Leptin and Neutrophil-Activating Peptide 2 Promote Mesenchymal Stem Cell Senescence Through Activation of the Phosphatidylinositol 3-Kinase/Akt Pathway in Patients With Systemic Lupus Erythematosus

Abstract

Objective.

Mesenchymal stem cells (MSCs) derived from patients with systemic lupus erythematosus (SLE) exhibit enhanced senescence. Cellular senescence has been reported to be induced by several inflammatory cytokines, including interferon-α (IFNα) and IFNγ, that are involved in the pathogenesis of SLE. We undertook this study to investigate whether the inflammatory environment in SLE could affect MSC senescence.

Methods.

Cellular senescence was measured by staining of senescence-associated β-galactosidase and by expression of the cell cycle inhibitors p53 and p21. Eighty cytokines and chemokines in serum from healthy controls and patients with SLE were identified by cytokine antibody array.

Results.

SLE serum promoted senescence of MSCs, which was reversed by the phosphatidylinositol 3-kinase (PI3K)/Akt signaling inhibitor LY294002 but not by the JAK/STAT inhibitor AG490 and not by the MEK/ERK inhibitor PD98059. Cytokine antibody array analysis revealed that leptin and neutrophil-activating peptide 2 (NAP-2) were the 2 factors most significantly elevated in SLE serum compared with normal serum. Blockade of leptin or NAP-2 in MSC cultures abolished SLE serum–induced senescence, while direct addition of these 2 factors could promote senescence in cultures of normal MSCs. Inhibition of PI3K/Akt signaling with LY294002 reduced leptin- and NAP-2–induced senescence in MSCs.

Conclusion.

Taken together, our data show that leptin and NAP-2 act synergistically to promote MSC senescence through enhancement of the PI3K/Akt signaling pathway in SLE patients.

Cellular senescence is a critical cellular response to continuous replication and environmental stress. It is characterized by cell cycle arrest, altered gene expression of growth regulatory proteins (such as p53, p21, and p16), morphologic transformations, and senescence-associated β-galactosidase (SA β-gal) activity (1). In this process, cells undergo an irreversible loss of growth and proliferation (2). Multiple types of stress can induce cells to undergo senescence, including oxidative stress, oncogene activity, lack of nutrients or growth factors, improper cell contacts, persistent cell replication, and stimulation with inflammatory cytokines, such as interferon-α (IFNα), IFNγ, and transforming growth factor β (3–8). It has also been demonstrated that interleukin-8, growth-related oncogene α/β/γ, neutrophil-activating peptide 2 (NAP-2), and CXCR2 can significantly promote cellular senescence (9). In addition, many signaling pathways have been reported to be involved in cellular senescence, including the JAK/STAT, Ras/Raf/MEK/ERK, NF-κB, and phosphatidylinositol 3-kinase (PI3K)/Akt signaling pathways (10–13).

Mesenchymal stem cells (MSCs) are fibroblast-like adherent cells with multilineage differentiation potential as well as potent immune modulation function (14). It has been shown that MSCs from patients with systemic lupus erythematosus (SLE) exhibit deficient function and fail to effectively inhibit T and B lymphocyte responses. Correspondingly, allogeneic MSC transplantation rather than autologous lupus MSC transplantation could ameliorate the disease and reduce proteinuria (15,16). We have previously demonstrated increases in apoptosis and senescence in bone marrow–derived MSCs (BM-MSCs) from SLE patients (13), yet the underlying mechanisms remain unclear.

Both intrinsic (genetic) and extrinsic (environmental) factors may affect cell state and function in SLE. Given the fact that MSCs from young lupus mice (ages 5–6 weeks) but not those from old lupus mice (ages 26–27 weeks) were comparable to normal MSCs from C57BL/6J (B6) mice in their effectiveness in ameliorating SLE-like disease, it is reasonable to hypothesize that MSC dysfunction in SLE might be mainly attributable to the abnormal microenvironment (17). To understand the role of the inflammatory microenvironment in SLE in MSC senescence, we cultured normal MSCs with serum from SLE patients and showed that SLE serum promoted the senescence of umbilical cord–derived MSCs (UC-MSCs) through activation of the PI3K/Akt signaling pathway. We further determined that leptin and NAP-2, the top 2 up-regulated factors in SLE serum, played a major role in inducing MSC senescence by activating the PI3K/Akt pathway.

PATIENTS AND METHODS

Study subjects.

All SLE patients fulfilled the 1997 update of the revised criteria of the American College of Rheumatology (18) and had SLE Disease Activity Index (SLEDAI) (19) scores of >5 (see Supplementary Table 1, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract). Sera were collected from 9 healthy female donors (mean ± SD age 27.3 ± 2.3 years) and 8 female SLE patients (mean ± SD age 29.8 ± 7.9 years). Sera were heated at 56°C for 30 minutes to deactivate complement and stored at −80°C until assayed. BM-MSCs were isolated from iliac crest bone marrow obtained from age-matched female SLE patients (n = 10) and normal controls (n = 9) (13). This study was approved by the Ethics Committee at The Affiliated Drum Tower Hospital of Nanjing University Medical School. All study subjects provided written informed consent.

Fresh umbilical cords were obtained from healthy mothers after normal deliveries in The Affiliated Drum Tower Hospital and were processed immediately. The cords were rinsed twice in phosphate buffered saline (PBS) with 1% penicillin and streptomycin, with the cord blood removed during this process. The washed cords were cut into 1-mm² pieces, floated in the medium, and then incubated at 37°C in a humidified atmosphere containing 5% CO₂. The medium was replaced every 3 days after the initial plating. When well-developed colonies of fibroblast-like cells appeared, the cultures were trypsinized and passaged into a new flask for further expansion.

Culture of MSCs.

MSCs were cultured in Dulbecco’s modified Eagle’s medium (DMEM)–Ham’s F-12 (Gibco) containing 10% fetal bovine serum, 10% normal serum, or 10% SLE serum. All MSCs used in this study were at passages 3–4. Flow cytometric analysis was performed to identify cell phenotype and showed CD29, CD73, CD90, and CD105 expression of >95% in parallel with CD45, CD34, CD14, CD79, and HLA–DR expression of <2%.

To investigate signaling pathways involved in MSC senescence in response to stimulation with SLE serum, the specific inhibitors PD98059, AG490, or LY294002 (Merck Millipore) targeting the MEK/ERK, JAK/STAT, or PI3K/Akt signaling pathways, respectively, were incubated at 20 μM with MSCs in DMEM–Ham’s F-12 for 20 minutes before the addition of SLE serum. In some experiments, cytokine-neutralizing antibodies including anti-IFNγ, anti-IFNα, antileptin, or anti–NAP-2 (all from R&D Systems) were added to the SLE serum–MSC cocultures at 10 μg/ml. To verify the role of leptin or NAP-2, UC-MSCs were treated with leptin (0, 10, 20, 50, or 100 ng/ml; R&D Systems) and/or NAP-2 (0, 10, 20, 50, or 100 ng/ml; R&D Systems) with or without LY294002 (20 μM).

SA β-gal assay.

SA β-gal in MSC cultures after 72 hours was assayed at pH 6.0 according to the instructions of the manufacturer (KAA002; Chemicon). Briefly, cells were washed twice with PBS, fixed in the wells using 1 ml fixing solution per well, and incubated at room temperature for 15 minutes. Following washes with PBS, cells were incubated in freshly prepared SA β-gal staining solution at 37°C without CO₂ and protected from light for ~18 hours. Senescent cells were determined by counting the number of blue-stained cells under light microscopy.

RNA isolation and real-time polymerase chain reaction (PCR).

Cultured cells were harvested after 48 hours. Total cellular RNA was extracted using TRIzol reagent (Invitrogen), and complementary DNA was synthesized with a PrimeScript RT reagent kit (Takara). Primers for p53, p21, and the housekeeping gene GAPDH were designed using Primer Express 2.0 software (Applied Biosystems) and synthesized by Takara Biotechnology. Primer sequences were as follows: for p53, 5′-TCAGCATCTTATCCGAGTGGAA-3′ (forward) and 5′-TGTAGTGGATGGTGGTACAGTCA-3′ (reverse); for p21, 5′-GAAGACCATGTGGACCTGTCACT-3′ (forward) and 5′-GAAGATCAGCCGGCGTTTG-3′ (reverse); for GAPDH, 5′-AGGGGCCATCCACAGTCTTC-3′ (forward) and 5′-AGAAGGCTGGGGCTCATTTG-3′ (reverse).

Real-time PCR was carried out in triplicate for each sample in 96-well plates using SYBR Green I dye in an Applied Biosystems 7500 real-time PCR system. Reactions were performed in a 20-μl reaction volume, and cycling was set as follows: initial denaturation for 10 seconds at 95°C, followed by 45 cycles of denaturation at 95°C for 5 seconds, and combined primer annealing/extension at 60°C for 34 seconds. Data were analyzed using SDS software, version 2.0 (Applied Biosystems). The ΔC_t value was determined by subtracting the GAPDH C_t value from the target gene C_t value. Relative gene expression was calculated as 2^{−(ΔC_teach−ΔC_tmean)}, where ΔC_t each is the ΔC_t value of each sample and ΔC_t mean is the mean of the ΔC_t values of normal controls.

Western blot analysis.

UC-MSCs were harvested and centrifuged at 500g for 10 minutes at 4°C after 72 hours of culture. The pellets were resuspended in lysis buffer (radioimmunoprecipitation assay buffer + phenylmethylsulfonyl fluoride + propidium iodide) for 30 minutes and then centrifuged at 14,000g for 10 minutes at 4°C. Supernatants were collected and quantified by the Bradford protein assay. To analyze levels of specific proteins, supernatants with equal amounts of proteins were mixed with 5× sodium dodecyl sulfate (SDS) sample buffer, boiled for 5 minutes and then separated by 8–15% SDS–polyacrylamide gel electrophoresis. After electrophoresis, proteins were transferred to PVDF membranes (Millipore) by a wet blotting system. After blocking with 5% skim milk, membranes were incubated with primary antibodies overnight at 4°C, including antibodies against p21 (Santa Cruz Biotechnology) and against p53, phospho–STAT-1, STAT-1, phospho–STAT-3, STAT-3, phospho–STAT-5, STAT-5, phospho-Akt, Akt, phospho-MEK, MEK, phospho–ERK-1/2, ERK-1/2, phospho–p38 MAPK, p38 MAPK, p65, Ras, and GAPDH (all from Cell Signaling Technology). The membranes were then washed in Tris buffered saline–Tween 20 and incubated for 2 hours with horseradish peroxidase–conjugated secondary antibody (1:5,000–1:10,000 dilution), and enhanced chemiluminescence (Millipore) was used to detect specific proteins. Anti-GAPDH antibody was used as loading control.

Cytokine and chemokine array.

Serum samples from 3 healthy individuals (mean ± SD age 25.7 ± 0.58 years) and 5 patients with active SLE (mean ± SD age 27.4 ± 5.7 years; each with a SLEDAI score of >5) were applied to an array-based cytokine screen. A RayBio human cytokine antibody array (G-Series 5, catalog no. AAH-CYT-G5–4; RayBiotech) consisting of 8 subarrays in 1 slide was used to interrogate 80 different human cytokines and chemokines (see Supplementary Table 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract). Briefly, serum samples were diluted 5-fold with sample buffer (1% bovine serum albumin in PBS), hybridized to the array, incubated with a cocktail of biotinylated secondary antibodies, and conjugated with Cy3-labeled streptavidin. The slides were scanned using a GenePix 4000B scanner (Molecular Devices) with signals acquired and transformed to digits using GenePix software. The GenePix photomultiplier setting was 33%, and the gain setting was 550 for all scans in this study. Positive control spots (POS1, POS2, POS3) consisting of standardized amounts of biotinylated IgG were printed directly onto the array and were the same in each subarray, allowing for normalization of results from different samples.

Enzyme-linked immunosorbent assay (ELISA).

Serum levels of leptin were determined with a sandwich solid-phase ELISA kit according to the manufacturer’s (Merck Millipore). Serum levels of NAP-2 were also determined with an ELISA kit (RayBiotech). Optical densities were measured in a Universal Microplate Spectrophotometer (Bio-Rad) at a wavelength of 450 nm and then transformed to values according to their standard curves.

Flow cytometry analysis.

Monoclonal R-Fluorescein–conjugated anti-human leptin antibody (R&D Systems) and Alexa Fluor 647–conjugated anti-human CXCR2 antibody (BioLegend) were applied to detect the leptin and NAP-2 receptors on MSCs by flow cytometry. MSCs were incubated with fluorescent antibodies for 30 minutes at 4°C. Subsequently, these cells were washed with PBS and fixed for flow cytometry (FACSCalibur; BD Biosciences). Data were analyzed with FlowJo software (Tree Star).

Statistical analysis.

All experiments were performed at least in triplicate, except that the array was analyzed in duplicate. Data were analyzed with SPSS software, version 16.0. All values are presented as the mean ± SD. Comparisons between 2 groups were assessed by 2-sample t-tests or Mann-Whitney U tests (in the case of non-normal distribution). P values less than 0.05 were considered significant.

RESULTS

Increased MSC senescence in response to stimulation with SLE serum.

To determine the senescence status of BM-MSCs from SLE patients, we measured SA β-gal staining and expression levels of cell cycle inhibitors (p53 and p21) in BM-MSCs. The frequency of SA β-gal–positive BM-MSCs was significantly increased in SLE patients (35.4 ± 1.0%) as compared with that in healthy controls (14.1 ± 0.9%) (P < 0.001) (n = 4) (Figures 1A and B). The expression of the cell cycle inhibitory molecules p53 and p21 was also higher in MSCs from SLE patients than in those from healthy controls at both the messenger RNA (mRNA) and protein levels (Figures 1C and D), which suggested that senescence was increased in SLE MSCs.

Figure 1.

Increased senescence in mesenchymal stem cells (MSCs) from patients with systemic lupus erythematosus (SLE) and in SLE serum–treated MSCs from normal controls (Nor). A, At passage 4, MSCs from SLE patients were remarkably positive for senescence-associated β-galactosidase (SA β-gal) staining (arrows) compared with MSCs from normal controls. B, The percentage of SA β-gal–positive MSCs was significantly increased in SLE patients (n = 4) compared with that in normal controls (n = 4). C, Western blot analysis showed that protein levels of p53 and p21 were elevated in MSCs from SLE patients (n = 3). D, Levels of mRNA for p53 and p21 were examined by semiquantitative real-time polymerase chain reaction and were found to be higher in MSCs from SLE patients (n = 10) than in MSCs from normal controls (n = 9). E, Shown is SA β-gal staining (arrows) of umbilical cord–derived MSCs (UC-MSCs) treated with normal serum or SLE serum. F, The percentage of SA β-gal–positive cells among cells treated with SLE serum was significantly higher than that among MSCs treated with normal serum (n = 5). G, Western blot analysis showed that treatment with SLE serum significantly elevated protein levels of p53 and p21 (n = 4). H, UC-MSCs treated with SLE serum (n = 8) also showed higher gene expression of p53 and p21 than those treated with normal serum (n = 8). Values are the mean ± SD. * = P < 0.05; ** = P < 0.01; *** = P < 0.001. Original magnification × 100.

Since UC-MSCs have been shown to exhibit a phenotype and function compatible with those of normal BM-MSCs (20), we next cultured UC-MSCs with SLE serum to investigate the role of the SLE microenvironment in MSC senescence. As shown in Figures 1E and F, the frequency of SA β-gal–positive cells among cells treated with SLE serum (27.2 ± 0.9%) was significantly higher than that among MSCs treated with normal serum (13.5 ± 0.6%) (P < 0.001) (n = 5). The expression of p53 and p21 was also significantly increased in MSCs treated with SLE serum compared with those treated with normal serum (P < 0.05) (Figures 1G and H). These results indicate that SLE serum enhances the senescence of MSCs in vitro and may play an important role in increased MSC senescence in SLE patients.

Involvement of the PI3K/Akt signaling pathway in MSC senescence induced by SLE serum.

Several signaling pathways have been reported to participate in cellular senescence, including the JAK/STAT, Ras/Raf/MEK/ERK, NF-κB, and PI3K/Akt pathways. Our Western blot analysis showed that phospho–STAT-1, phospho–STAT-5, phospho-Akt, phospho-MEK, and phospho–ERK-1/2 were increased in UC-MSCs after stimulation with SLE serum, whereas levels of phospho–STAT-3, Ras, and NF-κB (p65) were unchanged, suggesting potential involvement of the JAK/(STAT-1 and STAT-5), MEK/ERK, and/or PI3K/Akt pathways in senescence (see Supplementary Figure 1, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract).

To determine the pathway(s) involved in this process, UC-MSCs were treated with the MEK/ERK inhibitor PD98059, the JAK/STAT inhibitor AG490, or the PI3K/Akt inhibitor LY294002 before being cultured with SLE serum. As shown in Figures 2A and B, compared with the frequency of SA β-gal–positive MSCs that were not treated with a signaling pathway inhibitor (28.6 ± 1.4%), the frequency of SA β-gal–positive MSCs was significantly decreased by LY294002 treatment (12.9 ± 0.8%) but not by PD98059 (29.2 ± 2.3%) or AG490 (26.2 ± 1.1%). As expected, the mRNA and protein levels of p53 and p21 were also remarkably decreased after LY294002 treatment (P < 0.05) (Figures 2C and D). Thus, the PI3K/Akt signaling pathway plays a critical role in MSC senescence induced by SLE serum.

Figure 2.

Blockade of the phosphatidylinositol 3-kinase (PI3K)/Akt pathway inhibits senescence in UC-MSCs treated with SLE serum. A, Shown is SA β-gal staining (arrows) of UC-MSCs cocultured with SLE serum with or without the MEK/ERK inhibitor PD98059, the JAK/STAT inhibitor AG490, or the PI3K/Akt inhibitor LY294002, compared with SA β-gal staining of UC-MSCs cocultured with normal serum. Original magnification × 100. B, SA β-gal–positive cells were quantified. There was a higher percentage of SA β-gal–positive cells among UC-MSCs cocultured with SLE serum than among UC-MSCs cocultured with normal serum. The PI3K/Akt inhibitor LY294002 significantly decreased the number of SA β-gal–positive cells induced by SLE serum, while PD98059 and AG490 did not (n = 3). C, Western blotting demonstrated increased protein levels of p53 and p21 in UC-MSCs cocultured with SLE serum, and this was reversed after treatment with LY294002 but not after treatment with PD98059 or AG490. D, Levels of mRNA for both p53 and p21 were increased in UC-MSCs cocultured with SLE serum, and this was completely reversed by LY294002 (n = 5) but not by PD98059 or AG490. Values are the mean ± SD. * = P < 0.05; ** = P < 0.01. See Figure 1 for other definitions.

Elevation of leptin and NAP-2 levels in SLE serum.

We next investigated the active factors in SLE serum that induce MSC senescence. Microarray analysis of sera from 5 SLE patients and 3 healthy controls revealed increased expression of 12 cytokines in SLE serum, with leptin and NAP-2 being the top 2 upregulated factors (Figure 3A). To verify the data from microarray analysis, serum levels of leptin and NAP-2 in an independent cohort of 14 healthy controls and 24 SLE patients were measured by ELISA, and significant elevations were also detected in sera from SLE patients (Figures 3B and C).

Figure 3.

Differential expression of cytokines and chemokines between serum from SLE patients and serum from normal controls. A, Shown are the 12 cytokines or chemokines with the greatest increases in SLE serum (all P < 0.05 versus serum from normal controls). Among these, leptin and neutrophil-activating peptide 2 (NAP-2) were the most significantly up-regulated in SLE. The y-axis unit in the graph is pixel density, which represents percent difference. B, The elevation of leptin in SLE serum was further supported by an independent analysis of samples from 14 normal controls and 24 SLE patients, using enzyme-linked immunosorbent assay (ELISA). C, Increased expression of NAP-2 in SLE serum versus normal serum was also verified by ELISA (same samples as in B). In A, values are the mean ± SD. In B and C, symbols represent individual samples; bars show the mean ± SD. * = P < 0.05; ** = P < 0.01. HGF = hepatocyte growth factor; IL-1β = interleukin-1β; TIMP-1 = tissue inhibitor of metalloproteinases 1; FGF-9 = fibroblast growth factor 9; LIF = leukocyte inhibitory factor; LIGHT = tumor necrosis factor (ligand) super family, member 14 (see Figure 1 for other definitions).

MSC senescence reversed by blockade of leptin or NAP-2 in SLE serum.

We next studied whether leptin and NAP-2 are functionally involved in MSC senescence mediated by SLE serum. We added neutralizing antibodies targeting leptin, NAP-2, or IFNα (a cytokine reported to be crucial in SLE pathogenesis) to UC-MSCs treated with SLE serum. We found that treatment with antileptin or anti–NAP-2, but not with anti-IFNα, significantly reversed MSC senescence induced by SLE serum, as demonstrated by decreased frequencies of SA β-gal–positive cells (P < 0.05) (Figures 4A and B) and by reduced expression of the cell cycle inhibitors p53 and p21 (P < 0.05) (Figures 4C and D). Moreover, human sera with high levels of leptin had greater expression of p53 and p21 as well as a significantly increased frequency of SA β-gal–positive cells (see Supplementary Figure 2, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract). The antisenescence effects of antileptin and anti–NAP-2 were both dose dependent (see Supplementary Figure 3, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract).

Figure 4.

Effects of blocking of leptin, neutrophil-activating peptide 2 (NAP-2), and interferon-α (IFNα) on cellular senescence induced by SLE serum. A, Shown is SA β-gal staining (arrows) of UC-MSCs cocultured with SLE serum and either left untreated or treated with antileptin, anti–NAP-2, or anti-IFNα, compared with SA β-gal staining of UC-MSCs cocultured with normal serum. Original magnification × 100 B, SA β-gal–positive cells were quantified. Both antileptin and anti–NAP-2 partially decreased the percentage of SA β-gal–positive cells induced by SLE serum, while anti-IFNα did not change this percentage (n = 3). C, Western blotting verified that enhanced protein levels of p53 and p21 in UC-MSCs treated with SLE serum were partially reversed after antileptin or anti–NAP-2 treatment but not after anti-IFNα treatment. D, Levels of mRNA for p53 and p21 were increased in UC-MSCs treated with SLE serum, and this was partially reversed by antileptin or anti–NAP-2 (n = 6) but not by anti-IFNα. Values are the mean ± SD. * = P < 0.05; ** = P < 0.01. See Figure 1 for other definitions.

IFNγ was previously shown to induce senescence in human endothelial cells (5). Although IFNγ was not increased in SLE serum, we also studied the role of IFNγ in MSC senescence mediated by SLE serum. Blockade of IFNγ with IFNγ neutralizing antibody in SLE serum had no effect on mRNA levels of p53 and p21 (see Supplementary Figure 4, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract), suggesting that IFNγ was not the key factor affecting MSC senescence in SLE. Thus, leptin and NAP-2 play critical roles in MSC senescence in SLE.

Induction of MSC senescence by leptin and NAP-2 through PI3K/Akt signaling.

Since blockade of leptin or NAP-2 effectively reversed MSC senescence mediated by SLE serum, we next investigated the underlying molecular mechanisms mediating leptin and NAP-2 activity. We first showed that UC-MSCs expressed both leptin receptor and the NAP-2 receptor CXCR2, levels of which were comparable between cells treated with normal serum and cells treated with SLE serum (see Supplementary Figure 5, available on the Arthritis & Rheumatology web site at http://onlinelibrary.wiley.com/doi/10.1002/art.39196/abstract) (isotype control data are not shown). When cultured with UC-MSCs, both leptin and NAP-2 increased the percentage of SA β-gal–positive cells and up-regulated expression of p53 and p21 in a dose-dependent manner (Figures 5 and 6A). Strikingly, concomitant stimulation of UC-MSCs with leptin and NAP-2 further increased the frequency of SA β-gal–positive cells (38.2 ± 1.4% versus 24.3 ± 0.6% with leptin stimulation alone and 26.0 ± 2.0% with NAP-2 stimulation alone) (Figure 6A) and increased expression of p53 and p21 compared with stimulation of cells with leptin or NAP-2 alone (Figures 6B and C), which suggests that these 2 factors act synergistically to affect MSC senescence.

Figure 5.

Effects of leptin and neutrophil-activating peptide 2 (NAP-2) on senescence of umbilical cord–derived mesenchymal stem cells. A, Protein levels of p53 and p21 increased in a dose-dependent manner after treatment with leptin for 72 hours. Phospho-Akt levels were also elevated in a dose-dependent manner. B, Treatment with leptin for 48 hours enhanced gene expression of p53 and p21 in a dose-dependent manner. C, Protein levels of p53 and p21 were up-regulated in a dose-dependent manner after treatment with NAP-2 for 72 hours, as were levels of phospho-Akt. D, Administration of NAP-2 for 48 hours enhanced gene levels of p53 and p21 in a dose-dependent manner. Values are the mean ± SD. * = P < 0.05.

Figure 6.

Leptin and neutrophil-activating peptide 2 (NAP-2) act through the phosphatidylinositol 3-kinase/Akt pathway to modulate senescence of UC-MSCs. A, Shown is SA β-gal staining (arrows) of UC-MSCs treated with leptin (100 ng/ml) and/or NAP-2 (50 ng/ml) with or without LY294002 (20 μM) for 72 hours compared with SA β-gal staining of UC-MSCs in fetal bovine serum (FBS) (control). Both leptin and NAP-2 upregulated the percentage of SA β-gal–positive cells compared with FBS control, and leptin and NAP-2 together further increased the percentage of SA β-gal–positive cells. LY294002 treatment significantly decreased the percentage of SA β-gal–positive cells that had increased after treatment with leptin and/or NAP-2 (n = 3). Original magnification × 100. B, Protein levels of p53 and p21 were notably increased after leptin or NAP-2 treatment, and this increase was greater following treatment with both leptin and NAP-2. The p53 or p21 proteins were almost undetectable in cells treated with 20 μM LY294002 (n = 3). C, Leptin together with NAP-2 significantly increased gene levels of p53 and p21, and this was completely reversed by LY294002 (n = 3). Values are the mean ± SD. * = P < 0.05; ** = P < 0.01; *** = P < 0.001. See Figure 1 for other definitions.

To confirm that the PI3K/Akt signaling pathway is involved in leptin- and NAP-2–induced cellular senescence, we measured protein levels of phospho-Akt and Akt. We indeed demonstrated a substantial, dose-dependent elevation of phospho-Akt expression in UC-MSCs treated with leptin or NAP-2 (Figures 5A and C). Consistent with this, the senescence induced by leptin and/or NAP-2 was almost completely reversed after LY294002 treatment, which was demonstrated by a decreased number of SA β-gal–positive cells (from 38.2 ± 1.4% for leptin and NAP-2 to 2.17 ± 1.0% for leptin and NAP-2 plus LY294002) (Figure 6A) and by diminished levels of p53 and p21 (Figures 6B and C). Thus, we have confirmed that leptin and NAP-2 in SLE serum promote MSC senescence through activation of the PI3K/Akt signaling pathway.

DISCUSSION

MSCs in lupus patients display enhanced senescence (13), yet the underlying mechanisms remain largely unknown. In this study, we found that the inflammatory factors leptin and NAP-2 in SLE serum promote the senescence of MSCs, which signal through activation of the PI3K/Akt pathway.

Both genetic and environmental factors may participate in the modulation of cell senescence. Investigators in our group previously showed that BM-MSCs from young lupus-prone mice, but not those from old lupus-prone mice, could ameliorate disease in lupus mice as effectively as those from normal B6 mice (17), suggesting that dysfunction of MSCs in lupus could be attributed to environmental factors. This notion is further supported by the fact that the systemic environment in young mice significantly rejuvenates the functions of progenitor cells in aged mice (21,22). In the present study, serum from SLE patients and normal subjects was used to represent the complicated environmental factors, and our data demonstrate that SLE serum clearly promotes senescence of UC-MSCs.

A variety of substances in SLE serum, including cytokines and chemokines, play important roles in both the innate and adaptive immune responses (23–26). Two cytokines that are importantly involved in SLE pathogenesis, IFNγ and IFNα, have been reported to be capable of inducing the senescence of endothelial cells (5,6,27). However, our data showed that blocking either IFNγ or IFNα did not reverse the senescence of MSCs induced by SLE serum. To search for the components that are differentially expressed between SLE serum and normal serum, we performed a protein microarray analysis and showed that the top 2 increased factors in SLE serum were leptin and NAP-2 (Figure 3). Leptin deficiency can significantly ameliorate disease in lupus mice (28), indicating that this factor plays an important role in the pathogenesis of SLE. NAP-2 has not previously been associated with SLE disease. However, its ligand CXCR2 is reported to be up-regulated in several rheumatic diseases (29–31), and blocking it promoted recovery from multiple sclerosis in a mouse model (32). In our study, blockade of leptin or NAP-2 significantly reversed the senescence induced by SLE serum, while direct addition of these factors to MSC cultures induced cellular senescence in a dose-dependent manner, which suggests their importance in SLE serum in the senescence of MSCs. Moreover, it has been reported that leptin promotes proliferation and inhibits apoptosis of autoreactive T cells (33). Hyperleptinemia is characterized by increased Th17 cells (34) and decreased Treg cells (35), which together, accelerate the development of SLE. Our results provide new evidence that leptin may contribute to the pathogenesis of SLE.

Published studies have demonstrated that several signaling pathways might be involved in cellular senescence. For example, Ras/Raf/MEK/ERK is reported to be activated in oncogene-induced senescence (11). In tumor treatment, clinically used anticancer chemotherapeutics induce cancer cell senescence through the JAK/STAT signaling pathway (10). Additionally, evidence suggests that elevated NF-κB activity is correlated with aging, and IKK/NF-κB inhibitors could attenuate oxidative DNA damage and stress and delay cellular senescence (12,36). The role of the PI3K/Akt signaling pathway in cell senescence is a subject of controversy; some studies show that it supports cell growth (37,38), while others show that it triggers senescence with sustained activation of Akt (39–41). Hyperactivated Akt is reported to induce premature senescence, and Akt deficiency confers resistance to premature senescence (39). Our data showed that the level of phospho-Akt was elevated in MSCs cultured in SLE serum, which was also observed in BM-MSCs from SLE patients (13). Importantly, blockade of the PI3K/Akt signaling pathway almost completely reversed the senescence induced by SLE serum. However, inhibition of other pathways, such as the JAK/STAT and MEK/ERK pathways, did not change the senescence of MSCs. Thus, the PI3K/Akt pathway plays an important role in SLE serum–induced senescence.

In summary, we have discovered previously unrecognized roles of leptin and NAP-2 in lupus serum in the senescence of MSCs in SLE, and these factors signal through activation of the PI3K/Akt pathway. Blockade of leptin or NAP-2 may help to restore the function of BM-MSCs in SLE and thus have therapeutic effects in SLE patients.