Abstract

   The present study explores for the first time the effect of hyperbaric
   oxygen (HBO) on gingival mesenchymal stem cells’ (G-MSCs) gene
   expression profile, intracellular pathway activation, pluripotency, and
   differentiation potential under an experimental inflammatory setup.
   G-MSCs were isolated from five healthy individuals (n = 5) and
   characterized. Single (24 h) or double (72 h) HBO stimulation (100% O2,
   3 bar, 90 min) was performed under experimental inflammatory [IL-1β (1
   ng/mL)/TNF-α (10 ng/mL)/IFN-γ (100 ng/mL)] and non-inflammatory
   micro-environment. Next Generation Sequencing and KEGG pathway
   enrichment analysis, G-MSCs’ pluripotency gene expression,
   Wnt-/β-catenin pathway activation, proliferation, colony formation, and
   differentiation were investigated. G-MSCs demonstrated all mesenchymal
   stem/progenitor cells’ characteristics. The beneficial effect of a
   single HBO stimulation was evident, with anti-inflammatory effects and
   induction of differentiation (TLL1, ID3, BHLHE40), proliferation/cell
   survival (BMF, ID3, TXNIP, PDK4, ABL2), migration (ABL2) and osteogenic
   differentiation (p < 0.05). A second HBO stimulation at 72 h had a
   detrimental effect, significantly increasing the inflammation-induced
   cellular stress and ROS accumulation through HMOX1, BHLHE40, and ARL4C
   amplification and pathway enrichment (p < 0.05). Results outline a
   positive short-term single HBO anti-inflammatory, regenerative, and
   differentiation stimulatory effect on G-MSCs. A second (72 h)
   stimulation is detrimental to the same properties. The current results
   could open new perspectives in the clinical application of short-termed
   HBO induction in G-MSCs-mediated periodontal reparative/regenerative
   mechanisms.

   Keywords: inflammation, stem cells, hyperbaric oxygen, regeneration,
   periodontitis, ROS, next-generation sequencing

1. Introduction

   Periodontitis is a multifactorial inflammatory disorder associated with
   microbial dysbiosis [[42]1]. The initial localized inflammatory
   reaction is characterized by an increased expression of the
   pro-inflammatory cytokines interleukin-1β (IL-1β), interleukin-4
   (IL-4), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α)
   [[43]2]. On prolonged periodontal inflammation, further expression of
   interferon-γ (INF-γ) results in additional cellular stress, osteoclast
   activation, bone resorption, and tissue destruction
   [[44]3,[45]4,[46]5]. Aside from the regenerative potential of
   periodontal mesenchymal stem/progenitor cells (MSCs), MSCs possess the
   ability to modulate such inflammatory responses via NF-κB- and
   Wnt-/β-catenin-pathway-dependent manners [[47]6,[48]7,[49]8]. Yet, with
   increased periodontal disease progression, such MSCs’
   reparative/regenerative potential would substantially be hampered by
   rising intracellular levels of reactive oxygen species (ROS) with
   subsequent caspase-3-dependent cellular apoptosis
   [[50]1,[51]5,[52]9,[53]10,[54]11,[55]12].

   Short-termed hyperbaric oxygen (HBO; 100% pressurized oxygen (≥1 atm))
   [[56]13,[57]14], therapeutically employed in non-healing chronic
   injuries, is believed to increase tissue oxygenation and ROS
   elimination, through elevation of the superoxide dismutase expression
   (SOD2), the antioxidant protein that catalyzes the mitochondrial ROS
   dismutation into oxygen and hydrogen peroxide [[58]15,[59]16,[60]17].
   HBO further reduced TNF-α, IL-6, and IL-10 expression and enhanced
   survival in a sepsis model [[61]18]. HBO stimulation was reported to
   activate the Wnt-/β-Catenin pathway in adipose-derived stem cells,
   Wharton jelly mesenchymal stem cells, and neural stem cells
   [[62]19,[63]20,[64]21] and to improve MSCs proliferation and
   differentiation [[65]22,[66]23,[67]24]. Earlier results further
   suggested a positive impact of HBO-therapy on clinical attachment
   levels, bleeding scores, and microbial compositions in periodontitis
   patients [[68]25,[69]26]. Thus, the current study explored for the
   first time the effect of a single (24 h) or double (72 h) short-termed
   HBO therapy on G-MSCs’ mRNA expression and intracellular pathway
   activation via Next Generation Sequencing (NGS) and KEGG pathway
   analysis, pluripotency gene expression, Wnt-/β-catenin pathway
   activation, proliferation, colony-formation and differentiation under
   experimental inflammatory and non-inflammatory conditions.

2. Materials and Methods

2.1. Cell Culturing

   The study’s protocol was approved by the ethical committee of the
   Christian Albrechts University of Kiel, Kiel, Germany (IRB-Approval
   D513/17). After obtaining the patients’ informed consents, retromolar
   gingiva samples were surgically obtained from five randomly recruited
   healthy individuals (n = 5; [70]Supplementary Table S1), during
   elective interventions. Cells were isolated, cultured and STRO-1
   immunomagnetically sorted ([71]Table 1). Subsequently, cultures of
   STRO-1+ sorted G-MSCs [[72]27] were purified by single colony
   selection.

Table 1.

   Antibodies used for MACS sorting and flow cytometry.
   Antibody Clone Color Distributor Product/Cat. No.
   anti-human
   CD14 REA599 FITC Miltenyi Biotec (Bergisch Gladbach, Germany)
   130-110-576
   anti-human
   CD34 AC136 PE Miltenyi Biotec 130-113-741
   anti-human
   CD45 REA747 APC Miltenyi Biotec 130-110-771
   anti-human
   CD73 REA804 PE Miltenyi Biotec 130-112-060
   anti-human
   CD90 REA897 FITC Miltenyi Biotec 130-114-901
   anti-human
   CD105 REA794 FITC Miltenyi Biotec 130-112-327
   REA Control (S) REA293 FITC Miltenyi Biotec 130-113-437
   REA Control (S) REA293 PE Miltenyi Biotec 130-113-438
   anti-human
   STRO-1 STRO-1 Alexa Fluor 647 BioLegend (San Diego, CA, USA) 340103
   anti-Cy5/Alexa Fluor 647
   MicroBeads - - Miltenyi Biotec 130-091-395
   [73]Open in a new tab

2.2. Experimental Groups

   Third passage G-MSCs were preconditioned in the following four groups
   and received a single treatment at 24 h or a double treatment over 72 h
   (US-Navy treatment protocol Table 9) [[74]28].

2.3. Control Group: Basic Medium

   HBO-group: basic medium supplemented with hyperbaric oxygen (100% O2,
   [Hüllmann Gas, Hans Hüllmann GmbH, Kiel, Germany], 90 min, 2–3 bar).

   Inflammatory group: basic medium supplemented with inflammatory
   cytokines (IL-1β [1 ng/mL], TNF-α [10 ng/mL], and IFN-γ [100 ng/mL])
   [[75]29,[76]30,[77]31,[78]32,[79]33,[80]34].

   HBO/inflammatory group: basic medium with hyperbaric oxygen (100% O2,
   [Hüllmann Gas], 90 min, 2–3 bar) and inflammatory cytokines (IL-1β [1
   ng/mL], TNF-α [10 ng/mL] and IFN-γ [100 ng/mL]).

   Media were renewed every second day. To create hyperbaric conditions,
   samples were placed in an HBO-chamber system (Testcom200/20,
   Haux-Life-Support, Karlsbad-Ittersbach, Germany), nitrogen layers were
   exhausted, by applying pure O2, and the pressure of 2–3 bar was
   built-up, following an established protocol [[81]31] composed of
   compression (10 min), supplementation of isobaric oxygen (90 min) and
   finally decompression (10 min). Pneumatic use of pure oxygen is
   prohibited due to its high reactivity (IGV-SH-07S-Rev1). Instead,
   samples were placed in a compressed air chamber and received constant
   oxygen ventilation through individually constructed covers
   ([82]Supplementary Figures S1 and S2).

2.4. Flow Cytometric Analysis

   Third passage G-MSCs were characterized for their surface marker
   expression (CD14, CD34, CD45, CD73, CD90, CD105)
   [[83]35,[84]36,[85]37]. Cells were bound to fluorochrome-linked
   antibodies and measured against their corresponding isotype controls
   (FACSCalibur E6370 and FACSComp 5.1.1 software [BD Biosciences,
   Franklin Lakes, NJ, USA]). Background noise was reduced with an
   FcR-blocking reagent (Miltenyi Biotec) according to the manufacturers’
   instructions.

2.5. Real-Time-Polymerase-Chain-Reaction

   In the third passage G-MSCs, NANOG (NANOG homeobox protein), OCT4A
   (octamer-binding transcription factor 4A; Pou5f1) and SOX2 (sex
   determining region Y-box containing gene 2) mRNA expression was
   measured via real-time quantitative polymerase chain reaction
   (RT-qPCR). Reaching 80% confluence, mRNA extraction and cDNA synthesis
   were performed using Rneasy^® Protect Mini Kit (250) and QuantiTect^®
   Reverse Transcription Kit (Qiagen, Hilden, Germany). Each sample
   received a 20 µL reaction mixture [4 pmol primer, 10 μL LightCycler
   Probes Master mixture (Roche Life Science, Prenzberg, Germany), and 5
   μL specimen cDNA]. RT-qPCR was run for 40 cycles (Light Cycler 96;
   Roche). All trials were performed thrice and averaged. mRNA expression
   was normalized to PGK1 (phosphoglycerate kinase 1) housekeeping gene.
   PGK1 was identified as the most suitable reference of 19 reference
   genes via NormFinder algorithm (v.05/01-2015; Department of Molecular
   Medicine; Aarhus University Hospital, Denmark). Relative quantification
   of gene expression levels was calculated via the 2^−ΔΔCt method.

2.6. Next Generation Sequencing

   The setup described under “Experimental groups” allows a variety of
   possible comparisons and modeling methods with regard to differential
   RNA expression. In total, RNA was extracted from 28 samples from 4
   probands. However, due to sequencing failures in some samples, to keep
   this analysis as concise as possible and in response to the dominating
   effects of inflammation conditions, the main focus of NGS analyses was
   put on differential expression analysis (DEA) in HBO/inflammation
   samples compared to the inflammation samples directly, as well as on
   HBO treatment under normal (non-inflammation) conditions. For this, 16
   paired samples taken from two different probands were used. DEA was
   done after single (24 h) and double (72 h) HBO stimulation. Next
   Generation Sequencing was performed at the NGS lab at the Institute of
   Clinical Molecular Biology (IKMB) in Kiel (Illumina MiSeq; San Diego,
   CA, USA). Raw FastQ files were aligned, quality controlled, and
   transformed into read counts using the Nextflow nfcore/RNAseq pipeline
   [[86]35]. Read counts were analyzed in R v3.6.2 via edgeR and DeSEQ2
   packages. Gene counts were rlog transformed and plotted in heatmaps in
   DeSEQ2. Differential expression analysis (DEA) was done in edgeR using
   the Quasi-Likelihood F-test function with multi-level contrasting.
   Contrasts in the main analysis were modeled for the effect of HBO under
   inflammation and normal conditions at 24 h and 72 h. [87]Supplementary
   Table S2 provides an overview of additional contrasts for the effect of
   inflammatory conditions compared to normal conditions without HBO
   treatment. In the correction of false discovery rates (FDR), the
   Benjamin–Hochberg method was used. Pathway enrichment analysis was done
   for all differentially expressed genes using the Kyoto Encyclopedia of
   Genes and Genomes (KEGG) database and visualized using the R package
   “clusterprofiler”.

2.7. ELISA pS45/Total β-Catenin

   Phospho-β-catenin 45 (pS45) and total β-catenin were measured following
   a single HBO stimulation (24 h) in each group. ELISA was performed
   using ab205705β-Catenin (pS45) and total β-catenin SimpleStep
   ELISA^®Kit (Abcam, Cambridge, UK), following the manufacturer’s
   instructions. 2 × 104 G-MSCs/well were seeded out on 24-well-palates.
   After a single stimulation at 24 h, samples were washed twice (PBS
   [Biochrom]) and lysated via 1× Cell Extraction Buffer PTR (100–500
   µg/mL of protein). 50 µL of each sample were transferred to the
   antibody-coated 96-well-plate and primary and secondary antibodies
   (β-Catenin (Total) Capture Antibody/β-Catenin (Total) Detector Antibody
   and β-Catenin (pS45) Capture Antibody/b-Catenin (pS45) Detector
   Antibody) were added. Redundant antibodies were removed by washing four
   times with 350 µL 1× Wash Buffer PT one hour later. After pitching the
   supernatants, 100 µL of TMB Substrate was added followed by incubation
   (14 min, shaking 400 rpm, in the dark). Finally, the reaction was
   stopped by adding 100 µL Stop Solution. Optical density was measured
   and compared to standard curves (λ = 450 nm; MultiscanGo, Thermo Fisher
   Scientific, Waltham, MA, USA).

2.8. Cellular Proliferation and Colony-Forming Units (CFUs)

   Cellular proliferation was examined over 12 days. 1 × 10^3 third
   passage G-MSCs were seeded on 24-well-hard-shell plates and attached
   for 24 h. One well of each group was counted every 24 h (Neubauer
   chamber, Roth, Karlsruhe, Germany).

   1 × 102 third passage G-MSCs (1.65 cells/cm^2) were seeded out on cell
   culture dishes and grew for 12 days. Colonies were fixed with 80%
   ethanol (Sigma-Aldrich, St. Louis, MO, USA) and stained (crystal
   violet, Merck, Darmstadt, Germany). Clusters consisting of a minimum of
   50 cells were counted as colonies. Three independent investigators
   counted colonies and the arithmetical averages were calculated and
   statistically evaluated.

2.9. Multilinear Differentiation

   Following a single stimulation in the respective groups, G-MSCs were
   examined for their multilinear differentiation potential into
   osteogenic, adipogenic, and chondrogenic directions as described before
   [[88]38].

2.10. Statistical Analysis

   Data was checked for normality, using the Shapiro-Wilk-test and tested
   for homogeneity of variance by Levene testing and graphical
   exploration. Significances between groups regarding gene expression,
   β-catenin (total/pS45) activation, colony formation, and multilinear
   differentiation were tested via Friedman non-parametric ANOVA.
   Post-hoc, pairwise comparisons were performed via
   Wilcoxon-signed-rank-test. The level of significance was set as p ≤
   0.05. Multiple testing was adjusted via the Bonferroni method. Data
   transformation and testing were calculated with MS Excel (Microsoft,
   Washington, DC, USA, v.16.0) and SPSS (IBM, Chicago, IL, USA,
   v.18.0.0).

2.11. Data Availability

   FastQ files were uploaded to the Gene Expression Omnibus (GEO) platform
   and can be accessed publicly ([89]GSE244404). Further, all data will be
   available from the corresponding author upon reasonable request.

3. Results

3.1. G-MSCs’ Characterization

   Cells grew out of their soft tissue masses ([90]Figure 1A) and
   demonstrated fibroblast-like condensed colonies ([91]Figure 1B).
   G-MSCs’ flowcytometric ([92]Figure 1C) surface marker profiling showed
   negative CD14, CD34, and minute CD45 expression, whilst high expression
   of CD73, CD90, and CD105. Following two weeks of osteogenic
   stimulation, cultures developed distinct calcification ([93]Figure 1D)
   compared to their controls ([94]Figure 1E). Three weeks after
   adipogenic induction, G-MSCs exhibited Red-Oil-O positive lipid droplet
   formation ([95]Figure 1F) compared to their controls ([96]Figure 1G).
   Chondrogenic differentiated cell masses ([97]Figure 1H) demonstrated
   dense rather red-stained cortex (1) and loose blue-colored central (2)
   areas five weeks following stimulation, in contrast to their controls
   ([98]Figure 1I) showing homogenous Nuclear-Fast-Red staining.

Figure 1.

   [99]Figure 1
   [100]Open in a new tab

   Tissue cultivation and stem cell characteristics. Gingiva sample and
   outgrowing mixed culture (A). Colony formation of MACS^+ sorted cells
   after 14 days (B). G-MSC specific phenotyping via flow cytometry (C).
   Multilinear differentiation potential (D–I): Alizarin red staining
   after osteogenic stimulation (D) and control (E). Red Oil O staining
   after adipogenic stimulation (F) and control (G). Alcian Blue and
   Nuclear Fast Red stained paraffin-embedded histological samples after 5
   weeks of chondrogenic stimulation (H) and control (I).

3.2. Next Generation Sequencing

   DEA for HBO treatment under different conditions revealed 255 genes
   that were differentially regulated for the effect of HBO treatment
   under standard conditions after a single (24 h) stimulation, whereas
   the number of DE genes increased to 1999 with double HBO treatment
   after 72 h ([101]Figure 2A,B). Likewise, for the effect of HBO
   treatment under inflammation conditions 134 genes were differentially
   expressed at 24 h, and 1536 genes following 72 h HBO stimulation
   ([102]Figure 2C,D). The top three differentially expressed genes in
   both groups are shown below ([103]Table 2); 24 h HBO treatment under
   standard conditions resulted in three overrepresented KEGG pathways:
   Human papillomavirus infection, calcium signaling pathway, and breast
   cancer ([104]Figure 2C). After 72 h, we observed an activation of many
   genes in the Zinc-finger-protein family, with an underexpression of the
   Herpes simplex virus infection pathway. In contrast, genes in the BHLE
   family were overexpressed, resulting in the activation of the circadian
   rhythm pathway ([105]Figure 2C). The breast cancer pathway was further
   activated through the FGF18, HEYL, and WNT11 genes. Regarding 24 h HBO
   treatment under inflammation, the downregulation of several genes in
   the Inhibitor of the DNA binding family resulted in the activation of
   the TGF-β signaling pathway, regulating the pluripotency of stem cells
   and notch signaling ([106]Figure 2D). With 72 h HBO treatment an
   activation of the circadian clock pathway through genes in the ZNF
   family, as well as the breast cancer pathway was observed ([107]Figure
   2D). For completeness, [108]Supplementary Table S2 provides results of
   additional contrasts for the effect of inflammatory conditions compared
   to normal conditions without HBO treatment at 24 h and 72 h. Comparison
   of inflammatory medium to standard medium revealed, as expected, a
   number of DE genes involved in inflammatory processes (e.g., TNFAIP3)
   and the activation of the cytokine–cytokine interaction KEGG pathway
   ([109]Table 3).

Figure 2.

   [110]Figure 2
   [111]Open in a new tab

   Visualization of gene counts and differentially expressed genes in KEGG
   pathway analysis. Visualization of gene counts and differentially
   expressed genes in control-HBO (A) and Infl.-inf./HBO (B) comparisons:
   Heatmap of rlog transformed gene counts for all samples following 24 h
   and 72 h prestimulation. Kyoto Encyclopedia of Genes and Genomes (KEGG)
   Pathway analysis (C,D) of all DE genes during 24 h/72 h HBO-stimulation
   under non-inflammation (C) and inflammation conditions (D). Fold
   changes between control and treatment gene expression were highlighted
   via a color gradient.

Table 2.

   Top 3 differentially expressed genes during HBO stimulation under
   inflammation and non-inflammation conditions.
   Contrast Time Gene Name Name LogFC p-Value p-Value[BH Adjust]
   non-inflammation 24 h PTK2B Protein Tyrosine Kinase 2 Beta −1.0 2.28 ×
   10^−17 3.24 × 10^−13
   BMF Bcl2 Modifying Factor −1.0 2.34 × 10^−15 1.66 × 10^−11
   TLL1 Tolloid Like 1 −1.0 6.72 × 10^−13 3.19 × 10^−9
   72 h TXNIP Thioredoxin Interacting Protein −2.4 5.57 × 10^−27 7.90 ×
   10^−23
   PDK4 Pyruvate Dehydrogenase Kinase 4 −2.5 2.21 × 10^−26 1.57 × 10^−22
   ABL2 ABL Proto-Oncogene 2, Non-Receptor Tyrosine Kinase 1.7 2.24 ×
   10^−24 1.06 × 10^−20
   inflammation 24 h PI16 Peptidase Inhibitor 16 −6.0 8.90 × 10^−22 1.24 ×
   10^−17
   HMOX1 Heme Oxygenase 1 −1.0 7.60 × 10^−15 5.28 × 10^−11
   ID3 Inhibitor Of DNA Binding 3, HLH Protein −1.0 1.20 × 10^−14 5.56 ×
   10^−11
   72 h BHLHE40 Basic Helix-Loop-Helix Family Member E40 3.0 1.06 × 10^−8
   0.0001
   ARL4C ADP Ribosylation Factor Like GTPase 4C 1.6 5.16 × 10^−7 0.003
   HMOX1 Heme Oxygenase 1 1.7 6.10 × 10^−7 0.003
   [112]Open in a new tab

   Effects have been adjusted for the influence of different probands
   between control and HBO (non-inflammation) such as infl. and HBO/infl
   (inflammation) groups. LogFC = log Fold Change, LogCPM = log counts per
   million. Correction for multiple testing was done with the
   Benjamini-Hochberg method, level of significance was set to FDR <0.05.

Table 3.

   Primers used for RT-qPCR.
   Gene Assay ID Gene
   Symbol Accession ID Primer Sequence
   PGK-1 102083 PGK1
   H. sapiens ENST00000373316 FWD: GGAGAACCTCCGCTTTCAT
   REV: GCTGGCTCGGCTTTAACC
   NANOG 148147 NANOG
   H. sapiens ENST00000229307 FWD: GAGATGCCTCACACGGAGAC
   REV: AGGGCTGTCCTGAATAAGCA
   OCT4A 113034 OCT4A
   H. sapiens ENST00000259915 FWD: GCAAAACCCGGAGGAGTC
   REV: TCCCAGGGTGATCCTCTTCT
   SOX2 111867 SOX2
   H. sapiens ENST00000325404 FWD: ATGGGTTCGGTGGTCAAGT
   REV: GGAGGAAGAGGTAACCACAGG
   RUNX2 113380 RUNX2
   H. sapiens ENST00000359524 FWD: GCCTAGGCGCATTTCAGAT
   REV: CTGAGAGTGGAAGGCCAGAG
   ALP 103448 ALP
   H. sapiens ENST00000374840 FWD: AGAACCCCAAAGGCTTCTTC
   REV: CTTGGCTTTTCCTTCATGGT
   PPAR γ 110607 PPARγ
   H. sapiens ENST00000287820 FWD: GACAGGAAAGACAACAGACAAATC
   REV: GGGGTGATGTGTTTGAACTTG
   LPL 113230 LPL
   H. sapiens ENST00000311322 FWD: TCGTTCTCAGATGCCCTACA
   REV: GCCTGATTGGTATGGGTTTC
   ACAN 138057 ACAN
   H. sapiens ENST00000439576 FWD: GAACGACAGGACCATCGAA
   REV: AAAGTTGTCAGGCTGGTTGG
   [113]Open in a new tab

   PGK-1 (phosphoglycerate kinase 1), NANOG (NANOG Homeobox), OCT4A
   (octamer-binding transcription factor 4), SOX2 (sex determining region
   Y-box 2), RUNX2 (Runt-related transcription factor 2), ALP (alkaline
   phosphatase), PPARγ (peroxisome proliferator-activated receptor γ), LPL
   (lipoprotein lipase), ACAN (aggrecan). All primers were supplied by
   Roche (Mannheim, Germany).

3.3. Pluripotency Genes

   NANOG expression after a single (24 h) stimulation (median gene
   expression/PGK1; Q25/Q75) was highest in inflammatory—(0.0002;
   0.0001/0.0008), followed by HBO—(0.0001;0.0001/0.0012) with equally low
   levels in control—(0.0000; 0.0000/0.0002) and HBO/inflammatory-group
   (0.0000; 0.0000/0.0001; p > 0.05). Following a double (72 h)
   stimulation, NANOG was highest in inflammatory—(0.0003; 0.0001/0.0150)
   and control—(0.0002; 0.0001/0.0003), but nearly absent in
   HBO/inflammatory—(0.0000; 0.0000/0.0009) and HBO-group (0.0000;
   0.0001/0.0003; p > 0.05; [114]Figure 3A). SOX2 expression after 24 h
   stimulation was highest in HBO—(0.0003; 0.0001/0.0045) followed by
   control—(0.0003; 0.0001/0.0007) and HBO/inflammatory—(0.0001;
   0.0001/0.0002) with almost no expression in inflammatory-group (0.0000;
   0.0000/0.0050). After 72 h, expression was highest in
   inflammatory—(0.0006; 0.0000/0.0011), followed by control—(0.0003;
   0.0000/0.0007), HBO/inflammatory—(0.0001; 0.0000/0.0033) and HBO-group
   (0.0001; 0.0000/0.0023; p > 0.05; [115]Figure 3B). OCT4A expression
   after 24 h was highest in HBO—(0.0007; 0.0002/0.0016) followed by
   control—(0.0002; 0.0001/0.0007), HBO/inflammatory—(0.0002;
   0.0000/0.0007) and inflammatory-group (0.0001; 0.0000/0.0017). After 72
   h, OCT4A was almost equally expressed in control—(0.0005;
   0.0001/0.0008) and inflammatory—(0.0005; 0.0001/0.0007) as well as
   HBO—(0.0002; 0.0000/0.0005) and HBO/inflammatory-group (0.0001;
   0.0000/0.0015; p > 0.05; [116]Figure 3C).

Figure 3.

   [117]Figure 3
   [118]Open in a new tab

   Key transcription factor expression profile under 24 h and 72 h
   HBO/inflammation conditions. Real-time PCR analysis of NANOG (NANOG
   Homeobox; (A)), SOX2 (sex determining region Y-box 2; (B)) and OCT4A
   (octamer-binding transcription factor 4; (C)) expression profiles after
   24 h and 72 h inflammation/HBO prestimulation. (n = 5; box-and-whiskers
   diagrams with medians and quartiles; outliers are shown as dots).

3.4. ELISA for Total and pS45-β-Catenin

   Total β-catenin levels (median concentration [µg/mL]; Q25/Q75) were
   highest in inflammatory—(497.00; 417.07/551.82), followed by
   HBO/inflammatory—(457.38; 389.20/554.59), control—(447.08;
   413.85/472.13) and HBO-group (415.92; 362.21/496.54; p > 0.05;
   [119]Figure 4A). pS45 β-catenin was highest in HBO—(175.36;
   131.72/182.88), followed by inflammatory—(170.03; 134.39/201.30),
   HBO/inflammatory—(168.57; 134.15/207.77) and control-group (162.27;
   126.55/182.63; p > 0.05). Median pS45/total-β-catenin (Q25/Q75) was
   highest in HBO/inflammatory—(0.356; 0.329/0.401), followed by
   control—(0.353; 0.307/0.392), HBO—(0.346; 0.335/0.447) and
   inflammatory-group (0.336; 0.272/0.444; p > 0.05).

Figure 4.

   [120]Figure 4
   [121]Open in a new tab

   Proliferation assay, colony formation and ELISA after 24 h
   HBO/inflammation prestimulation. G-MSCs were stimulated over 24 h.
   pS45- and total β-Catenin expression were analyzed in enzyme-linked
   immunosorbent assays (A). Proliferation assay was observed over 12 days
   (B) and colonies were counted after 12 days (C1,C2). (n = 5;
   box-and-whiskers diagrams with medians and quartiles; outliers are
   shown as dots). Significance in Wilcoxon-testing is asterisked (* p ≤
   0.05).

3.5. Cellular Proliferation and CFUs

   Cell counts increased in all groups over 12 days. After 3 days, median
   cell numbers (Q25/Q75) were highest in control—(67500; 27500/68750),
   followed by inflammatory—(40000; 31250/102500), HBO/inflammatory
   (27500; 21250/53750) and HBO-group (25000; 18750/56250; p = 0.005;
   padjust = 0.02; [122]Figure 4B). After 6 days, cells grew most in the
   control group (162500; 118750/167500) followed by a nearly equal growth
   in inflammatory—(122500; 98750/186250) and HBO group (122500;
   100000/155000) and were least in HBO-inflammatory-group (110000;
   73750/157500; p = 0.088; padjust = 0.352). After 9 days, counts were
   highest in inflammatory—(200000; 156250/213750), followed by control
   (192500; 153750/233750), HBO/inflammatory—(180000; 123750/202500) and
   HBO-group (170000; 131250/222500; p = 0.069; padjust = 0.276). Finally,
   after 12 days, cell count was highest in inflammatory—(237500;
   185000/266250), followed by control—(227500; 180000/276250),
   HBO/inflammatory—(220000; 165000/243750) and HBO-group (217500; 167500/
   268750; p = 0.472; padjust > 1). Colony formation (median CFUs count;
   Q25/Q75) was highest in HBO—(73.67; 13.16/81), followed by
   control—(67.00; 14.00/77.00), inflammatory—(53.00; 14.33/108.16) and
   HBO/inflammatory-group (37.67; 9.83/78.17; p > 0.05; [123]Figure 4C).

3.6. Multilinear Differentiation Potential

   Osteogenic differentiation (relative Ca2+ content; Q25/Q75) was
   significantly higher in HBO—(0.201; 0.128/0.251) and
   HBO/inflammatory—(0.158; 0.096/0.222) compared to control—(0.016;
   0.008/0.178) and inflammatory-group (0.013; 0.008/0.115; p < 0.05;
   [124]Figure 5(A1)). Following adipogenic differentiation, lipid
   formation (relative lipid content; Q25/Q75) was highest in HBO—(0.027;
   0.016/0.046) and HBO/inflammatory—(0.024; 0.018/0.045), followed by
   control—(0.023; 0.020/0.031) and finally inflammatory-group (0.020;
   0.012/0.036; p > 0.05; [125]Figure 5(B1)). Regarding chondrogenic
   differentiation, Alcian-Blue staining (relative pixel content Q25/Q75)
   was highest in HBO—(0.658; 0.576/0.704), closely followed by
   control—(0.641; 0.487/0.661), HBO/inflammatory—(0.625; 0.608/0.711) and
   finally inflammatory-group (0.549; 0.467/0.627; p > 0.05).
   Nuclear-Fast-Red staining was highest in inflammatory—(0.438;
   0.361/0.518) followed by HBO/inflammatory—(0.357; 0.278/0.379),
   control—(0.347; 0.327/0.475) and HBO group (0.330; 0.286/0.411; p >
   0.05; [126]Figure 5(C1)).

Figure 5.

   [127]Figure 5
   [128]Open in a new tab

   Multilineage differentiation potential under 24 h/72 h HBO/inflammation
   stimulation. Multilinear differentiation was induced after 24 h
   inflammation/HBO prestimulation over 3 weeks (osteogenic
   differentiation; (A)), 4 weeks (adipogenic differentiation; (B)), and 5
   weeks (chondrogenic differentiation; (C)). Dyes (A2,B2) were quantified
   by elution of Alizarin Red S (A1) and Red Oil O (B1), and Alcian Blue
   (C2) was quantified in digital image quantification (C1).
   Real-time-qPCR analysis was performed for osteogenic markers RUNX2
   (Runt-related transcription factor 2; (A3)) and ALP (alkaline
   phosphatase; (A4)), adipogenic markers PPARγ (peroxisome
   proliferator-activated receptor γ; (B3)) and LPL (lipoprotein lipase;
   (B4)) such as chondrogenic marker ACAN (aggrecan; (C3)) (n = 5;
   box-and-whiskers diagrams with medians and quartiles, outliers are
   shown as dots). Significance in Wilcoxon-signed-testing is asterisked
   (* p ≤ 0.05).

   Regarding differentiation gene expressions, following osteogenic
   differentiation and 24 h pre-stimulation, the expression of the RUNX2
   early osteogenic marker was highest in HBO/inflammatory—(0.1020;
   0.0512/0.1430) and HBO—(0.0813; 0.0131/0.1580) but lower in
   inflammatory—(0.1020; 0.0521/0.1430) and control group (0.0356;
   0.023/0.1487; p > 0.05; [129]Figure 5(A3)). ALP was highest
   HBO/inflammatory—(0.0006; 0.0004/0.0019), followed by nearly equal
   expression in inflammatory—(0.0004; 0.0004/0.0008) and HBO—(0.0004;
   0.0001/0.0029) and lowest expression in control (0.0003; 0.0001/0.0019;
   p > 0.05; [130]Figure 5(A4)). In adipogenic differentiation and 24 h
   pre-stimulation, PPARγ expression was highest in inflammatory—(0.2700;
   0.1050/0.3650), followed by control—(0.1200; 0.1200/0.6450),
   HBO—(0.0740; 0.0480/0.1500) and HBO/inflammatory-group (0.0680;
   0.0210/0.1750; p > 0.05; [131]Figure 5(B3)). LPL expression after 24 h
   pre-stimulation was nearly equal in inflammatory—(0.0170;
   0.0145/0.0420) and HBO—(0.0170; 0.0091/0.0265), followed by control
   (0.0150; 0.0067/0.0505) and HBO/inflammatory-group (0.0100;
   0.0039/0.0845; p > 0.05; [132]Figure 5(B4)). ACAN expression after 24 h
   HBO/inflammation stimulation was highest in inflammatory—(3.2772;
   0.8235/8.8987), followed by HBO/inflammatory—(1.5899; 0.4853/8.1237),
   control—(1.4651; 0.4623/2.423) and HBO-group (0.6404; 0.2040/4.2633; p
   < 0.05; [133]Figure 5(C3)).

4. Discussion

   G-MSCs represent a readily available minimally invasive source of
   stem/progenitor cells, with significant proliferation, differentiation,
   regeneration, and inflammation-modulatory attributes
   [[134]7,[135]11,[136]38,[137]39,[138]40,[139]41,[140]42]. Similar to
   previous investigations, G-MSCs isolation was performed, employing the
   STRO-1 immunomagnetic separation method [[141]27,[142]35,[143]43]. The
   isolated G-MSCs demonstrated all predefined MSCs’ characteristics, with
   remarkable CFUs, surface markers expression pattern, and multilineage
   differentiation aptitude [[144]27,[145]37]. The current study employed
   an experimental IL-1β/TNF-α/IFN-γ inflammatory cocktail to test the
   effect of a single (at 24 h) or double (at 72 h) HBO induction, with or
   without inflammatory stimuli, on the G-MSCs intracellular pathway
   activation, pluripotency gene expression, Wnt-/β-catenin pathway
   activation, proliferation, colony-formation and differentiation in
   vitro. The current investigation employed a short-termed 90 min HBO
   stimulation scheme, following the US Navy clinical protocol standard
   [[146]28] and earlier clinical investigations [[147]25,[148]26],
   avoiding long-term HBO stimulation (i.e., continuous 24 h or 48 h 100%
   O2 stimulation) [[149]44,[150]45], which could result in oxygen
   intoxication.

   Differential gene expression is by nature versatile. To describe the
   dominant alterations, we primarily focused on the top three HBO-induced
   DE genes. Interestingly, these genes strongly correlated with
   differentiation (TLL1, ID3, BHLHE40), proliferation/cell survival (BMF,
   ID3, TXNIP, PDK4, ABL2), and migration (ABL2). Similar to previous
   investigations [[151]46], current NGS results revealed that HBO
   treatment counteracted inflammation-induced gene activation in G-MSCs.
   A single HBO stimulation activated the calcium signaling pathway,
   silencing the phosphatidylinositol signaling branch via PTK2B, FGF18,
   and GRPR inhibition [[152]47], with a subsequent hampering of
   PIP3-dependent ROS synthesis, strengthening oxidative stress resistance
   [[153]48] at 24 h. The anti-inflammatory effect of a single HBO
   application was further evident through the downregulation of HMOX1,
   primarily implicated in a multitude of inflammatory diseases (i.e.,
   Crohn’s disease and ulcerative colitis). Further, downregulation of the
   Wnt11 dependent branches in Brest Cancer and human papillomavirus
   infection pathways indicated reduced signaling within Wnt/β-Catenin
   signaling. Wnt/β-Catenin activation, however, is crucially associated
   with the early cytokine-induced inflammatory reaction and cell cycle
   progression [[154]31]. HEYL, FGF18, LFNG, and PHKG1 are part of MAPK
   [[155]49] and Notch signaling branches, affecting the overall
   expression of the key inflammatory transcription factor NFκB [[156]50]
   and proliferation/cell survival. The effect eventually affected the
   anti-inflammatory protection against an inflammatory stimulus after 24
   h. Interestingly, the effect primarily targeted pro-inflammatory
   signaling (TGF-β/Notch signaling) and signaling pathways regulating
   pluripotency of stem cells but strongly enhanced the BMPR1B-associated
   differentiation potential [[157]51].

   Yet, with a second HBO treatment an opposite effect was evident at 72
   h, with upregulated HMOX1, indicating an HBO/inflammation augmented ROS
   production [[158]52], in addition to FGF18, IGF1 knock-down in Breast
   Cancer signaling [[159]53] and p53 linked GADD45 and HES1 activation,
   both signifying reduced cell survival and apoptosis [[160]54].
   Furthermore, following the second HBO application at 72 h, PER2
   activation in Circadian Rhythm signaling indicated ubiquitin-mediated
   proteolysis [[161]55] degrading oxidized proteins, suggesting
   HBO-induced protein damage. The enrichment of the herpes simplex virus
   1 infection pathway was created by a large suppression of different ZNF
   genes. ZNFs are discussed to be tumor suppressor genes and alterations
   are involved in inflammatory skin diseases and diabetes development
   [[162]56,[163]57]. As the beneficial effect of HBO on the intracellular
   survival and differentiation pathways was solely evident on a single
   stimulation, the subsequent multilineage experiments were conducted
   following a single short-termed HBO pre-stimulation.

   Regarding cellular properties, an inflammation-induced enhanced
   cellular proliferation was evident in the G-MSCs, similar to previous
   study results [[164]6,[165]31,[166]58,[167]59]. This effect was
   counteracted by the anti-inflammatory effect of the HBO treatment,
   which downregulated intracellular pathways associated with
   proliferation and cell survival (BMF, ID3, TXNIP, PDK4). Yet, no
   differences were notable in the pluripotency NANOG/OCT4A/SOX2 gene
   expression, Wnt-/β-Catenin signaling, or CFUs formation. The absence of
   Wnt-/β-Catenin signaling pathway activation, in contrast to earlier
   findings [[168]19,[169]21,[170]60], could rely on the short-termed HBO
   stimulation protocol employed in the current study. A single HBO
   pre-stimulation downregulated BMF expression, induced via TNF-α in a
   RIP1-dependent mechanism [[171]61,[172]62], necessary for
   death-receptor-induced apoptosis/necroptosis [[173]63], with improved
   G-MSCs’ survival and differentiation potential [[174]64]. The single
   HBO pre-stimulation further downregulated the inflammation-induced ID3
   in G-MSCs. Such a decrease in ID1-3 and increased BMPR1B expression
   would result in TGF-β signaling pathway activation and upregulation of
   pluripotency and differentiation potentials of G-MSCs
   [[175]51,[176]65], explaining the concurrently observed G-MSCs’
   stronger osteogenic and chondrogenic differentiation upon a single HBO
   pre-stimulation [[177]66].

5. Conclusions

   Taken together, current results point to a positive short-term
   stimulatory effect of a single HBO induction on G-MSCs’ regenerative
   and differentiation properties relying primarily on its ability to
   hinder/downregulate ROS inflammation-associated apoptotic pathways,
   while inducing regenerative intracellular signaling pathways. Yet, a
   second stimulation within a short time (72 h) would augment the
   inflammatory stress inside the cells, occurring during periodontitis,
   with accelerated signs of cellular damage. Thus, a single short-term
   HBO induction could be considered a beneficial priming method of G-MSCs
   for periodontal regenerative approaches. Still, further exploration of
   the HBO ideal stimulation duration and frequency in addition to
   clinical interventional studies is required to confirm the observed
   effects.

Acknowledgments