Metformin decreases LPS-induced inflammatory response in rabbit annulus fibrosus stem/progenitor cells by blocking HMGB1 release

Isolation and identification of rabbit AFSCs

In order to study the cellular and molecular pathway of disc degeneration, stem cells were initially isolated from rabbit AF tissues (AFSCs) and the stemness of these AFSCs was identified by three stem cell markers, namely octamer-binding transcription factor-4 (Oct-4), stage-specific embryonic antigen-4 (SSEA-4) and nucleostemin (NS). Immunostaining results indicated that more than 92% of AFSCs were positively stained with all three stem cell markers (Figure 1), suggesting these AFSCs could be used for the following experiments.

The effect of metformin and LPS on cell morphology and proliferation

The AFSCs isolated from rabbit AF tissues were treated with various concentrations of metformin (0–10 mM) for 7 days. Although metformin did not change the morphology of AFSCs (Figure 2A–2D), it decreased the proliferation of rabbit AF cells at a concentration dependent manner (Figure 2E). However, the morphology of rabbit AF cells was altered from spindle-like (pink arrows in Figure 3A, 3E) to epithelial-like cells following LPS treatment (yellow arrows in Figure 3B–3D, 3F–3H). Semi-quantification indicated that the number of round-shape cells was increased following an increase in the concentration of LPS (Figure 3I). Certain adipocyte-like cells were also evident in LPS-treated cells (red arrows in Figure 3C, 3D, 3G, 3H).

Metformin inhibits LPS-induced HMGB1 release from the nuclei of AFSCs

Immunostaining indicated that more than 85% of HMGB1 was localized in the nuclei of normal rabbit AF cells (Figure 4A–4D, 4Q). However, LPS induced HMGB1 release from the nuclei of rabbit AF cells to the cytoplasm at a concentration dependent manner (Figure 4E–4Q). Less than 20% of HMGB1 expression was noted in the nuclei of rabbit AF cells following treatment with 100 ng/ml LPS (Figure 4N–4Q). Metformin inhibited the translocation of HMGB1 in rabbit AF cells as demonstrated by the high percentage of HMGB1 expression noted in the nuclei of rabbit AF cells treated either with metformin alone (Figure 5C, 5G, 5K, 5O, 5Q), or with metformin and LPS (Figure 5D, 5H, 5L, 5P, 5Q).

Metformin inhibits LPS- induced cell senescence of AFSCs

SA-β-gal staining indicated that LPS (100 ng/ml) induced senescence of more than 75% of the total population of AFSC cells (Figure 6B, 6F, 6I). Moreover, metformin could inhibit LPS-induced cell senescence (Figure 6D, 6H).

Metformin inhibits LPS-induced inflammation in AFSCs

The effects of metformin and LPS treatment on the inflammatory response of rabbit AF cells were further investigated by qRT-PCR on AF stem cells (Figure 6). LPS treatment increased inflammatory gene expression levels in AF cells (Figure 6). The expression of TNF-α was 7 fold-higher in AF stem cells treated with LPS (Figure 6A). Furthermore, the expression levels of IL-β1 (Figure 6B), IL-6 (Figure 6C) and COX-2 (Figure 6D) were also increased in LPS-treated rabbit AF cells.

The expression of the catabolic genes in AF stem cells was also studied. AF stem cells treated with LPS indicated significant upregulation of catabolic gene expression including MMP-3 (Figure 7E) and MMP-13 (Figure 7F) compared with those noted in the untreated control samples. Moreover, metformin treatment significantly decreased the expression levels of these genes in AF stem cells (Figure 7). The gene expression levels of collagen type I (Col-I) were further decreased in LPS-treated AF cells, while they were increased by metformin treatment (Figure 7G). Similarly, collagen type II (Col-II) gene expression levels were also decreased in AF cells treated with LPS (Figure 7H), while they were increased in metformin-treated cells (Figure 7H).

PGE2 production in LPS-treated rabbit AF cells was two times higher than that of the control cells (Figure 8A). Metformin decreased PGE2 production significantly in rabbit AF cells (Figure 8A).

The concentration of HMGB1 in the culture medium of rabbit AF cells was also increased by LPS treatment (Figure 8B). Metformin decreased the levels of HMGB1 in the medium of the cells treated with LPS (Figure 8B). Western blot analysis highlighted that the protein levels of TNF-α, IL-β1 and COX-2 were enhanced by LPS treatment and decreased by metformin treatment (Figure 8C).

Isolation of stem cells from AF tissues

The AF tissue samples were obtained from the lumbar spines of three New Zealand white rabbits (5 months old). The protocol for the use of the rabbits was approved by IACUC of the Tongji University, China. The AF tissues were cut into small pieces and digested with phosphate buffer solution containing 0.4% pronase (100 mg tissue/ml enzyme) at 37°C for 1 h. Subsequently, the samples were incubated with phosphate buffered 0.04% collagenase P solution (100 mg tissue/ml enzyme) overnight. The undigested tissue pieces were removed by a 70 μm filter, the supernatant was centrifuged at 500 g for 10 min and the pellets were washed three times with PBS. A single-cell suspension was obtained by re-suspending the cell pellets in culture medium consisting of F-12 medium (Lonza, Walkersville, MD) supplemented with 20% fetal bovine serum (FBS; Atlanta Biological, Lawrenceville, GA), 100 U/ml penicillin and 100 μg/ml streptomycin (Atlanta Biologicals, Lawrenceville, GA). The cells were cultured in T25 flasks at a density of 2 × 10⁵/flask. Following 10 days of culture, the stem cells that were isolated from AF tissues (AFSCs) formed colonies on the surface of the flask.

Purification and identification of AFSCs

For stem cell purification, individual cell colonies were collected by local application of trypsin under microscopic visualization and transferred to the individual wells of 6-well plates for further culture using published previously published method [37].

To identify the stem cells, the purified AFSCs at passage 2 were cultured in 12-well plates (3 × 10⁴ cells/well) for 5 days. The cells were fixed at room temperature with phosphate buffer containing 4% paraformaldehyde for 30 min. The fixed cells were directly reacted with mouse anti-stage-specific embryonic antigen-4 (SSEA-4; 1:500, Cat. #414000, Invitrogen, Carlsbas, CA) at 4°C overnight. For octamer-binding transcription factor-4 (Oct-4) and nucleostemin (NS) staining, the fixed cells were further treated with 0.1% triton X-100 for 30 min and subsequently incubated either with mouse anti-Oct-4 antibody (1:350, Cat. #MAB4401, Millipore, Temecula, CA), or goat anti-nucleostemin antibody (1:500, Cat. # GT15050, Neuromics, Edina, MN) at 4°C overnight. Following washing of the cells three times with PBS, cyanine 3 (Cy3)-conjugated goat anti-mouse immunoglobulin G (IgG) secondary antibody (1:500, Cat. # A10521, Invitrogen, Carlsbas, CA) was used for SSEA-4 and Oct-4 determination. Furthermore, Cy-3 conjugated donkey anti-goat IgG antibody (1:500, Cat. # AP180C, Millipore, Temecula, CA) was used for nucleostemin testing. The cells were also counterstained with Hoechst 33342 (1 μg/ml, Cat. #33270, Sigma, St Louis).

The effect of metformin on cell proliferation

To determine the effect of metformin on cell proliferation, the AFSCs at passage 2 (1.5 × 10⁴/well) were cultured in a 12-well plate with culture medium overnight. The following day, various concentrations of metformin (0, 1, 5, 10 mM) were added into the culture medium and the culture was performed for an additional 5 days. The morphology of the cells was observed by a microscope. Cell proliferation was determined by estimating the population doubling time, (PDT) which was defined as the total culture time divided by the number of generations described as reported in a previous study.²⁸ The number of generations was expressed as log₂Nc/No, where No was the population of the cells seeded initially, and Nc the population at confluence [37].

The effect of LPS on morphological changes and HMGB1 release of AFSCs

The effect of LPS on AFSCs was determined by morphological changes. The AFSCs at passage 2 (1.5 × 10⁴/well) were cultured in a 12-well plate with culture medium overnight. The following day, various concentrations of LPS (0, 10, 50, 100 ng/ml) were added into the culture medium and the culture was carried out for an additional 5 days. The morphology of the cells was observed by microscopy. The number of the cells with different shape was counted manually and expressed as percentage.

HMGB1 expression in AFSCs was determined by immunostaining. Briefly, the cells were cultured with various concentrations of LPS and fixed at room temperature with 4% phosphate-buffered paraformaldehyde for 30 min. The fixed cells were further treated with 0.1% triton X-100 for 30 min and washed with PBS for three times. Subsequently, the fixed cells were incubated with mouse anti-HMGB1 antibody (1:300, antibodies online.com, Cat. #ABIN1176834) at room temperature for 2 h. Following washing for three times with PBS, the cells were incubated with cy3-conjugated goat anti-mouse IgG secondary antibody (1:500, Cat. # A10521, Invitrogen, Carlsbas, CA) at room temperature for 1 h. The total number of the cells was determined by counterstaining with Hoechst 33342 (1 μg/ml, Cat. #33270, Sigma, St Louis).

The inhibitory effect of metformin on HMGB1 release and cell senescence

To determine the inhibitory effect of metformin on the LPS-induced nuclear translocation of HMGB1, AFSCs at passage 2 (1.5 × 10⁴/well) were cultured overnight in a 12-well plate with culture medium. The following day, the cells were treated with four different conditions. Group-1 contained the cells that were cultured with culture medium only (Control), group-2 contained the cells that were cultured in culture medium with 100 ng/ml LPS (LPS), group-3 included the cells that were cultured in culture medium with 1 mM of metformin (Met) and group 4 contained the cells that were cultured in culture medium with 100 ng/ml LPS and 1 mM metformin (LPS+Met). Following 5 days of culture, the medium from each group was collected for the determination of the prostaglandin E2 (PGE2) production and for the measurement of HMGB1 concentration using the commercial kits (Cat. #514010, for PGE2 determination, Cayman Chemical and Cat. #ST51011 for HMGB1 determination, Ann Arbor, MI; IBL International GmbH, Hamburg, Germany) according to the manufacturer’s protocols. The HMGB1 expression in AFSCs was also determined by immunostaining as described by the aforementioned protocol. The cell senescence in AFSCs was determined by the senescence β-galactosidase staining kit (Cell Signaling Technology, Cat. #9860, Danvers, MA) according to the manufacturer’s instructions.

Quantitative real-time RT-PCR for gene analysis

The metformin and LPS effects on AFSCs were further assessed by qRT-PCR. Following 5 days of culture with four different conditions, RNA was extracted from the cells using an RNeasy Mini Kit (Qiagen, Cat. #74104, Pudong, Shanghai, China). The gene expression was tested using the Qiangen QuantiTeact SYBR Green RT-PCR Kit (Qiagen, Cat. #204243, Pudong, Shanghai, China) in a Real Time PCR System (Step One Plus, AB Applied Biosystems). Rabbit specific primers were used for the detection of collagen type I (Col-I), collagen type II (Col-II), MMP-3, MMP-13, IL-6, IL-β1, COX-2, and TNF-α genes. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. The forward and reverse primer sequences were designed according to previous studies and are listed in Table 1 [38–40]. Following an initial denaturation step for 10 min at 95°C, PCR was performed for 80 cycles. Each cycle consisted of the following steps: denaturation for 50 sec at 95°C, followed by annealing for 50 sec at 58°C and extension for 40 sec at 72°C. The PCR reaction was terminated following a 10-min extension at 70°C. At least three independent experiments were performed to obtain the relative expression levels of each gene.

Western blot analysis

The effects of metformin and LPS on the expression of specific proteins of AFSCs were also assessed by western blot analysis. Following 5 days of culture with four different conditions, the protein was extracted from the cells using a lysis buffer (Thermo Scientific, Cat. #78501, Pittsburgh, PA). In order to ensure equal loading, the total protein concentration in each sample was measured by a BCA protein assay kit (Thermo Scientific, Cat. #23225, Pittsburgh, PA). A total of 30 μg protein was separated in 10% SDS gels by PAGE prior to their transfer to PVDF membranes (Bio-Rad, Hercules, CA). The blots were blocked using 5% Non-Fat dry milk (Bio-Rad, Hercules, CA) at room temperature for 1 h and subsequently incubated at 4°C overnight with primary antibodies against TNF-α (Cell Signaling, 1:1,000), IL-β1 (Cell Signaling, 1:1,000), COX-2 (Cell Signaling, 1:1,000) and β-actin (Abcam, 1:10,000). The following day, the blots were washed three times with PBS-T buffer and incubated with the corresponding secondary antibodies (LI-COR Biosciences, 1:15,000) for 1 h at room temperature. Following another three washes with PBS-T, the blots were subjected to the LiCoR Odyssey imager (LI-COR Biosciences, Lincoln, NE) for visualization of the protein bands, and semi-quantification was performed using the software of the LiCoR Odyssey imager.

Statistical analysis

All experiments were performed in three replicates. The data are presented as mean ± SD.

One-way analysis of variance (ANOVA), followed by Fisher’s predicted least-square difference (PLSD) for multiple comparisons, or two tailed student t-test wherever applicable, were used for statistical analysis. The differences between the two groups were considered significant when the p-value was less than 0.05.