Baricitinib ameliorates inflammatory and neuropathic pain in collagen antibody-induced arthritis mice by modulating the IL-6/JAK/STAT3 pathway and CSF-1 expression in dorsal root ganglion neurons

doi:10.21203/rs.3.rs-4119106/v1

Download PDF

Research Article

Baricitinib ameliorates inflammatory and neuropathic pain in collagen antibody-induced arthritis mice by modulating the IL-6/JAK/STAT3 pathway and CSF-1 expression in dorsal root ganglion neurons

https://doi.org/10.21203/rs.3.rs-4119106/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Jun, 2024

Read the published version in Arthritis Research & Therapy →

You are reading this latest preprint version

Background Janus kinase (JAK) inhibitors, such as baricitinib, are widely used to treat rheumatoid arthritis (RA). Clinical studies show that baricitinib is more effective at reducing pain than other similar drugs. Here, we aimed to elucidate the molecular mechanisms underlying the pain relief conferred by baricitinib, using a mouse model of arthritis.

Methods We treated collagen antibody-induced arthritis (CAIA) model mice with baricitinib, celecoxib, or vehicle, and evaluated the severity of arthritis, histological findings of the spinal cord, and pain-related behaviors. We also conducted RNA sequencing (RNA-seq) to identify alterations in gene expression in the dorsal root ganglion (DRG) following baricitinib treatment. Finally, we conducted in vitro experiments to investigate the direct effects of baricitinib on neuronal cells.

Results Both baricitinib and celecoxib significantly decreased CAIA and improved inflammatory pain as determined by grip strength, while only baricitinib remarkably suppressed neuropathic pain as determined by the von Frey test. CAIA induction of inflammatory cytokines in ankle synovium, including interleukin (IL)-1β and IL-6, was suppressed by treatment with either baricitinib or celecoxib. In contrast, RNA-seq analysis of the DRG revealed that baricitinib, but not celecoxib, restored gene expression alterations induced by CAIA to the control condition. Among many pathways changed by CAIA and baricitinib treatment, the interferon-alpha/gamma, JAK-signal transducer and activator of transcription 3 (STAT3), and nuclear factor kappa B (NF-κB) pathways were considerably decreased in the baricitinib group compared with the celecoxib group. Notably, only baricitinib decreased the expression of colony-stimulating factor 1 (CSF-1), a potent cytokine that causes neuropathic pain through activation of the microglia–astrocyte axis in the spinal cord. Accordingly, baricitinib prevented increases in microglia and astrocytes caused by CAIA. Baricitinib also suppressed JAK/STAT3 pathway activity and Csf1expression in cultured neuronal cells.

Conclusions Our findings demonstrate the effects baricitinib has on the DRG in relation to ameliorating inflammatory and neuropathic pain.

Rheumatoid arthritis (RA) is a common disorder that occurs together with synovitis and subsequent joint destruction. Previous studies have revealed the involvement of inflammatory cytokines in this pathology, including interleukin (IL)-1β, IL-6, tumour necrosis factor (TNF) and potent signalling pathways, such as the nuclear factor kappa B (NF-κB) and Janus kinase (JAK)-signal transducer and activator of transcription 3 (STAT3) pathways. NF-κB is closely associated with both acute and chronic inflammatory responses. IL-1β and TNF are representative cytokines that activate NF-κB, which in turn upregulates IL-1β, TNF and IL-6 [1, 2]. IL-6 activates the JAK/STAT3 pathway and exerts various effects. IL-6 is also a downstream molecule of the JAK/STAT3 pathway, and the feedback loops of these cytokines and pathways enhance inflammation in RA [1, 2]. The JAK/STAT3 and NF-κB signalling pathways interact and work collaboratively with each other [3].

The management of RA has improved dramatically with the development of antibody-based drugs against TNF and IL-6 (known as biological disease-modifying antirheumatic drugs, DMARDs) and JAK inhibitors, in addition to methotrexate. However, residual pain continues to be observed in a large proportion of patients who are experiencing biological remission of RA [4]. It remains difficult to clearly assess and classify pain, although recent findings suggest the presence of neuropathic pain in rheumatic conditions [4].

Neuropathic pain arises from complex processes. The primary mediators derived from various cells sensitise sensory nerve endings and enhance the generation of secondary mediators from primary afferent neurons [5]. These secondary mediators, including colony-stimulating factor 1 (CSF-1), chemokine C-C motif ligand (CCL-21) and Wnt ligands, alter the properties of cells in the spinal dorsal horn (SDH) [5]. Microglia and astrocytes in the SDH are involved in the pathophysiology of neuropathic pain [6–9]. Generally, microglial activation precedes then subsides prior to astrogliosis, which involves the proliferation and activation of astrocytes [10, 11]. Astrogliosis is maintained long after the reactions of microglia subside and is closely associated with neuropathic pain [11].

A recent clinical study indicated that baricitinib, an orally taken active, reversible selective inhibitor of JAK1 and JAK2, provides greater and more rapid pain relief than adalimumab, in addition to greater clinical improvement in patients with RA who had previously shown an inadequate response to methotrexate [12, 13]. The JAK/STAT3 pathway is required for nerve injury-induced astrocyte proliferation in the SDH of spinal nerve injury model rats [9]. Considering these findings, baricitinib may regulate pain in patients with RA independently of arthritis improvement; however, the details of how baricitinib relieves pain are not well understood.

Here, we investigate the molecular mechanisms underlying baricitinib-induced pain relief. We treated collagen antibody-induced arthritis (CAIA) model mice with baricitinib or celecoxib, a representative nonsteroidal anti-inflammatory drug that is widely used for the management of joint pain, and compared pain behaviours. We conducted RNA sequencing (RNA-seq) of the dorsal root ganglion (DRG) neurons to identify baricitinib-specific changes and the neuropathic pain mediators involved. We also undertook histological examinations of the SDH and directly observed the effects of baricitinib on cultured nerve cells.

Animals

All animal experiments were carried out in accordance with the Institutional Animal Care and Use Committee (IACUC) guidelines of The University of Tokyo. We complied with all relevant ethical regulations. Male DBA/1JJmsSlc mice aged 8 to 10 weeks were purchased from the Sankyo Labo Service Corporation (Tokyo, Japan). All animals were housed in cages with free access to food and water in a facility approved by the IACUC. The mice were separated in each group with a maximum of four mice per cage. The temperature was maintained at 18 to 22℃ with a 12-hour light/dark cycle. The mice were habituated to the experimental environment for at least one week prior to the experiments. The experiments were conducted during the light phase (from 9:00 am to 9:00 pm).

CAIA induction and drug administration

CAIA was induced by an intravenous injection of 1.5 mg anti-collagen monoclonal antibody cocktail (#53040, Chondrex, WA, USA) at day 0 and intraperitoneal injection of 50 µg lipopolysaccharide from E. coli (#53040, Chondrex) at day 3. The mice were divided into the following groups, which were compared with six mice per group: CAIA group (CAIA mice with administration of vehicle), baricitinib group (CAIA mice with administration of baricitinib), celecoxib group (CAIA mice with administration of celecoxib) and control group (wild-type mice with administration of vehicle). Baricitinib (4 mg/mL; HY-15315, Med Chem Express, NJ, USA) and celecoxib (3 mg/mL; C-2816, Tokyo Chemical Industry, Tokyo, Japan) were dissolved in distilled water, and 200 µL baricitinib (40 mg/kg/day) and celecoxib (30 mg/kg/day) were intragastrically administered once a day from day 0. The equivalent quantity of vehicle was administered to both the CAIA and control groups in the same manner. All tissue samples were harvested 2 hours following the administration of the drugs or vehicle.

Evaluation of arthritis

The severity of arthritis in each mouse’s paws was evaluated by visual scoring (arthritis score) and measurement of the joint width (paw width). The arthritis score is given as the total of all paw visual scores. The visual score for each paw was graded on a scale of 0 to 4, with increments of 0.25: 0, normal; 1, mild redness, slight swelling of the ankle or wrist; 2, moderate swelling of the ankle or wrist; 3, severe swelling, including some digits, ankle or foot; 4, maximally inflamed. The maximum possible score per mouse was 16 [14]. If the arthritis score was less than 4 at day 8, the mouse was excluded in all the experiment without the control group. The paw width is given as the sum of the thickness of all wrists and the width of all ankles, measured using a digital calliper (C110T, Kloeplin, Schlüchtern, Germany) [15]. Measurements were taken every 2 to 3 days.

Histological analysis

Prior to the histological analysis, the mice were euthanised with carbon dioxide. Perfusion fixation was immediately performed, using 10% formalin. Two different sections of the L4-5 segments of the spinal cords and the ankles from each mouse were analysed [16]. All samples were fixed in 10% formalin at room temperature overnight, decalcified in 10% EDTA (pH 7.4) at 4℃ for 5 days, embedded in paraffin and cut into 5- or 6-µm sections. Haematoxylin and eosin (H&E) and safranin O staining were performed according to standard protocols. The degree of joint arthritis was histologically evaluated based on four categories: synovial inflammation, bone erosion, proteoglycan loss and cartilage erosion. Synovial inflammation and bone erosion were evaluated by H&E staining, and proteoglycan loss and cartilage erosion were evaluated by safranin O staining. Each sample was graded from 0 (healthy) to 3 (severe), with incremental grading scores of 0.25, according to a previously described protocol [17]. The total histological score was defined as the sum of the four scores.

Grip strength

Grip strength was measured using a digital grip-strength meter for mice (GPM-101B, Melquest, Toyama, Japan). Each mouse was gently placed on a wire mesh connected to a force transducer. After allowing the mouse to rest for 3 minutes, its tail was pulled and we continuously measured the force until the mouse released the wire mesh. This was repeated five times for each mouse, with intervals of more than 30 seconds. The grip strength power reported is the average of the maximum forces recorded.

von Frey test

We evaluated tactile hypersensitivity using a dynamic plantar aesthesiometer (#37550, Ugo Basile, Gemonio, Italy). A mouse was randomly placed in each box on a wire mesh, and the withdrawal pressure was measured after 30 minutes of rest. This measurement was repeated five times for both hind limbs of each mouse, with 2-minute intervals, and the average values were reported.

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

Immediately after mice were euthanised and perfused with chilled phosphate buffered saline (PBS), bilateral ankles, L4-5 DRG and L4-5 spinal cord segments were harvested and placed in TRI Reagent (#TR118, Cosmo Bio, Tokyo, Japan). Total RNA was purified using a Direct-zol RNA Kit (#R2062, Zymo Research, Irvine, CA, USA). The purified total RNA was reverse-transcribed into cDNA using ReverTra Ace qPCR RT Master Mix (#FSQ-201, TOYOBO, Osaka, Japan), and qRT-PCR was performed with THUNDERBIRD SYBR qPCR Mix (#FSQ-201, TOYOBO), using a Thermal Cycler Dice Real-Time System III (Takara Bio, Kusatsu, Japan). Relative quantification using a standard-curve method was used to compare gene expression levels. Target gene expression levels were normalised using glyceraldehyde-3-phosphate dehydrogenase (Gapdh) as an internal control. The primers used are listed in Supplementary Table 1.

RNA sequencing

DRG samples were extracted from three mice per group on day 8. The purified total RNA was confirmed by the ratio of A260/A280 (1.8-2.0), and the RNA integrity number was > 7. Sequencing was performed on an Illumina NovaSeq 6000 System with two 150-bp paired-end reads and 26.7 million reads per sample were extracted. Quality checks and trimming were performed using fastp v0.23.4 [18]. Genome mapping was performed using STAR 2.7.10b [19]. The reference index was created using GRCm39.primaryassembly.genome.fa and gencode.vM31.primary_assembly.annotation.gtf, each downloaded from GENCODE [20]. Mapping information was quantified by RSEM v1.3.3 [21], then aggregated to count the matrix using script_rnakato [22]. Genes detected in 10 or more of the 12 samples were selected. The counts were normalised using DESeq2 [23]. Principal component analyses (PCAs) were performed using the prcomp in stats built in R 4.3.2 [24]. The variance (%) of the first and second principal components is shown in the PCA plot. Heatmaps were visualised using pheatmap: pretty heatmaps v1.0.12. In the four-group analysis, differentially expressed genes (DEGs) and their rankings were determined using TCC v1.42.0 [25] and EBSeq v2.0.0 [26]. The significant genes among CAIA + Baricitinib and others were extracted in Fig. 3G. The scores corresponding to each term in mSigDB (Gene Ontology (GO), m5.all.v2023.2.Mm.symbols.gmt; Pathway, m2.cp.v2023.2.Mm.symbols.gmt) were calculated sample-by-sample using AUCell v1.24.0. Differences among groups were determined using a one-way ANOVA, and significantly different GO terms and pathways were defined as P-values < 0.0001 and < 0.01, respectively. In the two-group analysis, DEGs were extracted using DESeq2 with the following cutoff values: adjusted P-values < 0.01 and log2Fc > 0.3. GO analyses were performed using clusterProfiler, referring to the same mSigDB as previously described. Bar plots of log10(p-value) values marked as positive or negative according to the sample were illustrated using ggplot2 [27]. Hallmark GO scores were calculated by referring to mh.all.v2023.2.Mm.symbols.gmt using AUCell, then performed t.test in stats with the significant criteria P-value < 0.05. Gene set enrichment analysis (GSEA) of the Hallmark gene set was performed using GSEA 4.2.3 [28, 29].

Immunohistochemistry

The paraffin-embedded sections were incubated with primary antibodies against glial fibrillary acidic protein (GFAP, 1:2,000, 16825-1-AP, Proteintech, Rosemont, IL, USA) or ionized calcium-binding adapter molecule 1 (Iba1, 1:250, NB100-1028, Novus Biologicals, Centennial, CO, USA) for 1 hour at room temperature. The sections were then incubated for 30 minutes at room temperature with horseradish peroxidase-conjugated secondary antibodies, and Histofine Simple Stain DAB (3,3'-diaminobenzine) (#415172, Nichirei Biosciences, Tokyo, Japan) was used for visualisation. Positive signal areas were measured using ImageScope (Leica Microsystems, Wetzlar, Germany) [30].

Cell culture

Cells of the mouse cell line Neuro-2a were purchased from the Japanese Collection of Research Bioresources (JCRB) Cell Bank. Neuro-2a cells were cultured in Minimum Essential Medium (#M5650, Sigma-Aldrich, St. Louis, MO, USA) with 10% fetal bovine serum (#F7524, Sigma-Aldrich), 1% penicillin–streptomycin (#09367-34, Nacalai Tesque, Kyoto, Japan) and 1% L-glutamine (#073-05391, Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Neuro-2a cells were seeded into 24-well plates at a density of 2.5 × 10⁴ cells per well with 500 µL of medium and incubated at 37°C for 24 hours. The Neuro-2a cells were then incubated with a mixture of recombinant human IL-6 (50 ng/mL, 206-IL, R&D Systems, Minneapolis, MN, USA) and recombinant human sIL-6R (50 ng/mL, 227-SR, R&D Systems) or recombinant human TNF-α (50 ng/mL, 300-01A, Peprotech, Rocky Hill, NJ, USA). Baricitinib (400 ng/mL, HY-15315, Med Chem Express), celecoxib (300 ng/mL; C-2816, Tokyo Chemical Industry) and BMS-345541 (50 ng/mL, S8044, Selleck Chemicals, Tx, USA) were added to the treatment groups.

Statistical analysis

A one-way ANOVA followed by Tukey’s post hoc test was used to establish statistical significance in comparisons between the four groups, and values are expressed as means ± standard deviation (SD). A two-way ANOVA followed by Tukey’s post hoc test was used for time-course comparisons of the four groups, and the values are expressed as means ± standard error of the mean (SEM). The sample size was determined based on previous studies that employed similar methods. A P-value of less than 0.05 was considered to be statistically significant. The statistical analyses were performed using PRISM version 9.3.1 (GraphPad Software, San Diego, California, USA).

Arthritis-induced neuropathic pain is suppressed by baricitinib

We initially examined pain-related behaviours in CAIA model mice treated with or without baricitinib or celecoxib. Arthritis, as determined by the gross appearance of the paw and the paw width, developed several days after being induced, reached a peak between days 7 and 10 and then gradually decreased (Fig. 1A and B). These arthritic changes were suppressed to a comparable extent by baricitinib or celecoxib treatment throughout almost the entire course of the treatment (Fig. 1A and B). Histological analysis of samples from day 8 indicated that both drugs suppressed synovial inflammation, including hyperplasia and cell infiltration, to a similar extent (Fig. 1C). Synovial inflammation and proteoglycan loss were improved by baricitinib, but the erosion of bone and cartilage was not significantly changed (Fig. 1C).

We next evaluated the pain-related behaviours in these groups. Grip strength, a widely used index for inflammatory pain, decreased until 8 days and began to increase thereafter in the CAIA group (Fig. 1D). The temporal evolution of this change exhibited a consistent concordance with the arthritis score and paw width (Fig. 1A-B and D). The decrease in grip strength caused by CAIA was significantly suppressed by treatment with baricitinib or celecoxib (Fig. 1D), exhibiting a consistent effect in line with their impact on arthritis (Fig. 1A and B). The nociceptive threshold as determined by the von Frey test showed a gradual decrease in the CAIA group (Fig. 1E). Notably, the changes in the threshold were remarkably suppressed by baricitinib treatment throughout the course of the treatment but were not suppressed by celecoxib (Fig. 1E). These data indicate that only baricitinib remarkably suppressed the neuropathic pain caused by arthritis, while inflammatory pain was suppressed by both baricitinib and celecoxib.

Inflammatory responses in ankle synovium are suppressed by baricitinib or celecoxib

To investigate the involvement of cytokines in the arthritis, we conducted an analysis of mRNA expression levels pertaining to the associated genes within the synovial tissue of ankle joints. At day 8, suppressor of cytokine signalling 3 (Socs3), which is a downstream molecule of the JAK/STAT3 pathway [31], Il6 and Il1b were significantly induced by CAIA (Fig. 2A). The mRNA expression levels of these genes were suppressed by treatment with either baricitinib or celecoxib (Fig. 2A). The enhancements of these genes induced by CAIA had already subsided by day 14 (Fig. 2B). When we further examined cytokines associated with RA pathogenesis, the expression of Tnf and interferon-alpha (Ifna1) was not significantly increased in response to CAIA at days 8 or 14 (Fig. 2A and B). Interferon-gamma (Ifng) gene expression was barely detectable.

Comprehensive analysis of mRNA in DRG

To examine the mechanism underlying the suppression of CAIA-induced neuropathic pain by baricitinib, we next performed RNA-seq analysis of DRG obtained from mice at day 8. All PCAs and heatmap analyses based on genes, GO terms and pathways revealed the proximity of the baricitinib and celecoxib groups to the control and CAIA groups, respectively (Fig. 3A-F, Supplementary Fig. 1–3). Socs3 and Il1b were detected as genes significantly altered by baricitinib treatment (Fig. 3G). In particular, terms related to neurons and macrophage inflammation were altered in the baricitinib group, with a similar trend observed in the control group (Fig. 3H and I).

We next performed comparison between respective two groups. According to a clusterProfiler analysis comparing the control and CAIA groups, inflammation- and immune-related terms were enriched in the CAIA group, while those related to physiological development or maintenance of neuronal systems were enriched in the control group (Fig. 4A). Some of these neuron-related terms, and immune-related terms, including chemotaxis, were elevated in the baricitinib and CAIA groups, respectively (Fig. 4B). Of note and similar to the comparison between the CAIA and baricitinib groups, neuron-related terms were enriched in the baricitinib group, while immune-related terms were enriched in the celecoxib group (Fig. 4C). Hallmark pathway analyses [32] indicated the upregulation of various pathways by CAIA, such as inflammatory response, Notch, Wnt, interferon-alpha and p53, while many of them were downregulated by baricitinib (Fig. 4D and E). The IL-6/JAK/STAT3 signalling pathway was one of the pathways downregulated by baricitinib, but Hallmark pathway analysis showed that these changes were not significantly different between the baricitinib and celecoxib groups (Fig. 4E and F).

We also conducted GSEA. Various pathways were changed in three sets of comparisons between two groups (Fig. 5A-C). Among them, the interferon-gamma/alpha response, the IL-6/JAK/STAT3 signalling and the TNFA-signalling via NF-κB pathways were greatly enriched in the baricitinib and celecoxib groups (Fig. 5C). The top-30 DEGs in each pathway are displayed as heatmaps (Fig. 5D). In addition to pathogenesis-related cytokines such as Il1b and Il6, activating transcription factor 3 (Atf3), a rapidly expressed marker of nerve injury, and Socs3 were decreased by baricitinib (Fig. 5D). Interestingly, Csf1, a representative secondary mediator of neuropathic pain, was also downregulated in the baricitinib group (Fig. 5D).

JAK/STAT3 pathway and microglia and astrocyte stimulators are specifically suppressed by baricitinib in DRG

Next, we then used qRT-PCR to investigate mRNA expression in DRG. At day 8, the CAIA-induced expression of Socs3 and Il6 was significantly suppressed by baricitinib but not by celecoxib (Fig. 6A). This was in agreement with the aforementioned RNA-seq data and an earlier finding that the IL-6/JAK/STAT3 pathway enhances IL-6 expression [33]. The expression of Atf3 was also enhanced by CAIA induction and efficiently suppressed only by baricitinib (Fig. 6A). By day 14, the changes in these markers had subsided (Fig. 6B).

We further investigated whether primary afferent neuron-derived neuropathic pain mediators were changed by treatment with baricitinib. According to the RNA-seq data, the CAIA-induced expression of Csf1 was downregulated more by baricitinib than it was by celecoxib (Fig. 6A and B). The expression of other activators for microglia, such as Ccl21 and Wnt5a, was not obviously changed by CAIA or baricitinib (Fig. 6A and B). Wnt3a expression was barely detectable in DRG.

Proliferation of microglia and astrocytes in the SDH is suppressed by baricitinib

Next, we histologically analysed the SDH. Iba1-positive microglia were increased by CAIA at both days 8 and 14 (Fig. 7A). Only baricitinib treatment significantly inhibited the increase in microglial proliferation induced by CAIA at days 8 and 14 (Fig. 7A). GFAP-positive astrocytes were increased by CAIA and were not suppressed by baricitinib or celecoxib treatment at day 8 (Fig. 7B). However, at day 14, the number of astrocytes was decreased by baricitinib but not by celecoxib (Fig. 7B). These histological findings suggested that neuropathic pain-related changes in the spinal cord were only suppressed by baricitinib, at least on day 14.

Direct effects of baricitinib on nerve cells

Our in vivo data strongly suggest that baricitinib suppresses neuropathic pain primarily through the targeting of CSF-1 production in the DRG. To investigate whether baricitinib directly affects nerve cells in the DRG, we examined the effects of baricitinib on nerve cells in vitro. We treated cells of the neuronal cell line Neuro-2a with rhIL-6 and rhIL-6R, or rhTNF-α. When the Neuro-2a cells were treated with rhIL-6 and rhIL-6R, Csf1 expression was upregulated, in accordance with the remarkable changes seen in Socs3 and Il6 expression (Fig. 8A). These changes in expression levels were significantly suppressed by baricitinib but not by celecoxib (Fig. 8A). When Neuro-2a cells were treated with rhTNF-α, the expression levels of Il6 and Csf1, but not Socs3, were upregulated (Supplementary Fig. 4A). However, neither baricitinib nor celecoxib suppressed these changes (Supplementary Fig. 4A). When we added BMS-345541, a potent I-kappaB kinase inhibitor, to the Neuro-2a cells stimulated with rhTNFα, Il6 and Csf1 expression levels were clearly suppressed by the addition of more than 50 ng/mL BMS-345541 (Supplementary Fig. 4B), while the increased gene expression resulting from stimulation with rhIL-6 and rhIL-6R was not affected by 50 ng/mL BMS-345541 (Fig. 8B). These in vitro data confirmed that baricitinib suppresses the expression of Il6 and Csf1 induced by the IL-6/JAK/STAT3 pathway in nerve cells.

Here, we have shown that CAIA mice exhibit both inflammatory and neuropathic pain. CAIA-induced inflammatory pain was improved by treatment with either baricitinib or celecoxib, but neuropathic pain was remarkably suppressed only by baricitinib. As well as arthritis, inflammatory responses in ankle synovium were significantly suppressed by both baricitinib and celecoxib. RNA-seq analysis of DRG revealed that baricitinib, but not celecoxib, restored gene expression alterations induced by CAIA to the control condition. Among many pathways changed by CAIA and baricitinib treatment, the JAK/STAT3 and NF-κB pathways were decreased in the baricitinib group compared with the celecoxib group. Notably, only baricitinib decreased the expression of CSF-1 in the DRG. Accordingly, baricitinib prevented increases in microglia and astrocytes caused by CAIA. JAK/STAT3 pathway activity and Csf1 expression were suppressed by baricitinib in cultured neuronal cells. These findings indicate that baricitinib ameliorates CAIA-induced neuropathic pain by suppressing the IL-6/JAK/STAT3 pathway and further expression of CSF-1 in the DRG.

Despite major advances in the treatment of RA in recent years, dealing with residual pain remains a major challenge. Fautrel et al. used a matching-adjusted indirect comparison to compare the pain-reducing ability of baricitinib and other comparators and concluded that baricitinib monotherapy demonstrated greater pain reduction compared with tocilizumab and adalimumab monotherapy [34]. In the present study, the results of the grip strength test were closely associated with the degree of arthritis, while the decrease in pain threshold as determined by the von Frey test, an indicator of allodynia, was not improved even after the arthritis had subsided (Fig. 1A-E). The superior effects of baricitinib for pain relief in patients with RA are also supported by our current data linked to pain-related behaviours and our histological findings of the SDH (Fig. 1D, E, and 7). It should be noted that we did not compare baricitinib with biological DMARDs, including antibodies for TNF or IL-6, mainly because mice antibodies are not comparable with those used in humans. Despite this limitation, the superiority of baricitinib in the amelioration of neuropathic pain in mouse arthritis models might be observed in comparison to the inhibitors of these cytokines [35], because their expression in the synovium was efficiently suppressed by celecoxib, as well as by baricitinib (Fig. 2A).

Tsuda et al. reported that the JAK/STAT3 pathway is activated in a rat spinal nerve injury model and plays essential roles in the development of astrogliosis and neuropathic pain [9]. Additionally, intrathecal administration of JAK2/3 inhibitors has been shown to improve astrocyte proliferation and alleviate pain behaviours [9]. However, the involvement of peripheral nerve cells or the JAK/STAT3 pathway in these cells was not evaluated [9]. We therefore sought to determine which tissue was the most essential for the baricitinib effect to suppress neuropathic pain. In general, most systemically administered chemical compounds can barely cross the blood–brain barrier (BBB). However, using a mouse RA model, Matsushita et al. showed that baricitinib acts on a part of brain where the BBB is weak [36]. In our present data, Socs3 expression in DRG and Neuro-2a cells was noticeably decreased only by baricitinib and not by celecoxib (Fig. 6A and 8). Although we could not directly determine the delivery of baricitinib in the synovium, DRG or SDH, our data indicate that baricitinib exerts specific effects on nerve cells in the DRG and that these effects were responsible for the difference between the baricitinib and celecoxib groups.

In the current study, both the JAK/STAT3 and NF-κB pathways in the DRG were suppressed more efficiently by baricitinib than they were by celecoxib (Fig. 5C and D). Il6 is a target gene of both the STAT3 and NF-κB pathways, and IL-6 induces further activation of the JAK/STAT3 pathway [33, 37]. Our RNA-seq data showed that, compared with celecoxib, baricitinib downregulated Il1b (Fig. 5D). Considering that IL-1β is both an upstream and downstream gene of NF-κB [37], the present data indicate that baricitinib suppressed the IL-6 Amp [38], the amplification of these two pathways mediated by IL-6, during arthritis.

When proteins secreted in response to inflammation sensitise sensory nerve endings, primary afferent neurons begin to increase the production of secondary mediators, including CSF-1, CCL-21, WNT-3A and WNT-5A, which contributes to the activation of microglia in the SDH and the development of neuropathic pain [5]. Among these secondary mediators, only Csf1 is upregulated in the DRG by CAIA and selectively suppressed by baricitinib (Fig. 6A and B). CSF-1 is the best characterised among the primary afferent neuron-derived secondary mediators for microglial activation [5]. CSF-1 is induced by inflammatory mediators, such as IL-1β, from satellite cells and macrophages around the DRG [5]. Yu et al. showed that Csf1 depletion in sensory neurons largely abrogated nerve injury-induced microglial activation and proliferation [39]. Csf1 is one of the transcriptional target genes of NF-κB [37, 40]. Additionally, Csf1 is induced by STAT3, while CSF-1, as well as IL-6, can activate the STAT3 pathway [41, 42]. Taken together, these data suggest that baricitinib remarkably suppressed neuropathic pain through the inhibition of IL-6 Amp and the subsequent decrease in Csf1 expression in the DRG.

We found that baricitinib suppresses prolonged spinal glial cell activity and persistent neuropathic pain in CAIA mice by inhibiting the upregulation of IL-6 Amp and CSF-1 in the DRG. These findings may contribute to improved management of residual pain in patients with RA and the further development of novel methods for the management of inflammatory and neuropathic pain.

Atf Activating transcription factor

BBB Blood–brain barrier

CAIA Collagen antibody-induced arthritis

CCL Chemokine C-C motif ligand

CSF Colony-stimulating factor

DEG Differentially expressed gene

DMARD Disease-modifying antirheumatic drug

DRA DNA Data Bank of Japan Sequence Read Archive

DRG Dorsal root ganglion

Gapdh Glyceraldehyde-3-phosphate dehydrogenase

GSEA Gene set enrichment analysis

GO Gene Ontology

H&E Haematoxylin and eosin

IACUC Institutional Animal Care and Use Committee

Ifna1 interferon-alpha

Ifng Interferon-gamma

IL Interleukin

JAK Janus kinase

JCRB Japanese Collection of Research Bioresources

NF-κB Nuclear factor kappa B

PBS Phosphate buffered saline

PCA Principal component analysis

qRT-PCR Quantitative reverse transcription-polymerase chain reaction

RA Rheumatoid arthritis

RNA-seq RNA sequencing

SD Standard deviation

SDH Spinal dorsal horn

Socs Suppressor of cytokine signalling

STAT Signal transducer and activator of transcription

SEM Standard error of the mean

TNF Tumour necrosis factor

Supplementary information

Supplementary information associated with this article can be found online.

Availability of data and materials

The data underlying this article are available in the DNA Data Bank of Japan Sequence Read Archive (DRA) at https://www.ddbj.nig.ac.jp/dra/index.html, and can be accessed with accession number PRJDB16633.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

This work was partially supported by Eli Lilly Japan K.K. NI and MT are employees of Eli Lilly Japan K.K. and a minor shareholders of Eli Lilly and Company. The authors declare no other potential competing interests.

Funding

This work was supported by Eli Lilly Japan K.K. and JSPS KAKENHI Grant Numbers 23H05484, 21K09294 and 19H05654.

Authors’ contributions

KM, YO, NI, MT, S Tanaka, and TS designed the study. KM, YO, NT, HI, RC, AT, and MA performed the experiments. KM, HO, AT, FY, DY, S Taketomi, TM, YO, S Tanaka, and TS analyzed the data. KM and TS wrote the manuscript. All authors read and approved the final version of the manuscript.

Acknowledgements

We would like to thank J Sugita, K Kaneko and R Homma for providing technical assistance.

Ibrahim SSA, Huttunen KM: Orchestrated modulation of rheumatoid arthritis via crosstalking intracellular signaling pathways. Inflammopharmacology 2021, 29(4):965-974.
Liu S, Ma H, Zhang H, Deng C, Xin P: Recent advances on signaling pathways and their inhibitors in rheumatoid arthritis. Clin Immunol 2021, 230:108793.
Fan Y, Mao R, Yang J: NF-kappaB and STAT3 signaling pathways collaboratively link inflammation to cancer. Protein Cell 2013, 4(3):176-185.
Berghea F, Berghea CE, Zaharia D, Trandafir AI, Nita EC, Vlad VM: Residual Pain in the Context of Selecting and Switching Biologic Therapy in Inflammatory Rheumatic Diseases. Front Med (Lausanne) 2021, 8:712645.
Boakye PA, Tang SJ, Smith PA: Mediators of Neuropathic Pain; Focus on Spinal Microglia, CSF-1, BDNF, CCL21, TNF-alpha, Wnt Ligands, and Interleukin 1beta. Front Pain Res (Lausanne) 2021, 2:698157.
Chen G, Zhang YQ, Qadri YJ, Serhan CN, Ji RR: Microglia in Pain: Detrimental and Protective Roles in Pathogenesis and Resolution of Pain. Neuron 2018, 100(6):1292-1311.
Inoue K, Tsuda M: Microglia in neuropathic pain: cellular and molecular mechanisms and therapeutic potential. Nat Rev Neurosci 2018, 19(3):138-152.
Ji RR, Chamessian A, Zhang YQ: Pain regulation by non-neuronal cells and inflammation. Science 2016, 354(6312):572-577.
Tsuda M, Kohro Y, Yano T, Tsujikawa T, Kitano J, Tozaki-Saitoh H, Koyanagi S, Ohdo S, Ji RR, Salter MW et al: JAK-STAT3 pathway regulates spinal astrocyte proliferation and neuropathic pain maintenance in rats. Brain 2011, 134(Pt 4):1127-1139.
Donnelly CR, Andriessen AS, Chen G, Wang K, Jiang C, Maixner W, Ji RR: Central Nervous System Targets: Glial Cell Mechanisms in Chronic Pain. Neurotherapeutics 2020, 17(3):846-860.
Ji RR, Donnelly CR, Nedergaard M: Astrocytes in chronic pain and itch. Nat Rev Neurosci 2019, 20(11):667-685.
Taylor PC, Keystone EC, van der Heijde D, Weinblatt ME, Del Carmen Morales L, Reyes Gonzaga J, Yakushin S, Ishii T, Emoto K, Beattie S et al: Baricitinib versus Placebo or Adalimumab in Rheumatoid Arthritis. N Engl J Med 2017, 376(7):652-662.
Taylor PC, Lee YC, Fleischmann R, Takeuchi T, Perkins EL, Fautrel B, Zhu B, Quebe AK, Gaich CL, Zhang X et al: Achieving Pain Control in Rheumatoid Arthritis with Baricitinib or Adalimumab Plus Methotrexate: Results from the RA-BEAM Trial. J Clin Med 2019, 8(6).
Khachigian LM: Collagen antibody-induced arthritis. Nat Protoc 2006, 1(5):2512-2516.
Pietrosimone KM, Jin M, Poston B, Liu P: Collagen-Induced Arthritis: A model for Murine Autoimmune Arthritis. Bio Protoc 2015, 5(20).
Harrison M, O'Brien A, Adams L, Cowin G, Ruitenberg MJ, Sengul G, Watson C: Vertebral landmarks for the identification of spinal cord segments in the mouse. Neuroimage 2013, 68:22-29.
Hayer S, Vervoordeldonk MJ, Denis MC, Armaka M, Hoffmann M, Backlund J, Nandakumar KS, Niederreiter B, Geka C, Fischer A et al: 'SMASH' recommendations for standardised microscopic arthritis scoring of histological sections from inflammatory arthritis animal models. Ann Rheum Dis 2021, 80(6):714-726.
Chen S, Zhou Y, Chen Y, Gu J: fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 2018, 34(17):i884-i890.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR: STAR: ultrafast universal RNA-seq aligner. Bioinformatics 2013, 29(1):15-21.
Frankish A, Diekhans M, Jungreis I, Lagarde J, Loveland JE, Mudge JM, Sisu C, Wright JC, Armstrong J, Barnes I et al: Gencode 2021. Nucleic Acids Res 2021, 49(D1):D916-D923.
Li B, Dewey CN: RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 2011, 12:323.
Scripts for NGS analysis [https://github.com/rnakato/script_rnakato]
Love MI, Huber W, Anders S: Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014, 15(12):550.
R Core Team: R: A Language and Environment for Statistical Computing,. In., 4.3.0, edn. Vienna, Austria,; 2023.
Sun J, Nishiyama T, Shimizu K, Kadota K: TCC: an R package for comparing tag count data with robust normalization strategies. BMC Bioinformatics 2013, 14:219.
Leng N, Kendziorski C: EBSeq: An R package for gene and isoform differential expression analysis of RNA-seq data. In.; 2022.
Wickham H: ggplot2: Elegant Graphics for Data Analysis: Springer-Verlag New York; 2009.
Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J, Puigserver P, Carlsson E, Ridderstrale M, Laurila E et al: PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 2003, 34(3):267-273.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES et al: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005, 102(43):15545-15550.
Sumizono M, Sakakima H, Otsuka S, Terashi T, Nakanishi K, Ueda K, Takada S, Kikuchi K: The effect of exercise frequency on neuropathic pain and pain-related cellular reactions in the spinal cord and midbrain in a rat sciatic nerve injury model. J Pain Res 2018, 11:281-291.
Tamiya T, Kashiwagi I, Takahashi R, Yasukawa H, Yoshimura A: Suppressors of cytokine signaling (SOCS) proteins and JAK/STAT pathways: regulation of T-cell inflammation by SOCS1 and SOCS3. Arterioscler Thromb Vasc Biol 2011, 31(5):980-985.
Liberzon A, Birger C, Thorvaldsdottir H, Ghandi M, Mesirov JP, Tamayo P: The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst 2015, 1(6):417-425.
Chang Q, Bournazou E, Sansone P, Berishaj M, Gao SP, Daly L, Wels J, Theilen T, Granitto S, Zhang X et al: The IL-6/JAK/Stat3 feed-forward loop drives tumorigenesis and metastasis. Neoplasia 2013, 15(7):848-862.
Fautrel B, Zhu B, Taylor PC, van de Laar M, Emery P, De Leonardis F, Kannowski CL, Nicolay C, Kadziola Z, De La Torre I et al: Comparative effectiveness of improvement in pain and physical function for baricitinib versus adalimumab, tocilizumab and tofacitinib monotherapies in rheumatoid arthritis patients who are naive to treatment with biologic or conventional synthetic disease-modifying antirheumatic drugs: a matching-adjusted indirect comparison. RMD Open 2020, 6(1).
Peterson JD, Labranche TP, Vasquez KO, Kossodo S, Melton M, Rader R, Listello JT, Abrams MA, Misko TP: Optical tomographic imaging discriminates between disease-modifying anti-rheumatic drug (DMARD) and non-DMARD efficacy in collagen antibody-induced arthritis. Arthritis Res Ther 2010, 12(3):R105.
Matsushita T, Otani K, Yoshiga M, Hirano M, Noda K, Kurosaka D: Inhibitory effect of baricitinib on microglia and STAT3 in a region with a weak blood-brain barrier in a mouse model of rheumatoid arthritis. Rheumatology (Oxford) 2023, 62(8):2908-2917.
Brasier AR: The nuclear factor-kappaB-interleukin-6 signalling pathway mediating vascular inflammation. Cardiovasc Res 2010, 86(2):211-218.
Hirano T: IL-6 in inflammation, autoimmunity and cancer. Int Immunol 2021, 33(3):127-148.
Yu X, Liu H, Hamel KA, Morvan MG, Yu S, Leff J, Guan Z, Braz JM, Basbaum AI: Dorsal root ganglion macrophages contribute to both the initiation and persistence of neuropathic pain. Nat Commun 2020, 11(1):264.
Hohensinner PJ, Kaun C, Rychli K, Niessner A, Pfaffenberger S, Rega G, de Martin R, Maurer G, Ullrich R, Huber K et al: Macrophage colony stimulating factor expression in human cardiac cells is upregulated by tumor necrosis factor-alpha via an NF-kappaB dependent mechanism. J Thromb Haemost 2007, 5(12):2520-2528.
Grillo M, Palmer C, Holmes N, Sang F, Larner AC, Bhosale R, Shaw PE: Stat3 oxidation-dependent regulation of gene expression impacts on developmental processes and involves cooperation with Hif-1alpha. PLoS One 2020, 15(12):e0244255.
Novak U, Harpur AG, Paradiso L, Kanagasundaram V, Jaworowski A, Wilks AF, Hamilton JA: Colony-stimulating factor 1-induced STAT1 and STAT3 activation is accompanied by phosphorylation of Tyk2 in macrophages and Tyk2 and JAK1 in fibroblasts. Blood 1995, 86(8):2948-2956.

Competing interest reported. This work was partially supported by Eli Lilly Japan K.K. NI and MT are employees of Eli Lilly Japan K.K. and a minor shareholders of Eli Lilly and Company. The authors declare no other potential competing interests.

SupFig1MakabeART.pptx
Additional file 1: Supplementary Figure 1. Heatmap of the top-200 significantly altered genes among the four groups. The heatmap without gene names is shown in Fig. 3B.
SupFig2MakabeART.pptx
Additional file 2: Supplementary Figure 2. Heatmap of the top-106 significantly altered GO biological processes terms among the four groups. The heatmap without term names is shown in Fig. 3D.
SupFig3MakabeART.pptx
Additional file 3: Supplementary Figure 3. Heatmap of the top-140 significantly altered pathways among the four groups. The heatmap without pathway names is shown in Fig. 3F.
SupFig4MakabeART.pptx
Additional file 4: Supplementary Figure 4. mRNA levels of Socs3, Il6 and Csf1 in Neuro-2a cells. Neuro-2a cells were cultured with or without 50 ng/mL rhTNF (A, B), 400 ng/mL baricitinib (A), 300 ng/mL celecoxib (A) or 50 ng/mL BMS (B) for 8 hours. All data are expressed as dot plots and means ± SD. P-values were determined by one-way ANOVA followed by Tukey’s post hoc test.
SupTable1MakabeART.pptx

Download PDF

Journal Publication

published 15 Jun, 2024

Read the published version in Arthritis Research & Therapy →

Editorial decision: Revision requested
10 Apr, 2024
Reviews received at journal
05 Apr, 2024
Reviewers agreed at journal
25 Mar, 2024
Reviewers invited by journal
25 Mar, 2024
Editor assigned by journal
19 Mar, 2024
Submission checks completed at journal
19 Mar, 2024
First submitted to journal
17 Mar, 2024

You are reading this latest preprint version

Baricitinib ameliorates inflammatory and neuropathic pain in collagen antibody-induced arthritis mice by modulating the IL-6/JAK/STAT3 pathway and CSF-1 expression in dorsal root ganglion neurons

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Methods

Animals

CAIA induction and drug administration

Evaluation of arthritis

Histological analysis

Grip strength

von Frey test

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR)

RNA sequencing

Immunohistochemistry

Cell culture

Statistical analysis

Results

Arthritis-induced neuropathic pain is suppressed by baricitinib

Inflammatory responses in ankle synovium are suppressed by baricitinib or celecoxib

Comprehensive analysis of mRNA in DRG

JAK/STAT3 pathway and microglia and astrocyte stimulators are specifically suppressed by baricitinib in DRG

Proliferation of microglia and astrocytes in the SDH is suppressed by baricitinib

Direct effects of baricitinib on nerve cells

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1