Effects Of Inflammatory Microenvironment On Human Gingiva-derived Mesenchymal Stem Cells

Human gingival tissues are prone to develop hyperplasia under multipleexternal stimili. We have recently shown that normal and hyperplastic gingivaltissue contains mesenchymal stem cells (GMSCs), but it is not known how thesecells are influenced by commonly seen inflammatory conditions. We isolatedGMSCs from human normal (N-GMSC) and inflammatory gingival tissues(I-GMSC). Meanwhile, an in vitro culture with TNF-α and IL-1β wasestablished to simulate the in vivo inflammatory situation. In these two settingswe observed consistent results that GMSC affected by inflammatory factorsproliferated more actively than those under normal condition, while lostpotential of osteogenic and adipogenic differentiation. The expression of matrixmetalloproteinases (MMP)-1/2, IL-1α, IL-6, TNF-α and COL-1wassignificantly higher in I-GMSC than that in N-GMSC. The comparison to dentalpulp stem cells disclosed their intrinsic differences with regard to in vitro geneexpression and in vivo extracellular matrix formation and regulation. Thesefindings suggest that inflammatory factors turn GMSC to a more differentiated and pro-fibrotic phenotype, which could underlie the clinical hyperplasticappearance of inflammatory gingiva.Section1, Histological observation of human normal andinflammation gingival tissuesExperiment1, inflammation gingival tissues exhibited macroscopic tissueover-growth (FIG.1A,B) and thickened layers and elongated rete ridges withabundant cellularity and extracellular matrix within lamina propria by H&Estaining (FIG.1C, D).Experiment2, immunofluorescence assay showed that inflammation gingivaltissues and cells expressed more abundant collagen I (COL-I, FIG.1K-M; N-O).We analyzed the expression of STRO-1, aputative mesenchymal stem cellmarker and SSEA-4, which was deemed as an early embryonic glycolipidantigen specific for pluripotent embryonic stem cells and had recently been usedto identify and purify stem cells from adult bone marrow in human normal andhyperplastic gingival tissues. Both normal and inflammation gingival propriadisplayed positive STRO-1(FIG.1E-G) and SSEA-4(FIG.1H-J) staining,with the latter marker being more sporadically distributed. These resultsindicated that despite the existence of differences between human normal andhyperplastic gingival tissues, they both showed evidence for harboring cells withfeatures of mesenchymal stem cells.Section2, Study of separation, culture and biological characteristicsof GMSCExperiment1, clonogenic MSC can be isolated from gingival tissue. Tocharacterize whether gingival propria-derived MSC is clonogenic, we generatedand cultured single cell suspensions from human normal and inflammationgingival tissues after separating and discarding gingival epithelium byincubating with3mg/ml dispase for30min. With2×103initially seeded cells,both cultures were observed with cells attaching onto the plastic in2days andthen they proliferate rapidly in next5days to form colonies. These colonies were stained with toluidine blue and exhibited typical fibroblastic morphology(FIG.3A). With above results, we termed these clonogenic populations gingivalmesenchymal stem cells (GMSC), with N-GMSC and I-GMSC referring toGMSC derived from normal and inflammation gingival propria respectively.The ability of gingival tissue derived cells to form adherent clonogenic cellclusters had no significant difference (FIG.3B).Experiment2, by using limiting dilution technique, single cell-derivedcolony cultures were obtained. For N-GMSC,5batches of culture produced42single cell-derived colonies. For I-GMSC,3batches of culture produced34single cell-derived colonies. These colonies were randomly selected forsubsequent analysis. We used flow cytometric analysis to characterize N-GMSCand H-GMSC by surface molecules. Both GMSC within10passages showedthe characteristic pattern of mesenchymal surface markers including STRO-1,CD29, CD44, CD90, CD105, CD146and negatively expressed hematopoieticmarkers CD34and CD45(Fig.3C).Experiment3, GMSC possesses self renewal and multilineage differentiationability in vitro. The multi-differentiation potential of both GMSC wasdetermined. Under osteogenic induction conditions for3weeks, both GMSCcould specifically differentiate and formed distinct nodules as stained byAlizarin Red S. After feeding GMSC with osteogenic inducing media, dark redmineralised bone matrix (bone nodules) was shown in alizarin red stainedsections. The cells were positively stained with ALP(FIG.4A①). PCR andRT-PCR analysis showed that the expression of the osteogenic markers bonesialoprotein runt-related transcription factor2(RUNX2), osteocalcin (OCN),alkaline phosphatase (ALP) were increased after osteogenic induction(FIG.4B③④). After3weeks of culture in adipogenic induction medium, GMSCproduced lipid droplets, the hallmark of functional adipogenesis. Alkalinephosphatase activity staining (2weeks) showed I-GMSC lighter thanN-GMSC.(FIG.4A④). Adipogenic differentiation of GMSC was confirmed by Oil Red-O staining.After feeding GMSCs with adipogenic inducing media for21days, oil dropletswere presented in cytoplasm(FIG.4B①). Adipogenic differentiation was furtherconfirmed by the increased expression of specific adipogenic markers includingperoxisome proliferator-activated receptor γ (PPARγ). All are determined byRT-PCR (FIG.4A⑤B④). Through quantitative comparison of areas fromrandomly selected staining images, or grayscale analysis of RT-PCR bands, wehave found significant differences between N-and I-GMSC groups (datashown). To further compare the enzymatic activity related to ECM remodeling,we performed immunostaining and found that Collagen type1(COL-1)expression within N-GMSC was significantly lower than that in I-GMSC (FIG.1NO), suggesting role of COL-1expression within I-GMSC in mediating thepathological processes.Section3, N-GMSC and I-GMSC show some special characteristicboth in vitro and in vivoExperiment1, MTT assay and Cell Cycle showed that the growth tendencyof I-GMSC waw higher than that on the healthy group (FIG.5A,B). Accordingto MTT assay, concentration sitmulates the most N-GMSC proliferation was asthe final extrinsic source inflammatory factor IL-1β (5ng/ml) and TNF-α(10ng/ml)(FIG.5C). With extrinsic source inflammatory factor to N-GMSCshowed more apoptosis than that without inflammatory factor, and especiallywith extrinsic source inflammatory factor and Osteogenesis to N-GMSC showedthe most apoptosis (FIG.5D). N-GMSC possessed self renewal and multilineagedifferentiation ability in inflammatory microenvironment being reduced byRT-PCR analysis and alkaline phosphatase activity staining (FIG.6A,B,C).Experiment2, GMSC expresses some cytokine in vivo. The present studyexamined HGFs from healthy and inflammatory gingival tissues using DNAmicroarray analysis. The expression levels of TIMP1, TIMP2, IL-1α, IL-6,TNF-α, and COL-1in the HGFs of the Inflammatory group were higher than those in the healthy group on DNA microarray analysis, MMP1MMP2lowerthan those in the N-GMSC (figure3.2.1).Experiment3, these results showed that the DPSCs in inflammatory produceelevated levels of these MMP1/2, IL-1α, IL-6, TNF-α and COL-1proteins(figure3.3.1). Cytokines are soluble, biologically active glycoproteins, whichare secreted by host immuno-inflammatory cells. They have many functions andexert their effects in a paracrine or autocrine fashion to modulate inflammatoryand immune responses. It is likely that elevated levels of pro-inflammatorycytokines play a role in stimulating osteoclast activity. These cytokines are thekey of mediators in immunological and inflammatory responses, and have beenfound in measurable quantities in areas of active gingical inflammation.Therefore, the gingival tissues may produce various inflammatory cytokines.We demonstrated that the I-GMSC of inflammatory gingival tissues expressedsignificantly higher levels of IL-1α, IL-6, and TNF-α than the N-GMSC ofgingival tissues from healthy individuals did. Our data in the present studysuggest that cytokines are involved in chronic inflammatory conditions such asmechanical irritation-induced human gingival hyperplasia or gingivalinflammation.Experiment4, to better confirm that the N-GMSC and I-GMSC showedsome special characteristic in vitro, we proposed cell sheet engineering in vivo.A cell sheet was constructed firstly; then the cell sheet shrank and condensedinto an intact cell pellet; finally in vivo differentiation assay ofN-GMSC/N-GMSC (exogenous inflammatory factor) and N-DPSC/N-DPSC(exogenous inflammatory factor) pellets was performed using immunodeficientmice. All the cell pellets had good biology properties. It also found that MMPand TIMP had different changing tendency with N-GMSC (exogenousinflammatory factor) and N-DPSC (exogenous inflammatory factor). Theseresults showed that the DPSCs in inflammatory produced elevated levels ofthese MMP1, MMP2oppositing to that of GMSC (FIG.7C,D). This study showed that the expression of IL-1α, IL-6, col-1and TNF-α ofboth GMSC and DPSC increased in a time-dependant manner in vitro under aninflammatory microenvironment. However, as both GMSC and DPSC were twotissue-specific stem cells, so the expression of IL-1α in GMSC was relativelystronger than that in DPSC and the expression of IL-6in DPSC was relativelystronger than that in GMSC. Stem cells turned into TA in advance. After that,TA involved in the inflammatory reaction and its autocrine inflammatory factorwent into a positive feedback loop again. Do the expression differences of IL-1and IL-6between the stem cells derived from2kinds of tissue origins (GMSCand DPSC) under inflammatory microenvironment cause the difference ofchange trend of MMPs and MIMPs expression?In the study, it was found that the I-GMSC was actually the TA which wasturned from N-GMSC in advance by the action of inflammatory factor. It stillcould be cloned and the osteogenic and adipogenic ability decreased in themulti-directional differentiation potential stage. It showed strong proliferativecapacity, and increased the extracellular matrix, being consistent with theclinical phenotype. The inflammatory factors could increase the apoptosis ofN-GMSC and more at the differentiation potential stage. The phenotype wasregulated by MMPs, TIMPs, and inflammatory factors. Currently, the influenceof inflammation on GMSC is clear. While it is still unknown that how dospecific expression of the MMPs, TIMPs and inflammatory factors have aneffect on their clinical proliferative phenotypes, and it can be screened in thefuture.