hTERT-immortalized gingival fibroblasts respond to cytokines but fail to mimic primary cell responses to Porphyromonas gingivalis

In periodontitis, gingival fibroblasts (GFs) interact with and respond to oral pathogens, significantly contributing to perpetuation of chronic inflammation and tissue destruction. The aim of this study was to determine the usefulness of the recently released hTERT-immortalized GF (TIGF) cell line for studies of host–pathogen interactions. We show that TIGFs are unable to upregulate expression and production of interleukin (IL)-6, IL-8 and prostaglandin E2 upon infection with Porphyromonas gingivalis despite being susceptible to adhesion and invasion by this oral pathogen. In contrast, induction of inflammatory mediators in TNFα- or IL-1β-stimulated TIGFs is comparable to that observed in primary GFs. The inability of TIGFs to respond directly to P. gingivalis is caused by a specific defect in Toll-like receptor-2 (TLR2) expression, which is likely driven by TLR2 promoter hypermethylation. Consistently, TIGFs fail to upregulate inflammatory genes in response to the TLR2 agonists Pam2CSK4 and Pam3CSK4. These results identify important limitations of using TIGFs to study GF interaction with oral pathogens, though these cells may be useful for studies of TLR2-independent processes. Our observations also emphasize the importance of direct comparisons between immortalized and primary cells prior to using cell lines as models in studies of any biological processes.


Results
TIGFs internalize P. gingivalis but fail to upregulate inflammatory mediators after bacterial challenge. To verify whether TIGFs provide a useful tool for studies of interactions between oral pathogens and host cells, we compared adhesion and invasion of P. gingivalis in TIGFs and primary cells using the colonyforming assay. Significantly more bacteria adhered to the surface of TIGFs compared to PHGFs, and comparable numbers of live intracellular bacteria were detected in both cell types (Fig. 1a), indicating that TIGFs retain the ability to internalize bacteria. Next, we compared the effects of bacterial challenge on PHGF and TIGF inflammatory activation. Infection with increasing multiplicity of infection (MOI) of P. gingivalis resulted in a dosedependent increase in IL-6 and IL-8 production by PHGFs (Fig. 1b). However, the presence of bacteria had little impact on cytokine production by TIGFs, shown by only a slight increase in IL-6 and IL-8 levels in cells infected with the highest tested P. gingivalis MOI, compared to uninfected cells (Fig. 1b). P. gingivalis also induced a twofold increase in prostaglandin E2 (PGE2) secretion in PHGFs, whereas in TIGFs, PGE2 production remained unaffected (Fig. 1c). Consistently, mRNA expression of IL6, IL8, COX2, and mPGES1 was significantly upregulated in PHGFs, but not in TIGFs following infection (Fig. 1d). Of note, basal cytokine production as well as expression of IL6, IL8, and COX2 was evidently higher in TIGFs than in PHGFs (Fig. 1b,d). Taken together, these data showed that TIGFs are unable to mount an inflammatory response to P. gingivalis, despite their ability to secrete large amounts of cytokines under homeostatic conditions and to internalize bacteria.

TNFα and IL-1β induce comparable levels of inflammatory mediators in TIGFs and
PHGFs. Next, we tested whether the inability of TIGFs to upregulate cytokine and COX2 expression is specific for bacterial challenge, or whether it represents a global defect in inflammatory transcriptional programs. To verify this, we stimulated PHGFs and TIGFs with cytokines that play a central role in the pathogenesis of periodontitis and are present in large quantities in the inflamed gingival tissue 19 . TNFα and IL-1β stimulation potently induced IL6, IL8, and COX2 expression in both cell types, though at 4 h post stimulation gene induction by IL-1β was more pronounced in PHGFs than in TIGFs (Fig. 2a). Cytokine-induced IL-8 production by both cell types was also comparable (Fig. 2b), suggesting that TIGFs are capable of eliciting an inflammatory response.
Cytokines, but not P. gingivalis, stimulate NF-κB and MAP kinase signaling in TIGFs. Engagement of pattern recognition and cytokine receptors converge on the activation of NF-κB and mitogen-activated protein kinase (MAPK) signaling pathways 20 . To identify the molecular mechanism underlying the inability of TIGF to upregulate inflammatory gene transcription upon P. gingivalis infection, we analyzed the activation status of the components of these pathways known to be activated by cytokines and by oral pathogens: the p65 subunit of NF-κB, extracellular signal-regulated kinase (ERK), and p38 MAPK [21][22][23] . As expected, TNFα stimulation induced rapid phosphorylation of p38 and the p65 subunit of NF-κB, though some differences in the activation kinetics and signal intensity have been noted (Fig. 3a). In contrast, infection with P. gingivalis resulted in significant activation of both signaling molecules only in PHGFs, but not in TIGFs (Fig. 3b,c). ERK displayed high basal phosphorylation levels that remained largely unaffected by TNFα stimulation or P. gingivalis infection in both cell types (Fig. 3a,b). This suggests that alterations in early signaling events or bacterial recognition are responsible for the observed defect in TIGF inflammatory activation.
TIGFs lack TLR2 expression. The innate immune response to P. gingivalis is induced by pattern recognition receptors, among which TLR2 plays the most prominent role 24 . We hypothesized that alternations in TIGF Western blot analysis showed that TLR2 is readily detectable in PHGFs but is almost completely absent in TIGFs (Fig. 4a). This difference was also evident at the mRNA level and, in contrast to primary cells, the TLR2 transcript was not inducible by TNFα or IL-1β stimulation in TIGFs (Fig. 4b).
To verify the functional consequences of diminished TLR2 expression, we analyzed the effects of TLR2-specific ligands, Pam2CSK4 and Pam3CSK4, on TIGF inflam-  www.nature.com/scientificreports/ matory activation. Both ligands significantly upregulated the expression of IL6, IL8, and COX2 in PHGFs, while their effects on TIGF activation were negligible (Fig. 4c). In line with mRNA data, Pam2CSK4 and Pam3CSK4 induced secretion of IL-6 and IL-8 only in primary cells, but not in TIGFs (Fig. 4d). Epigenetic changes in DNA methylation play an important role in regulation of TLR2 expression 25 . To verify whether the observed reduction in TLR2 levels is mediated by epigenetic alterations, we compared TLR2 promoter methylation in PHGFs and TIGFs using the EpiTect methyl II PCR assay. We found that the TLR2 promoter region is almost completely methylated in TIGFs whereas it remains mostly unmethylated in PHGFs (Fig. 4e). This observation may explain the absence of TLR2 expression in resting TIGFs and the inability of cytokines to upregulate TLR2 in TIGFs despite intact inflammatory signaling cascades. Notably, this defect is specific for TLR2 as we did not observe a similar reduction in expression of TLR4 (Fig. 4f) or NOD-like receptors NOD1 and NOD2 ( Supplementary Fig. S3). In fact, we noted a trend towards elevated TLR4 mRNA and protein expression in TIGFs compared to PHGFs, though this was not statistically significant ( Fig. 4f,g). To confirm that TLR4 expressed in TIGFs is functional, we infected cells with Fusobacterium nucleatum. In contrast to P. gingivalis, which is recognized by host cells predominantly through TLR2 26,27 , F. nucleatum activates cells through both TLR4 and TLR2 28,29 . TIGF infection with F. nucleatum induced IL-6 and IL-8 secretion in TIGFs, albeit to a lesser extent than in primary cells (Fig. 4h), indicating that TIGFs are capable of responding to infection through TLR4, but this response is diminished compared to primary cells in the absence of TLR2. Finally, to exclude the possibility that the defect in TLR2 expression is specific for the batch of cells used in our experiments, we compared two independent batches of TIGFs that are used in different laboratories and found no significant differences between their inflammatory responses. Both TIGF batches (TIGF-1 and TIGF-2) were unable to upregulate IL-6 and IL-8 production upon P. gingivalis infection (Fig. 4i). Collectively, these results indicate that the observed defects in TIGF responses to P. gingivalis challenge can be attributed to diminished expression of TLR2 and the resulting inability of TIGFs to respond to TLR2-activating signals. . Data are shown as mean signal intensity relative to actin. *P < 0.05; ***P < 0.001; 1-way ANOVA followed by Bonferroni multiple comparison test.

Discussion
Human cells immortalized through the overexpression of hTERT represent invaluable tools in cell biology research. A plethora of cell types from various organs and tissues have been immortalized using this method, including foreskin keratinocytes 30 , dermal papilla cells 31 , intestinal epithelial cells 32 , urothelial cells 33 , and liver endothelial cells 34 . Importantly, these cells maintain the morphology, expression profiles of cell type-specific markers, and functional characteristics of primary cells. Cells of the tooth-supporting tissues have also been successfully immortalized by hTERT overexpression. hTERT-immortalized human dental papilla cells retain the undifferentiated state of primary cells and represent a valuable model for odontoblast differentiation 35 . Similarly, the recently generated hTERT-immortalized human gingival keratinocyte (TIGK) cell line largely recapitulates the phenotypic properties and inflammatory responses of primary GECs, therefore, TIGKs are widely used for analyses of interactions between host gingival cells and oral bacteria 17,36-38 . www.nature.com/scientificreports/ Here, we identify both strengths and limitations of TIGFs as a representation of primary cells for studies of GF activation upon interaction with oral pathogens and the inflammatory tissue environment. Comparing the immortalized cell line to primary GFs derived from ten healthy individuals, we reveal that TIGFs have a similar activation profile in response to stimulation with TNFα and IL-1β. In contrast, the inflammatory responses of TIGFs to oral pathogens are defective, despite similar levels of P. gingivalis adhesion and invasion, compared to primary GFs. This is likely due to diminished TLR2 expression which is essential for host responses to P. gingivalis 26,27 , but is only partially involved in cell activation by F. nucleatum 28,29 . Consistently, while inflammatory mediator induction by P. gingivalis is almost completely absent, TIGFs partly maintain the ability to respond to F. nucleatum. Importantly, expression of TLR4 does not compensate for the absence of TLR2 in TIGFs despite the reported ability of P. gingivalis LPS to activate TLR4 39 . This is not surprising in light of in vitro and in vivo studies demonstrating that host inflammatory responses to live P. gingivalis rely predominantly on TLR2, but not TLR4 activation 26,27,40 .
We show that TIGFs lose some phenotypic features of primary cells, while retaining others, despite their genomic stability and diploid status confirmed during extended culture 14 . The defect in TIGF activation upon infection with P. gingivalis observed in this study is accompanied by our previously reported intact response to gingivitis and cariogenic biofilms in organotypic culture 16 . In the latter experiments, biofilm was applied topically to reconstructed epithelium grown on a TIGF-populated hydrogel 16 . Such a method of exposure would result in release of GEC-derived cytokines, including IL-6 and IL-8, which would in turn stimulate inflammatory cytokine release from fibroblasts 41 . This would explain the apparently contradictory findings since the current study describes a direct TIGF exposure whereas an indirect exposure method was used in the previous work 16 . TIGFs also display a pattern of chemokine receptor expression similar to that observed in primary GFs 42 . Similarly, hTERT-immortalized airway smooth muscle (ASM) cells mirror the behavior of primary cells in their proliferative capacity and inflammatory responses, but their production of the ECM components fibronectin and fibulin-1 is defective 43 . Consistently, the comparison of immortalized cell lines derived from healthy donors and patients with asthma revealed that they retain some of the unique characteristics of asthmatic cells, such as hyperproliferation and excessive release of inflammatory mediators, but not the differences in ECM deposition that are seen in primary ASM cells 43 . These observations highlight the limitations of hTERT-immortalized cell lines, which tend to lose certain unique phenotypic properties of parent cells, rendering these systems not reliable for some functional studies. These specific defects can be identified only through direct, side-by-side comparisons with primary cells.
Our data indicate that the lack of TLR2 expression in TIGFs may be driven by a specific epigenetic defect in DNA methylation. Cell immortalization may promote changes in DNA methylation profiles that alter cellular functions 44 . For example, hTERT-mediated immortalization of human foreskin keratinocytes and adenoid epithelial cells causes repression of the cell cycle inhibitor p16 INK4a which is driven by increased promoter methylation and can be reverted by treatment with DNA methyltransferase inhibitors 45 . Similarly, hypermethylation of genes involved in DNA damage repair, apoptosis, hormone responses, and tumor invasion has been observed in a range of in vitro immortalized prostate epithelial cell lines 46 . However, relatively few changes have been noted in hTERT-immortalized cells compared to cell lines generated using different immortalization methods 46 . Genome-wide profiling of DNA methylation in human bronchial epithelial cells immortalized through the introduction of hTERT and SV40 LT revealed global genomic hypomethylation, as well as site-specific changes in DNA methylation patterns that affect genes involved in cell proliferation 47 . Interestingly, many of the epigenetic defects found in immortalized cell lines mirror those found in spontaneously transformed cells and cancer tissues 46,47 . It is well-established that expression of TLR2 is dynamically regulated by epigenetic mechanisms. Increased methylation of the TLR2 promoter region has been reported in periodontitis 48 and is induced in GECs by chronic exposure to oral pathogens 25 . It is therefore not surprising that TLR2 expression is altered in a cell line that has dysregulated epigenetic regulatory mechanisms. Curiously, TLR4 expression was not affected in TIGFs compared to primary cells, suggesting that epigenetic regulation of TLR2 and TLR4 is likely driven by independent mechanisms. Consistently, distinct TLR2 and TLR4 promoter methylation patterns have been reported in periodontitis and Behçet's disease 49,50 . While our results identify an aberrant promoter methylation profile of a single gene, the full scope of possible alterations in DNA methylation in TIGFs remains to be determined.
Collectively, the results presented here show important limitations of using TIGFs to evaluate GF behavior under inflammatory conditions, and provide an independent confirmation that TLR2 is indispensable for GF activation in response to infection with P. gingivalis. Nonetheless, given their genetic stability and intact signal transduction pathways that drive inflammatory gene transcription, TIGFs may be a useful model for studies of TLR2-independent processes. Finally, our data also highlight the notion that the study of any biological process in an immortalized cell line must first be preceded by a direct comparison with parent cells to confirm the similarity of cell responses under the tested experimental conditions.

Materials and methods
Cell culture. PHGFs were isolated from gingival biopsies collected from 10 healthy subjects presenting for orthodontic treatment at the Department of Periodontology and Clinical Oral Pathology, Faculty of Medicine, Jagiellonian University Medical College, Kraków, Poland. The study was approved by the Bioethical Committee of the Jagiellonian University in Krakow, Poland (approval numbers 122.6120.337.2016 and 1072.6120.104.2019) and all methods were performed in accordance with the regulations and guidelines of the Committee. Written informed consent was obtained from all donors. PHGFs were isolated as described previously 38  www.nature.com/scientificreports/ tamicin (50 U/ml) (Biowest). One day prior to experiments, cells were seeded in medium supplemented with 2% FBS, without antibiotics.

Bacterial culture, cell infection and treatment with cytokines or TLR2 ligands.
Wild-type P. gingivalis (strain ATCC33277) and F. nucleatum (strain ATCC10953) were grown anaerobically for 7 days on blood agar plates (brain-heart infusion [BHI, Becton-Dickinson] with yeast extract containing 0.5 mg/mL L-cysteine, 10 µg/mL hemin and 0.5 µg/mL vitamin K). Bacterial suspensions at optical density (OD) 600 = 1 (corresponding to 10 9 colony-forming units/ml) were prepared for cell infections as described before 51 . Figure  Colony-forming assay. Cells were seeded in duplicate wells in 12-well plates (1.25 × 10 5 cells per well) and infected with P. gingivalis (MOI 100) for 1 h, then washed 3 times with PBS. Cells were then cultured either in fresh in medium without antibiotics, or in medium containing gentamicin (2.5 mg/ml) and metronidazole (2 mg/ml) for 1 h, to discriminate between adherent and internalized bacteria. Cells were subsequently lysed in distilled, sterile H 2 O for 40 min. Cell lysates were serially diluted and 10 µl of each dilution was plated on BHI blood agar plates in triplicate. Plates were incubated for 5 days at 37 °C in anaerobic conditions, then bacterial colonies were counted.

RNA isolation and quantitative polymerase chain reaction (qPCR). PHGFs and TIGFs cultured in
12-well plates (2.5 × 10 5 cells per well) were either stimulated with cytokines or TLR2 ligands, or infected with P. gingivalis (MOI 100) for 24 h. Total RNA was extracted using an ExtractMe Total RNA isolation kit (Blirt) and quantified using a BioPhotometer D30 (Eppendorf). RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). PowerUp SYBR Green PCR mix (Applied Biosystems) was used for qPCR reactions that were performed on a CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad) and the obtained results were analyzed using the CFX Manager (Bio-Rad). mRNA expression relative to housekeeping genes (ACTB or RPLP0) was calculated using the ΔΔCT method, unless otherwise indicated. Sequences of the primers (purchased from Genomed S.A., Poland) used are listed in Table 1.
Immunoblotting. Cells cultured in 12-well plates (2.5 × 10 5 cells per well) were lysed in Laemmli's buffer containing 2% SDS, 10% glycerol and 125 mM Tris-HCl, pH 6.8. Protein concentration was measured using the Bradford assay (BioShop) and equal amounts of protein were resolved by electrophoresis on 10% polyacrylamide gels. Proteins were then transferred to PVDF membranes using a Trans-Blot Turbo Transfer System (Bio-Rad). The membranes were then blocked using 5% milk (BioShop) in TBS containing 0.1% Tween-20 (TBS/T, Bio-Shop) prior to incubation with primary antibodies specific for phosphorylated (p)-p38 (#9211), pERK (#9101), p-p65 (#3033), β-actin (#4967) (all from Cell Signaling Technology), TLR2 (clone JM22-41, #MA5-32787) or TLR4 (#48-2300) (both from Invitrogen), at 4 °C overnight. Subsequently, the membranes were washed in TBS/T and incubated with anti-rabbit immunoglobulin secondary antibodies conjugated to horseradish peroxi-  IL6  GAC AGC CAC TCA CCT CTT CA  CCT CTT TGC TGC TTT CAC AC   IL8  GCT CTG TGT GAA GGT GCA GT  CCA GAC AGA GCT CTC TTC CA   IL1B  ACA GAT GAA GTG CTC CTT CCA  GTC GGA GAT TCG TAG CTG GAT   COX2  AGC CCT TCC TCC TGT GCC T  AAT CAG GAA GCT GCT TTT TACCT   mPGES1  CAC GCT GCT GGT CAT CAA GAT  CCG TGT CTC AGG GCA TCC T   TLR2  CGG AAT GTC ACA GGA CAG CA  TAC CAC AGG CCA TGG AAA CG   TLR4  ACC ATC ATT GGT GTG TCG GTC  AGC CAG CAA GAA GCA TCA GG   NOD1  ACA GCC AGG GCG AGA TAC  AAA GGT GCT AAG CGA AGA G   NOD2  GAT TGG CTG CCT TCC TTC TA  GAG CGT CTC TGC TCC ATC AT   RPLP0  GCG TCC TCG TGG AAG TGA CATCG TCA GGG ATT GCC ACG CAG  DNA isolation and methylation-specific qPCR. Genomic DNA was isolated using a DNeasy Blood and Tissue Kit (Qiagen) and quantified with a BioPhotometer D30. DNA was digested using the EpiTect Methyl II DNA Restriction Kit (Qiagen) that contains methylation-dependent and methylation-sensitive restriction enzymes. The reaction mixture was incubated overnight at 37 °C, then incubated at 65 °C for 20 min to terminate the reaction. qPCR was performed using the TLR2 promoter CpG-specific primers (EPHS111307-1A; CpG Island 111307) and RT2 SYBR Green qPCR Mastermix (both from Qiagen). The reaction was conducted on a CFX96 Touch™ Real-Time PCR Detection System. The obtained results were analyzed using CFX Manager (Bio-Rad) and the percentage TLR2 promoter methylation was calculated.
Statistical analyses. Data  www.nature.com/scientificreports/ Reprints and permissions information is available at www.nature.com/reprints.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.