders (neurodegenerative diseases, stroke or traumatic brain injury, etc). Three characteristic features of astrocyte activation have been identified: the upregulation of GFAP expression, massive proliferation of astrocytes, and hypertrophy of cell morphology. PPT1 effectively downregulated GFAP palmitoylation.17 When PPT1 was absent or knocked down, GFAP palmitoylation increased dramatically, leading to accelerated glial cell proliferation. In this study, GFAP palmitoylation was found to increase in astrocytes in the dorsal horn of the spinal cord along with tumor growth and invasion of the sciatic nerve, and the increased GFAP palmitoylation level led to accelerated glial cell proliferation and glial cell activation, which may contribute to the process of neuropathic pain.

Palmitoylation is a reversible modification of protein lipids, in which S-type palmitoylation is manifested by the covalent modification of a 16-carbon palmitoyl group to the sulfhydryl group of a protein cysteine through a thioester bond, which regulates protein trafficking, stability, and binding to cell membranes.28 Palmitoylation is catalyzed by a class of palmitoyl acyltransferases containing DHHC in the active center; the DHHC family has over 20 members in mammals.29 However, the role of palmitoyltransferases, which mediates GFAP modification, has not yet been reported. Further studies showed that GFAP significantly interacts with the palmitoyltransferase ZDHHC23 in astrocytes and that ZDHHC23 expression knockdown significantly downregulated GFAP palmitoyl modification levels. This suggests that ZDHHC23 is an important regulator of palmitoyl modification of GFAP.

Since current small-molecule inhibitors of palmitoylation, such as 2-bromopalmitate, cerulenin, and tunicamycin, lack selectivity for DHHC enzymes, this study attempted to design competitive palmitoylation inhibitors.30 31 Because the specificity of DHHC enzymes is determined by the substrate sequence, sequences adjacent to the modified site of GFAP can be modified by ZDHHC23, thereby reducing the modification of endogenous GFAP. Theoretically, this competitive peptide has less of an effect on other DHHC members and may, therefore, be more selective than the current palmitoylated small-molecule inhibitors. Therefore, the GFAP-competitor peptide should be derivatized and modified in the future to further improve its activity, stability, and drug-forming potential.

Conclusion

In conclusion, in a cancer pain model, tumor growth in the sciatic nerve increased the expression of GFAP, a marker of astrocytes in the dorsal glia of the spinal cord, and the levels of palmitoyl modifications, while activating glial cells to secrete inflammatory factors such as IL-6, which in turn further promoted astrocyte proliferation and activation. The design of competitive peptides targeting GFAP palmitoyl modifications not only effectively inhibited GFAP modifications but also significantly alleviated morphine tolerance and cancer pain signaling. These results provide ideas and approaches for the treatment of pain and suffering in patients with cancer .

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Not applicable.

Ethics statements

Patient consent for publication

Ethics approval

The rearing procedures were approved by the Committee on Ethics in Science and Technology of the Hefei Institutes of Physical Science, Chinese Academy of Sciences.

Acknowledgments

We thank the members of the technical assistance team at the Institute of Health and Medical Technology, Hefei Institute of Physical Science, and Chinese Academy of Sciences. We thank Dr Fukata for providing 23 plasmids encoding HA-tagged ZDHHC family proteins.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors XQF, XRC, and ZYF conceived and designed the experiments; XQF, SYZ, SSL, WXB, SYL, and WW conducted the experiments; XQF, SYZ, SSL, and XRC analyzed the data; and XQF and SSL wrote the manuscript. All the authors have read and approved the final version of the manuscript. ZYF, and XRC acted as guarantors to accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.

  • Funding This research was supported by the National Natural Science Foundation of China (grant numbers 82172663, 82104208, 81872066, 81773131, and 81972635), the Innovative Program of the Development Foundation of the Hefei Center for Physical Science and Technology (grant number 2021HSC-CIP011), and CASHIPS Director's Fund (YZJJ2023QN50, and YZJJ2022QN49).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.