The effects of nuclear factor‐kappa B in pancreatic stellate cells on inflammation and fibrosis of chronic pancreatitis

Abstract The activation of pancreatic stellate cells (PSCs) plays a critical role in the progression of pancreatic fibrosis. Nuclear factor‐kappa B (NF‐κB) is associated with chronic pancreatitis (CP). Previous evidence indicated that NF‐κB in acinar cells played a double‐edged role upon pancreatic injury, whereas NF‐κB in inflammatory cells promoted the progression of CP. However, the effects of NF‐κB in PSCs have not been studied. In the present study, using two CP models and RNAi strategy of p65 in cultured PSCs, we found that the macrophage infiltration and MCP‐1 expression were increased, and the NF‐κBp65 protein level was elevated. NF‐κBp65 was co‐expressed with PSCs. In vitro, TGF‐β1 induced overexpression of the TGF‐β receptor 1, phosphorylated TGF‐β1–activated kinase 1 (p‐TAK1) and NF‐κB in the PSCs. Moreover, the concentration of MCP‐1 in the supernatant of activated PSCs was elevated. The migration of BMDMs was promoted by the supernatant of activated PSCs. Further knockdown of NF‐κBp65 in PSCs resulted in a decline of BMDM migration, accompanied by a lower production of MCP‐1. These findings indicate that TGF‐β1 can induce the activation of NF‐κB pathway in PSCs by regulating p‐TAK1, and the NF‐κB pathway in PSCs may be a target of chronic inflammation and fibrosis.

to pancreatic fibrosis and ultimately dysfunction of the exocrine and endocrine glands. 3 Therefore, the activation of PSCs is considered a critical step in the development of pancreatic fibrosis. 4,5 Although TGF-beta 1 (TGF-β1) is thought to be the most likely regulator of PSC activation and proliferation, 3 its mechanism in regulating the downstream signalling pathway is unclear. Experiments showed that TGF-β1 signal transduction was linked to the transcription factor nuclear factor-kappa B (NF-κB); however, the results were contradictory, depending on the different cell types. 6,7 In one study, TGF-β inhibited the activation of NF-κB by increasing the expression of inhibitor of κB-alpha (IκB-α) in both salivary gland and breast cancer cells. 6 However, in rat hepatic stellate cells, TGF-β appeared to induce the activation of NF-κB. 7 In PSCs, the influence of TGF-β on NF-κB activation remains unclear.
The transcription factor NF-κB plays a crucial role in regulating inflammation, proliferation and apoptosis. 8 In a previous study, Steinle et al 9 reported that the NF-κBp65 DNA-binding activity was elevated in a murine model of acute pancreatitis induced by cerulein. Recently, some research studies attempted to explore the relationship between NF-κB and CP. 10,11 Matthias et al 11  in pancreatic acinar cells showed spontaneous acinar damage, obvious inflammation and fibrosis, in addition to NF-κB overexpression. Furthermore, after the administration of cerulein, the severity of pancreatitis was greater in mice overexpressing IKK2 compared with that observed in wild-type mice. However, the potential role of NF-κB in regulating the function of PSCs and inflammatory microenvironment during the progression of CP remains unclear.
In the present study, we used two CP models and a small interfering RNA (siRNA)-lipotransfection strategy to inhibit p65, a dominant subunit of NF-κB dimers, 10 in isolated and cultured PSCs. Our aim was to shed light on the effect and mechanism of NF-κB in PSCs, and identify a target for inflammation and fibrosis during the progression of CP.

| Animals
Kunming mice and C57BL/6 mice were purchased from the Experimental Animal Center of Xi'an Jiaotong University (Xi'an, China). All experimental procedures adhered strictly to the animal care and handling guidelines of the Committee on Animal Care of Shaanxi University of Chinese Medicine (Xi'an, China).

| Murine models of CP
Two murine models of CP were induced by administration of cerulein (C9026; Sigma-Aldrich) or L-arginine (A6969; Sigma-Aldrich). The C57BL/6 mice were intraperitoneally injected with cerulein (50 μg/ kg, six times per day, 3 d/wk) for 6 weeks. The Kunming mice were intraperitoneally injected with 20% l-arginine (3 g/kg, twice per day, 1 d/wk) for 6 weeks. A control group comprised mice injected with the same amount of sterilized saline. The mice were killed at weeks 2, 4 and 6 after modelling.

| Histological analysis
The pancreatic tissues were immediately immersed in 4% paraformaldehyde for 12 hours, dehydrated, embedded in paraffin and sectioned (2 μm). Haematoxylin and eosin and Masson's trichrome staining were performed. Pancreatic sections were assessed at 20× objective magnification over 10 separate fields to determine the severity of pancreatitis by scoring for oedema, inflammation, atrophy, necrosis and fibrosis.

| Immunohistochemistry
Immunohistochemistry (IHC) was performed using a ready-to-use SABC kit (Boster Biological Technology Co, Ltd). The sections were incubated with a murine monoclonal anti-p65 antibody (Santa Cruz Biotechnology, Inc), a monoclonal rat anti-F4/80 antibody (Santa Cruz) and a rabbit MCP-1 antibody (Biosynthesis Biotechnology Co., Ltd), all for 48 hours at 4°C. The corresponding secondary antibodies (antimouse secondary antibodies for p65 antibody, anti-rat secondary antibodies for F4/80 antibody, anti-rabbit secondary antibodies for MCP-1 antibody) were subsequently incubated for 1 hour at room temperature. Images were captured using a ZEISS Imager A1 Microscope (Carl Zeiss AG).
In vitro, PSCs cultured on 20 × 20-mm slides were fixed in 4% paraformaldehyde, permeabilized in 0.5% Triton X-100 (Solarbio Life Sciences Co., Ltd), blocked (normal goat serum and 0.5% Triton X-100) and incubated with a rabbit anti-p65 antibody or anti-α-SMA antibody at 4°C overnight. The cells were incubated for 1 hour with goat anti-rabbit fluorescein isothiocyanate secondary antibody, and the nuclei were counterstained with DAPI. Images were captured with a ZEISS Imager A2 Fluorescence Microscope (Carl Zeiss AG).

| Isolation of PSCs
For the isolation of PSCs, the Kunming mice (n = 6, aged 6-8 weeks) were killed using diethyl ether in accordance with the standard procedures. 12 The second passage of PSCs was used in subsequent analyses.

| Isolation of bone marrow-derived macrophages (BMDMs)
For the isolation of BMDMs, the Kunming mice (n = 6, aged 6-8 weeks) were killed and immersed in 75% ethanol. The skin was removed from the lower part of the body, and the remaining pelvic tissue and femoral bone tissue were cleaned, followed by separation at the knee joint. The bones were subsequently immersed in Dulbecco's phosphate-buffered saline for 5 minutes and placed in RPMI 1640 until the next step. Each end of the bone was then cut off, and the bone marrow from both ends of the bone was rinsed using a 1-ml syringe filled with RPMI 1640 medium. The bone marrow was collected in a 50-ml centrifuge tube, and the cells were filtered using a 70-μm sieve. The red blood cells were removed by red blood cell lysate, and the cells were washed twice with RPMI 1640 medium. Subsequently, 2 × 10 6 cells were added to 4 mL of complete medium and seeded in 35-mm culture dishes. After 16 hours of inoculation, adherent cells were discarded, and the cells in the supernatant were collected and re-seeded in a six-well plate at 1 × 10 6 , followed by incubation with medium containing 50 ng of macrophage colony-stimulating factor. Subsequently, 50% of the cell culture medium was replaced with a new 50% cell culture solution containing 50 ng of macrophage colony-stimulating factor on day 4. BMDMs obtained on day 5 of culture were identified through flow cytometry using macrophage markers (CD11b and F4/80).
The cell density was approximately 1 × 10 6 , and the purity of the macrophages was >98%.

| Scratch wound assay
The BMDMs were seeded in a six-well plate with serum-free culture medium at 1 × 10 6 . Cell proliferation was completely inhibited by the administration of 10 µg/mL of mitomycin C (Invitrogen) for 1 hour.
A straight scratch was performed using the tip of a P200 pipette.
The cells were subsequently washed thrice with phosphate-buffered saline and cultured in different culture supernatants, including a control group, TGF-β1 without PSC group, TGF-β1-stimulated PSC group, control RNA interference (RNAi) plus TGF-β1 group, and NF-κBp65 RNAi plus TGF-β1 group. After incubation for 24 hours, the gap width of the scratch repopulation was measured and compared with the initial gap size (0 hour).

| RNAi strategy
The effectiveness of Lipofectamine 3000 (Thermo Fisher Scientific) in  Table 1.

| Western blot analysis
For Western blotting, lysates from pancreatic tissue (loading protein = 40 µg) or PSCs (loading protein = 20 µg) were separated through sodium dodecyl sulphate-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane.

| Quantitative real-time PCR
Total RNA was extracted from pancreatic tissue or cultured PSCs and reverse transcribed using a reverse transcription kit (Takara Biotechnology). The real-time PCR was performed using a SYBR Green Kit (Takara Biotechnology). Amplification and detection were accomplished using an ABI 7500 fast real-time PCR system. GAPDH was used as an internal RNA loading control. The primer sequences are shown in Table 2.

| Enzyme-linked immunosorbent assay
The supernatant from the cultured PSCs was measured using enzymelinked immunosorbent assay kits (Boster Biological Technology Co., Ltd.) for quantification of the concentration of MCP-1, without dilution.

| Statistical analysis
Data were expressed as the mean ± standard deviation. Statistical analyses were performed using a t test and one-way analysis of variance (ANOVA). A P value <.05 was considered statistically significant.

| Pathological changes and overexpression of NF-κΒ in two different murine models of CP
As shown in Figure 1A,B, the mice were killed at weeks 2, 4 and

| Overexpression of NF-κB detected in activated PSCs in both CP models
Immunohistochemistry was performed to detect the localization of

| Overactivation of NF-κB in isolated PSCs stimulated with TGF-β1 in vitro
The second passage of PSCs was used in the experiment to obtain a sufficient amount of PSCs for the proteolytic samples. Experiments were performed to identify potential differences between the primary isolated PSCs and the second passage of PSCs at 5 days of culture. We found that both types of cells showed positive staining of Oil Red O and autofluorescence ( Figure 3A). The expression of α-SMA in the second passage of PSCs at 5 days of culture was only slightly higher than that observed in the primary isolated PSCs, without significant difference ( Figure 3B). These findings indicate that the activity of the second passage PSCs is similar to that of primary isolated PSCs, and the second passage of PSCs was suitable for subsequent experiments. TGF-β1 is a potent driver of pancreatic

Gene
Primer forward Primer reverse

| Successful knockdown of NF-κBp65 expression in PSCs through siRNA lipotransfection
As shown in Figure 4A the most effective inhibition observed at 36 hours (>90% inhibition; Figure 4B). As shown in Figure 4C, the level of p65 protein was markedly inhibited in the si-p65 (p65-1663) group compared with that measured in the negative control group. These results suggest that transfection of p65 siRNA can be used to knock down the expression of NF-κB in PSCs. showed that the expression of α-SMA was highly elevated after 30 minutes of stimulation with TGF-β1 and sustained at a higher level for 24 hours, suggesting that TGF-β1 may augment the activity of PSCs ( Figure 3C). The overexpression of TGF-β receptor 1 in the PSC lysate was observed from 15 minutes of incubation ( Figure 3D).

| Effect of NF-κBp65 knockdown on PSC activation and fibrosis-related factors
At the same time, the expression of p-TAK1, which is the upstream gene of the NF-κB signalling pathway, was highly elevated after stimulation with TGF-β1 from 30 minutes to 4 hours ( Figure 3E).
The expression of p65 was elevated 30 minutes after stimulation with TGF-β1, reached a peak after 1 hour and remained at a high F I G U R E 2 Localization of NF-κB p65 in the PSCs after the induction of CP at weeks 2, 4 and 6 in both murine models. A, B, IHC staining for NF-κBp65, the brown colour denoting positive cells (original magnification: 200×). C, D, Immunofluorescence double staining of α-SMA and p65 in the murine models of CP. DyLight 594-conjugated α-SMA (red), DyLight 488-conjugated p65 (green), DAPI (blue) and co-expression areas (orange).

Original magnification: 400×
After successful transfection of NF-κBp65 siRNA, the expression level of TIMP-1 was reduced compared with that detected in the TGF-β1 group ( Figure 5D).

| Effect of NF-κBp65 knockdown on the production of chemokine MCP-1 and its potential role in regulating pancreatic inflammation
The level of MCP-1 mRNA expression in PSCs stimulated with TGF-β1 was increased at 12 hours and 24 hours (P < .01 compared with the TGF-β1-untreated group); however, it was obviously reduced by

| D ISCUSS I ON
Progressive fibrosis is considered the final pathological manifestation of CP, and PSC activation is an early cellular event in the initiation of pancreatic fibrosis. 12,13 Previous studies proposed that a To investigate the role of NF-κB specifically in PSCs, we isolated PSCs from healthy mice, cultured and stimulated the second passage of PSCs with TGF-β1. The results showed that activated PSCs expressed a high level of α-SMA in response to TGF-β1 ( Figure 3C). At the same time, IκB-α was abundantly degraded, and p65 phosphorylation was increased. Thereafter, the expression of p65 protein was upregulated and reached a peak at 1 hour after incubation with TGF-β1. These results suggested that the NF-κB pathway in PSCs was activated by TGF-β1 ( Figure 3F). However, the mechanism through which TGF-β1 induced the activation of the NF-κB pathway in PSCs remains unclear.
In addition, we found that the TGF-β receptor 1 and p-TAK1 were rapidly overexpressed after stimulation with TGF-β1 ( Figure 3D,E). In experiments involving other types of cells, TGF-β1 activated TAK, with subsequent activation of the IKK complex, followed by degradation of phosphorylated IκB-α and nuclear translocation of NF-κB. 19, 20 Arsura et al 21 provided further evidence that TAK1 may participate in TGF-β-induced activation of NF-κB. In our experiment, the expression of p-TAK1 was highly elevated and induced earlier than the expression of NF-κB after stimulation with TGF-β1. These findings indicate that the process of NF-κB activation in PSCs induced by TGF-β1 may be related to TAK1 phosphorylation.
As previously reported, in the NF-κB 'classical pathway', p65/p50 is the predominant form of NF-κB heterodimers, and p65 contains a Rel homology domain responsible for DNA binding. 22 Therefore, in the present study, p65 was selected as a target to inhibit the activity of NF-κB. 23 The cells were transfected with p65 siRNA using Lipofectamine 3000 (Thermo Fisher Scientific) to ascertain the role of NF-κB in the activation of PSC induced by TGF-β1. After successful transfection of p65 siRNA, PSC activation was significantly decreased, as shown by a reduction in the level of α-SMA ( Figure 5).
This finding further indicated that NF-κB plays an important role in the regulation of PSC activation. Furthermore, shortly after transfection of p65 siRNA, the suppression of MMP-1 caused by TGF-β1 was relieved, and the elevation of TIMP-1 was inhibited ( Figure 5).
These results suggest that NF-κB activation in PSCs is responsible for the imbalance between the expression of MMP-1 and TIMP-1, both of which are related to the deposition and degradation of ECM in the progression of CP. [24][25][26] In this study, we found that infiltration of inflammatory cells increased with the progression of pancreatic fibrosis in both CP models ( Figure 6G). Many previous studies revealed that the chemokine MCP-1 contributed to inflammation by recruiting inflammatory cells to injured areas. [27][28][29] In the present study, MCP-1 was obviously increased in the pancreas of CP mice ( Figure 6C-F).  Figure 6H,I). These results further verified that the chemokine MCP-1 secreted by PSCs can recruit more macrophages to the injured pancreas and aggravate the pancreatic inflammatory progress. 30 In addition, macrophage migration induced by the supernatant of activated PSCs was dependent on the activity of the NF-κB pathway. In summary, NF-κB activation precipitated the production of MCP-1, which possibly recruited tissue-infiltrating macrophages to participate in the progression of pancreatic inflammatory damage.
In the present study, we demonstrated the mechanism of NF-κB signalling pathway in PSCs influencing pancreatic inflammation and fibrosis of CP (Figure 7). This study revealed the cascade as follows: In

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. Hong Zhang: Conceptualization (lead); Funding acquisition (lead); Project administration (lead); Supervision (lead); Writing -original draft (lead); Writing -review and editing (lead).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.