Detrimental alteration of mesenchymal stem cells by an articular inflammatory microenvironment results in deterioration of osteoarthritis

Abstract

Background: Articular injection of mesenchymal stem cells (MSCs) has been applied to treat knee osteoarthritis (kOA), but its clinical outcomes are controversial. This study investigated whether an articular inflammatory microenvironment (AIM) impacts MSC-based therapy in a rat model of kOA.

Methods: The biological change of MSCs and the functional change of MSCs on chondrocytes were evaluated under AIM. The key mediator and mechanism for the AIM impact on MSC therapy were explored via gain- and loss-of-function approaches.

Results: The results showed that MSCs exerted potent anti-kOA effects in vivo and in vitro, but that this therapy become chondrodestructive if a chronic AIM was present. Mechanistically, the overexpression of MMP13 in the injected MSCs via a MAPKs-AP1 signaling axis was revealed as the underlying mechanism for the detriment outcome.

Conclusions: This study thus clarifies recent clinical findings while also suggesting a means to overcome any detrimental effects of MSC-based therapy while improving its efficacy.

Keywords: Articular inflammatory microenvironment; Human umbilical cord mesenchymal stem cells; Knee osteoarthritis; MAPKs; MMP13; Synovial fluid; c-Jun.

Fig. 1

Histopathological observation, IHC analysis, and pain behavior tests on the in vivo effects of hucMSCs, SF, and SF-stimulated hucMSCs on kOA rats. A Safranin O/Fast green staining on cartilage with black arrows (hypertrophy or loss of chondrocytes). Scale bars = 50 μm. B IHC staining of Col2 and MMP13 on cartilage with red arrows (positive cells). Scale bars = 50 μm. C OARSI and Mankin’s scoring of histopathology. D Quantitative measurements of the percentage of Col2-positive area and MMP13-positive cells. E Measurement of MWT and TWL of rats. Data expressed as mean ± SD. The different letter symbols (a, b, bc, c, cd, and d) indicate significant difference between each other (Fisher’s LSD, P < 0.05 or P < 0.01) in the descending order of data from a to z. Although the overlapped letters (b versus bc, bc versus c, c versus cd, and cd versus d) were statistically different, their difference between each other was not significant. All experiments were repeated at least three times

Fig. 2

Evaluation of the wound healing and proliferation-regulatory effects of hucMSCs on chondrocytes before and after SF stimulation. A Wound healing of chondrocytes with hucMSCs and/or SF treatment at 0 h, 24 h, and 48 h. Scale bars = 200 μm. B Quantification of the wound area ratio (blank area/total area) at 0 h, 24 h, and 48 h. C Quantification of PCNA-expressed cell ratio (positive cells/total cells). D Cell immunofluorescence of chondrocytes with hucMSCs and/or SF treatment at 24 h, n = 3. E Expressions of anabolic and catabolic genes of chondrocytes. F Protein expression of MMP13 in chondrocytes. Values are presented as mean ± SD. The different letter symbols (a, ab, b, c, and d) indicate significant difference between each other (Fisher's LSD, P < 0.05 or P < 0.01) in descending order of data from a to z. Although the overlapped letters (a versus ab and ab versus b) were statistically different, their difference between each other was not significant. All experiments were repeated at least three times

Fig. 3

Evaluation of stemness and MMP13 expression of hucMSCs before and after SF stimulation. A Flow cytometry of the immunophenotype of mesenchymal stem cell surface markers of hucMSCs. B A WB of stemness-related proteins from hucMSCs. C qPCR and WB analyses of MMP13 expression in hucMSCs. Values are presented as mean ± SD. ^#P < 0.05; ^##P < 0.01. All experiments were repeated at least three times

Fig. 4

Evaluation of the anabolic and catabolic effects of MSC^OE and MSC^KD on chondrocytes by qRT-PCR and WB. A mRNA expression of Mmp13 and Col2 in chondrocytes treated with normal hucMSCs and MMP13-overexpressing hucMSCs. B Protein expression levels of MMP13, Adamts4, and Col2 in chondrocytes treated with normal hucMSCs or MMP13-overexpressing hucMSCs. C mRNA expression levels of Col2, Col10, Adamts4, and Adamts5 in chondrocytes treated with normal hucMSCs, normal hucMSCs in the presence of SF and MMP13-knockdown hucMSCs in the presence of SF. D Protein expression levels of Col2 and Col10 in chondrocytes treated with hucMSCs, hucMSCs in the presence of SF, and hucMSCs with MMP13 knockdown in the presence of SF. MSC^NC: nontargeting control siRNA-treated hucMSCs, MSC^KD: MMP13-knockdown siRNA-treated hucMSCs. The medium of MSC^NC group was replaced by IMDM containing 10% FBS (v/v) and 10% PBS (v/v), and the medium of MSC^NC + SF group and MSC^KD + SF groups were replaced by IMDM containing 10% FBS (v/v) and 10% SF (v/v). Values are presented as mean ± SD. ^#P < 0.05 versus MSC^NC group; ^##P < 0.01 versus MSC^NC group; ^**P < 0.01 versus MSC^NC + SF group. All experiments were repeated at least three times

Fig. 5

Histopathological and IHC analyses and pain behavior tests in kOA rats treated with normal hucMSCs, hucMSCs with MMP13 overexpression, SF-stimulated hucMSCs and SF-stimulated knockdown MMP13-hucMSCs. A Safranin O and Fast green staining of the cartilage, with black arrows indicating hypertrophy or loss of chondrocytes. Scale bars = 50 μm. B IHC staining of Col2 and MMP13 on cartilage, with red arrows indicating positive cells. Scale bars = 50 μm. C OARSI and Mankin’s scoring of histopathology. D Quantitative measurements of the percentage of Col2-positive area and MMP13-positive cells. E Measurement of MWT and TWL of rats. Data are expressed as mean ± SD. The different letter symbols (a, b, c, cd, d, and e) indicate significant difference between each other (Fisher's LSD, P < 0.05 or P < 0.01) in descending order of data from a to z. Although the overlapped letters (c versus cd, and cd versus d) were statistically different, their difference between each other was not significant. All experiments were repeated at least three times

Fig. 6

SF-mediated transcription factors upregulate MMP13. A Expression of hucMSC transcription factors treated with SF. B Expression of hucMSCs treated with SF and p65-SiRNA. C Expression of hucMSCs treated with SF and c-Fos-SiRNA. D Expression of hucMSCs treated with SF and c-Jun-SiRNA. E Expression of chondrocytes treated with hucMSCs (with c-Jun knockdown) and SF. Values are presented as mean ± SD. The different letter symbols (a, ab, b, bc, c, cd, bcd, d, and e) indicate significant difference between each other (Fisher's LSD, P < 0.05 or P < 0.01) in descending order of data from a to z, in which the overlapped letters (e.g., b versus bc, c versus cd, and cd versus d) were statistically but not significantly differed. All experiments were repeated at least three times

Fig. 7

SF upregulates MMP13 expression in hucMSCs via a MAPK-AP1 pathway. A Protein bands and quantitation of their expression in hucMSCs treated with SF. B hucMSCs treated with SF and p38 inhibitor. C hucMSCs treated with SF and Erk1/2 inhibitor. D hucMSCs treated with SF and JNK inhibitor. E hucMSCs nuclei treated with SF and p38 inhibitor. F hucMSCs nuclei treated with SF and Erk1/2 inhibitor. G hucMSCs nuclei treated with SF and JNK inhibitor. Values are presented as mean ± SD. ^#P < 0.05 and ^##P < 0.01 versus NC group; ^*P < 0.05 and ^**P < 0.01 versus SF group. All experiments were repeated at least three times

Fig. 8

Summary of the impact and underlying mechanism of an inflammatory microenvironment on hucMSCs-based therapy for kOA