ABSTRACT
The increased local concentration of calcium ions (Ca2+) and phosphate (Pi), a natural body process for bone healing and remodeling, as well as local delivery of these ions as signaling molecules by synthetic bone graft substitutes, may lead to cytotoxic ion levels that can result in Ca2+/ Pi mitochondria overload, oxidative stress, and cell death. In this research, the effect of H2S as a cytoprotective signaling molecule to increase the tolerance of mesenchymal stem cells (MSCs) in the presence of cytotoxic level of Ca2+/Pi was evaluated. Different concentrations of sodium hydrogen sulfide (NaSH), a fast-releasing H2S donor, were exposed to cells in order to evaluate the influence of H2S on MSC proliferation. The results suggested that a range of NaSH (i.e., 0.25 - 4 mM NaSH) was non-cytotoxic and could improve cell proliferation and differentiation in the presence of cytotoxic levels of Ca2+ (32 mM) and/or Pi (16 mM). To controllably deliver H2S over time, a novel donor molecule in thioglutamic acid (GluSH) was synthesized and evaluated for its H2S release profile. Excitingly, GluSH successfully maintained cytoprotective level of H2S over 7 days. Furthermore, MSCs exposed to cytotoxic Ca2+/Pi concentrations in the presence of GluSH were able to thrive and differentiate into osteoblasts. These findings suggest that the incorporation of a sustained H2S donor such as GluSH into CaP-based bone substitutes can facilitate considerable cytoprotection making it an attractive option for complex bone regenerative engineering applications.
INTRODUCTION
Bone defects are the second leading cause of disability affecting more than 1.7 billion people worldwide.1 When bone fractures occur, free radicals and reactive oxygen species (ROS) are generated within the damaged tissue causing an imbalance between these and antioxidants damaging cellular macromolecules and altering their functions through oxidative stress.2 This oxidative stress causes Ca2+/Pi influx into the cytoplasm from the extracellular environment followed by Ca2+/Pi passage into the cell mitochondria. The trauma associated with fractures or other major bone injuries also damages the blood supply, resulting in local hypoxia.3 An increase in intracellular Ca2+ levels is a primary response of many cell types to hypoxia.4 Finally, after considerable bone tissue damage, Ca2+/Pi concentrations in the blood and urine will decrease5 while levels local to the defect site will increase causing the formation of a soft callus which is necessary for bone remodeling.6 These increases in extracellular Ca2+/Pi concentrations can cause an intracellular overload.7 All three of these phenomena contribute to Ca2+/Pi mitochondrial overload disrupting and accelerating cellular metabolism leading to cell death.3 In addition, mitochondrial Ca2+/Pi overload has been found to lead to greater ROS production4 which further increases oxidative stress leading to even greater intracellular Ca2+/Pi concentrations.
While bones are capable of remodeling themselves, reconstruction of critical-sized bone defects is challenging as intervention is required to facilitate adequate healing.8 Since the body’s natural reaction to bone tissue damage is to locally increase Ca2+/Pi concentrations, many regenerative engineering approaches leverage these ions as signaling molecules in their design. Bioactive ceramics, especially calcium phosphates (CaPs), have been widely used in bone regeneration because of their similarity to native bone mineral content.9, 10 Once immersed in an aqueous solution, CaPs undergo dissolution and precipitation as the result of ion transfer at the solid–liquid interface yielding a net release of calcium ions (Ca2+) and phosphate ions (H2PO4-, HPO32-, or PO43- - Pi) from the material.11 These ions have been found to facilitate bone regeneration.12, 13 Ca2+ and Pi osteoinductivity exists in a concentration range termed the therapeutic window where enough ions are present to facilitate stem cell osteogenic differentiation, but not too many to overwhelm the cells inducing their death. Our previous efforts have shown that Ca2+ concentrations of 32 mM or more and Pi concentrations of 16 mM or more are cytotoxic for mesenchymal stem cells in vitro14 and are the limits that can be used for biomaterials-based bone regeneration engineering applications15,16.
Improving cell tolerance to increased Ca2+/Pi concentrations can led to a broader and enhanced therapeutic window by limiting the cytotoxic effects of high concentration Ca2+/Pi. One promising option to mitigate ion-mediated toxicity is hydrogen sulfide (H2S), a gasotransmitter signaling molecule17 that has been found capable of suppressing oxidative stress in mitochondria18–20 as well as moderating cellular oxygen consumption especially under hypoxic conditions.21 Unfortunately, there are only a few H2S donors commercially available with most research efforts utilizing sodium hydrosulfide (NaHS) or sodium sulfide (Na2S).22 As both of these rapidly dissociate in water, they yield an uncontrollable burst release of H2S that can be toxic to cells. L-cysteine is another option where H2S is enzymatically liberated from the donor molecule. This process is regrettably not externally controllable as enzymatic downregulation is tied to increased local H2S levels.23 Some synthetic H2S donors have been developed including GYY4137 for more sustained and hydrolytic cleavage, but the toxicity and clearance of the remaining small molecule backbone after H2S release has yet to be studied.24, 25 Zhou and colleagues developed H2S donors through thioacid substitution in amino acids (i.e., glycine and valine) to improve the compatibility, predictability, and degradation kinetics of a H2S donor as a cardioprotective reagent.24 While promising, the rapid release rate and lack of a conjugatable chemical group makes thioglycine and thiovaline difficult to incorporate into various biomaterials for sustained, localized H2S release to supplement site-directed Ca2+/Pi osteoinductivity. To address this issue, we have designed and synthesized a novel H2S donor with a conjugatable carboxylic acid (i.e., thioglutamic acid - GluSH) and evaluated its cytoprotective effect on mesenchymal stem cells exposed to cytotoxic Ca2+/Pi concentrations in vitro.
EXPERIMENTAL
Preparation of H2S, Ca2+, and Pi Containing Media
H2S stock solution (pH = 7.4) was prepared by dissolving 512 mM of NaSH (Sigma-Aldrich) in distilled, deionized water (ddH2O) at 37 °C. Ca2+ stock solution (pH = 7.4) was prepared by dissolving 512 mM calcium chloride (CaCl2; Sigma-Aldrich), 25 mM 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES; Sigma-Aldrich), and 140 mM sodium chloride (NaCl; Sigma-Aldrich) in ddH2O at 37 °C. Pi stock solution (pH = 7.4) was prepared by dissolving disodium hydrogen phosphate dihydrate (Na2HPO4 • 2 H2O; Sigma-Aldrich) and sodium dihydrogen phosphate dihydrate (NaH2PO4 • 2 H2O; Sigma-Aldrich) at a ratio of 4:1 Na2HPO4/NaH2PO4, 25 mM HEPES, and 140 mM NaCl in ddH2O at 37 °C. These stock solutions were made fresh for each experiment and sterilized by 0.22 μm syringe filter. Individual solution concentrations of 0.25, 0.5, 1, 2, 4, 8, 16, 32, and 64 mM H2S, 32 mM Ca2+, and 16 mM Pi, as well as combined solution concentrations of 32 mM: 1 mM Ca2+:H2S, 16 mM: 1 mM Pi:H2S, and 32 mM: 16 mM: 1 mM H2S:Ca2+:Pi:H2S were prepared by diluting stock solutions in growth media consisting of Dulbecco’s modified Eagle’s medium (DMEM, Invitrogen) supplemented with 10% fetal bovine serum (FBS, Invitrogen) and 1% Penicillin-streptomycin (Pen-Strep, Invitrogen).
Cell Culture and Seeding
Murine mesenchymal stem cells (MSCs) were purchased from Cyagen and initially cultured in T-75 cell culture flasks (Corning) in growth medium at 37 °C in a humidified incubator supplemented with 5% CO2. Media was changed every 48 h until cells approached ~ 80% confluency after which they were dissociated using a 0.05% trypsin-EDTA (Invitrogen) solution. Detached MSCs were counted by hemocytometer and passed to new T-75 flasks at a splitting ratio of 1:4 or 1:5 dependent on cell count. After the 5th passage, cells were used for in vitro bioactivity studies. Tissue cultured polystyrene 24-well plates (Corning) were seeded with 30,000 cells/well and exposed to growth media alone as a negative control or growth media containing 0.25, 0.5, 1, 2, 4, 8, 16, 32, 64, 128, or 256 mM NaSH, or 32 mM Ca2+ (Ca32), or 16 mM Pi (P16). Additional studies utilized growth media containing 32 mM: 16 mM Ca2+:Pi (Ca32/P16), 32 mM glutamic acid (Glu32), 32 mM GluSH (GluSH32), 32 mM: 1 mM Ca2+:NaSH (Ca32/NaSH1), 16 mM: 1 mM Pi:NaSH (P16/NaSH1), 32 mM: 16 mM: 1 mM Ca2+:Pi:NaSH (Ca32/P16/NaSH1), 32 mM: 32 mM Ca2+:Glu (Ca32/Glu32), 16 mM: 32 mM Pi:Glu (P16/Glu32), or 16 mM: 32 mM: 32 mM Ca2+:Pi:Glu (Ca32/P16/Glu32), 32 mM: 32 mM Ca2+:GluSH (Ca32/GluSH32), 16 mM: 32 mM Pi:GluSH (P16/GluSH32), or 32 mM: 16 mM: 32 mM Ca2+:Pi:GluSH (Ca32/P16/GluSH32) loaded into a Transwell membrane insert (Corning). MSCs were cultured with these solutions for up to 7 days at 37 °C in a humidified incubator supplemented with 5% CO2 and the media containing various compounds were added fresh every 2 days. After day 7, cells exposed to high Ca2+ and Pi concentrations were treated with 16 mM Ca2+ (Ca16), 8 mM Pi (P8), or 16 mM: 8 mM Ca2+:Pi (Ca16/P8) based on their group without any additional H2S releasing molecules with the media containing appropriate ion concentrations changed every 2 days. Cell proliferation, viability, alkaline phosphatase (ALP) activity, and mineralization were assessed at 1, 3, 7, and 14 days.
Proliferation Assay
Cell proliferation was determined using the Quanti-iT PicoGreen dsDNA Assay (Thermo Fisher Scientific). At each endpoint, the samples were rinsed with phosphate buffered saline (PBS) and exposed to 1% Triton X-100 (Sigma-Aldrich) followed by three freeze-thaw cycles in order to lyse the cells. Lysates were diluted with TE buffer (200 mM Tris-HCL, 20 mM EDTA, pH 7.5) and mixed with PicoGreen reagent according to the manufacturer’s protocol. A BioTek Cytation 5 fluorospectrometer plate reader was utilized to measure the fluorescence of each sample (ex. 480 nm, em. 520 nm) and the cell number was calculated using a MSC standard curve (0 - 200,000 cells/mL).
Viability Assay
Cell viability was evaluated at each time point using an MTS Cell Proliferation Colorimetric Assay Kit (BioVision). MTS reagent (20 μL) was added to growth media (500 μL) followed by 4 h incubation at 37 °C in a humidified incubator supplemented with 5% CO2. Absorbance of each sample was measured at 490 nm using a plate reader. Cell viability was reported as the ratio of absorbance in the experimental groups compared to the growth media negative control.
Alkaline Phosphatase Activity Assay
Cell ALP activity was quantified at each time point using an Alkaline Phosphatase Assay Kit (BioVision). In brief, 20 μL of cell lysate was combined with 50 μL of p-nitrophenyl phosphate (pNPP) solution in assay buffer. The mixture was incubated for 1 h at room temperature away from light. The reaction was stopped by adding 20 μL of the stop solution and the absorbance of the solution was measured at 405 nm using a plate reader. To eliminate any background effects, 1% Triton X-100 was incubated with pNPP, exposed to stop solution in assay buffer after 1 h, and its absorbance deducted from sample absorbance. The absorbance was converted to content of dephosphorylated p-nitrophenyl (pNP) using a pNP standard curve (0 - 20 nmol/mL) which was dephosphorylated using excess ALP Enzyme. ALP activity was reported as the pNP content normalized by cell count.
Mineralization Assay
Cell-based mineral deposition was measured using an Alizarin red assay. At each time point, the media was removed after which the cells were washed with ddH2O and fixed in 70% ethanol for 24 h. The ethanol was removed and the samples were incubated in 1 mL of 40 mM Alizarin Red solution (Sigma-Aldrich) for 10 minutes. The samples were rinsed with ddH2O several times to make sure all non-absorbed stain was removed. Absorbed Alizarin Red was desorbed using 1 mL of a 10% cetylpyridinium chloride (CPC, Sigma-Aldrich) solution after which stain concentration was measured at 550 nm using a plate reader. Absorbance of each sample was converted to the concentration of absorbed Alizarin Red using a standard curve (0 - 0.2740 mg/mL). Samples above the standard curve linear range were diluted with CPC solution until a reading in the linear range was obtained. Cell-based mineral deposition was calculated by subtracting mineralization found in blank wells exposed to the same experimental conditions. All results were normalized by cell count.
GluSH Synthesis
The three-step synthesis process is detailed in Scheme 1.
GluSH synthesis process.
Reaction A
Sodium bicarbonate (NaHCO3; Sigma-Aldrich) (1.5 mmol) was dissolved in water (50 mL) at room temperature after which 10 mL of dry tetrahydrofuran (THF; Sigma-Aldrich) was added. L-glutamic acid 5-benzyl ester (Alpha Aesar) (4.2 mmol) was dissolved in the reaction mixture followed by dropwise addition of benzyl chloroformate (Sigma Aldrich) (6.5 mL). After 4 hours stirring under an argon atmosphere, the reaction mixture turned to a clear solution. To remove the unreacted benzyl chloroformate, the reaction mixture was washed with diethyl ether (60 mL x 2). The aqueous layer was then acidified (pH ~ 2) with 1 mM hydrochloric acid (HCl). The mixture was then extracted with ethyl acetate (60 mL x 2), dried over anhydrous sodium sulfate (Na2SO4), and evaporated under reduced pressure using a rotary evaporator (Buchi). The final product, N-benzyloxycarbonyl-L-glutamic acid 5-benzyl ester (1), was a white powder and used without further purification. This product was dissolved in deuterated chloroform for 1H-NMR spectroscopy analysis (Fig. S1): 1H NMR (500 MHz, CDCl3) δ 7.35-7.27 (m, 10H), 5.56-5.54 (m, 1H), 5.11 (m, 4H), 4.43-4.40 (m, 1H), 2.51-2.45 (m, 2H), 2.27-2.24 (m, 1H), 2.04 (m, 1H).
Reaction B
N-benzyloxycarbonyl-L-glutamic acid 5-benzyl ester ((1), 10 mmol) were dissolved in 15 mL of dry THF in a round bottom flask. The solution was then treated with NaSH (20 mmol) and thioacetic acid (10 mmol) at room temperature. After the reaction mixture was stirred in open air for 48 hours, the solvent was evaporated under reduced pressure using a rotary evaporator. The residue was diluted with water (50 mL) and acidified with 1 mM HCl (pH ~ 2). This aqueous solution was extracted with ethyl acetate (30 mL x 3). The combined organic layer was washed with brine, dried over anhydrous Na2SO4, and evaporated under reduced pressure using a rotary evaporator. The final product was purified using a silica gel column chromatography and the pure N-benzyloxycarbonyl-L-thioglutamic acid 5-benzyl ester (2) was eluted at 40% ethyl acetate in hexane as a yellow oil. This product was dissolved in deuterated chloroform for 1H-NMR spectroscopy analysis (Fig. S2): 1H NMR (500 MHz, CDCl3) δ 7.35-7.27 (m, 10H), 5.53-5.51 (m, 1H), 5.12 (m, 4H), 4.46-4.45 (m, 1H), 2.56-2.48 (m, 2H), 2.31-2.29 (m, 1H), 2.07 (m, 1H).
Reaction C
N-benzyloxycarbonyl-L-thioglutamic acid 5-benzyl ester ((2), 10 mmol) was dissolved in methanol (40 mL) after which 10 wt % palladium on carbon (Pd/C; Sigma-Aldrich) was added to the mixture as a catalyst for the hydrogenation reaction. The reaction flask was sealed and vacuum purged prior to the addition of hydrogen gas. The reaction was stirred for 48 hours under hydrogen at room temperature. The reaction mixture was then dissolved in water and vacuum filtered. The clear filtrate was then evaporated under reduced pressure using a rotary evaporator which left a yellow powder of L-thioglutamic acid (GluSH, (3)). This product was dissolved in deuterium oxide for 1H-NMR spectroscopy analysis (Fig. S3): 1H NMR (500 MHz, D2O) δ 3.96-3.79 (m, 1H), 2.57-2.56 (m, 2H), 2.16 (m, 2H).
H2S Release Study
H2S release from GluSH was measured in ddH2O at 37 °C over 7 days. GluSH loaded into a Transwell membrane insert (Corning) was placed in centrifuge tubes containing Hank’s buffer saline solution (HBSS) supplemented with excess amounts of bicarbonate (1.5 mol per 1 mol GluSH). Each centrifuge tube was sealed with vacuum glue and parafilm. At certain time points, 1 mL of the release solution were extracted from the reaction tube using a needle and replaced with 1 mL of HBSS and bicarbonate solution. The solutions then immediately assayed to determine their H2S concentration using a fluorescent method measuring the formation of thiobimane.26 Briefly, dibromobimane (Sigma-Aldrich) was dissolved in HBSS at a concentration of 500 μM and incubated with the test samples at room temperature for 5 minutes after which florescence (ex. 340 nm, em. 465 nm) was measured. A standard curve was created by dissolving different concentrations of NaSH in the 500 μM dibromobimane in HBSS solution, incubating at room temperature for 5 minutes, and measuring the florescence. The concentration of the H2S in each time point were then calculated by comparing the test sample values to the NaSH standard curve.
Statistical Analysis
JMP software was used to make comparisons between groups with Tukey’s HSD test specifically utilized to determine pairwise statistical differences (p < 0.05). The statistical analysis results are reported in the supporting information section. Groups that possess different letters have statistically significant differences in mean whereas those that possess the same letter have means that are statistically insignificant in their differences.
RESULTS AND DISCUSSION
H2S Cytotoxicity
Proliferation and viability of MSCs exposed to different concentrations of NaSH are shown in Fig. 1. MSCs seeded on tissue cultured plastic (Ctrl) and incubated in growth media increased to 7 times initial cell seeding number over 14 days and cells exposed to concentrations of 4 mM NaSH or less expanded to more than 8 times initial cell seeding number (Fig. 1A). The cell number for those exposed to 0.5 and 1 mM of NaSH were statistically significantly higher than control group at day 14 (Table S1). On the other hand, proliferation of MSCs exposed to 8 mM NaSH or more were statistically significantly lower than those provided growth medium supplemented with 0 - 4 mM NaSH (Table S1). The mildly mitogenic behavior of H2S at lower concentrations is likely due to its ability to increase intracellular levels of cyclic guanosine monophosphate (cGMP),27 which is known to stimulate stem cell proliferation.28 H2S can also inhibit cytochrome c oxidase activity and cause oxidative stress29 or even directly cause a radical-associated DNA damages,30 which is likely responsible for the cell death found at higher H2S concentrations.
Proliferation and viability of MSCs exposed to different concentrations of NaSH supplemented media. (a) Cell count was measured by the Quanti-iT PicoGreen Assay over 14 days of cell culture in media supplemented with 0 - 256 mM NaSH for 7 days and then culture media with 0 mM NaSH for the next 7 days. The black line indicates the original cell seeding number (i.e., 30,000). (b) An indirect measure of cell metabolism and viability was determined by NAD(P)H activity over 14 days using a MTS assay for MSCs exposed to 0 - 256 mM NaSH for 7 days and then culture media with 0 mM NaSH for the next 7 days. Statistical analysis of the data is available in the supplementary information (Table S1 and Table S2).
The results of the MTS assay show that MSCs exposed to 4 mM NaSH or less were over 90% as viable as the control group through 14 days. Interestingly, MSCs subj ected to 8 mM NaSH or more were still more than 80% viable as the control group through 14 days (Fig. 1B). Taken with the proliferation results, these data indicate that though high concentrations of NaSH adversely affect MSC proliferation, the surviving cells are highly metabolically active over the 14 days of the study.
Cytoprotective Effect of H2S at High Ca2+ and/or Pi Concentrations
The cytotoxicity evaluation of NaSH revealed that high NaSH concentration (i.e., 8 mM or more) can decrease cell proliferation. Since there was no statistically significantly difference between cell proliferation and viability of the MSCs exposed to 0.25 to 4 mM NaSH, the rest of this research explored the cytoprotective effect of 1 mM NaSH (NaSH1) when co-delivered with higher concentrations of Ca2+ and/or Pi. Our previous research indicates that media supplemented with more than 16 mM Ca2+ and/or 8 mM Pi can be cytotoxic adversely affecting MSC proliferation and viability.14 Therefore, 32 mM Ca2+ (C32 - cytotoxic concentration), 16 mM Pi (P16 - cytotoxic concentration), 1 mM NaSH (NaSH1 - non-cytotoxic concentration), and combinations of Ca2+, Pi, and NaSH1 were used to explore the cytoprotective effect of hydrogen sulfide in the initial presence of excess calcium and phosphate ions. After 7 days of high concentration exposure, MSCs were exposed to inductive and non-cytotoxic Ca2+ (i.e., Ca16) and Pi (i.e., P8) concentrations without NaSH from day 7 to day 14 to better mimic the conditions within the bone fracture site.
Proliferation and viability of cells exposed to different combinations of ionic and gasotransmitter signaling molecules is demonstrated in Fig. 2. MSCs cultured with Ca32 and P16 showed less proliferation over 14 days as compared to those grown under control conditions, which is due to the likely cytotoxic effects and the possible osteoinductivity of high Ca2+ and Pi concentrations (Fig. 2A). However, when the MSCs were also supplied with NaSH1, their proliferation was statistically significantly greater through 14 days (Table S3). This result demonstrate the positive effect of NaSH on the proliferation of cells exposed to cytotoxic level of Ca2+ and Pi. It is known that increases in intracellular Ca2+ promote mitochondrial calcium uptake resulting in the loss of mitochondrial membrane potential which is interpreted as a danger signal initiating apoptosis.31
Proliferation and viability of MSCs exposed to different combinations of Ca2+, Pi, and/or NaSH supplemented media. (a) Cell count was measured by the Quanti-iT PicoGreen Assay over 14 days of cell culture in media supplemented with no signaling molecules (Ctrl), 32 mM Ca2+ (Ca32), 16 mM Pi (P16), Ca32/P16, 1 mM NaSH (NaSH1), Ca32/NaSH1, P16/NaSH1, or Ca32/P16/NaSH1 for the first 7 days. This was followed by the cells being exposed for the next 7 days to no signaling molecules or half the Ca2+ and/or Pi concentrations (i.e., Ca16 and/or P8) they were originally cultured in. The black line indicates the original cell seeding number (i.e., 30,000). (b) An indirect measure of cell metabolism and viability was determined by NAD(P)H activity over 14 days using an MTS assay. MSCs were exposed to Ca32, P16, and/or NaSH1 and combination of these molecules (Ca32/P16, Ca32/NaSH1, P16/NaSH1, and Ca32/P16/NaSH1) for the first 7 days followed by the aforementioned decrease in Ca2+ and/or Pi concentrations for the respective groups. All data were normalized against the results determined for MSCs given non-supplemented media. Statistical analysis of the data is available in the supplementary information (Table S3 and Table S4).
Addition of NaSH1 as a H2S donor can help regulate cytosolic Ca2+ levels and mitochondrial calcium uptake,32 resulting in higher cell survival compared to MSCs exposed to a high Ca2+ concentration alone. Higher Pi concentrations can dysregulate the mitochondrial permeability transition pore (MPTP) initiating a pro-apoptotic cascade33 or further enhancing ongoing apoptosis.34 By reducing oxidative stress, H2S indirectly protects the mitochondria from becoming damaged and activating pro-apoptotic signaling pathways.35
Cell viability results reveal that the metabolic activity of MSCs exposed to the Ca32, P16, and Ca32/P16 were statistically significantly lower than Ca32/NaSH1 and P16/NaSH1, and Ca32/P16/NaSH1 through 14 days (Fig. 2B and Table S3). High Ca2+ and/or Pi concentrations can even lead to viability less than 20% when compared to the control group whereas supplementing these same groups with NaSH1 preserved cell viability to over 90% of control. To investigate viability, the MTS assay was used which measures cell viability by indirectly determining NAD(P)H-dependent mitochondrial dehydrogenase activity essential to cell metabolism and proliferation oxidation/reduction reactions.36 The previously mentioned mitochondrial dysregulation due to Ca32 and P16 that limited cell proliferation understandably negatively impacted cell dehydrogenase activity as well.
ALP activity and cell-based mineralization of MSCs cultured with Ca32 or P16 with or without NaSH1 is described in Fig. 3. MSC osteogenic differentiation is known to proceed through 3 stages: proliferation, maturation, and mineralization. ALP is an enzyme that is expressed during the beginning of the osteogenic maturation period.37 The ALP results presented were normalized by cell number to focus on cell differentiation independent of proliferation. MSCs exposed to media (Ctrl) and media supplemented with NaSH1 showed background levels of ALP expression while cells subjected to Ca32, P16, Ca32/P16, Ca32/NaSH1, P16/NaSH1, and Ca32/P16 /NasSH1 showed statistically significant increased levels of ALP expression compared to control (Fig. 3A, Table S3). Also, cells cultured with Ca32 and P16 supplemented with NaSH1 (i.e., Ca32/NaSH1, P16/NaSH1, and Ca32/P16 /NasSH1) showed statistically significantly higher levels of ALP expression compared to those without NaSH1 (i.e., Ca32, P16, and Ca32/P16) at each time point assessed (Fig. 3A, Table S3 and S4). The low ALP activity in MSCs exposed to Ca32 and/or P16 alone is likely due to their diminished viability which when modulated by NaSH1 can improve overall ALP activity considerably. It is recently been reported that H2S can also promote osteoblast differentiation at sites of bone regeneration by triggering deposition of inorganic mineral matrix and promoting expression of osteogenic genes in human MSCs.38 However, our results did not show improved ALP activity after exposure of MSCs to NaSH1 alone compared to control group over the course of 14 days suggesting H2S can assist other inductive molecules (i.e., Ca2+ and Pi) though may not be osteoinductive in its own right.
ALP activity and cell-based mineralization of MSCs exposed to different combinations of Ca2+, Pi, and/or NaSH supplemented media. (a) Maturation enzyme activity was analyzed by an ALP Assay Kit over 14 days of cell culture in media supplemented with no signaling molecules (Ctrl), 32 mM Ca2+ (Ca32), 16 mM Pi (P16), Ca32/P16, 1 mM NaSH (NaSH1), Ca32/NaSH1, P16/NaSH1, or Ca32/P16/NaSH1 for the first 7 days. This was followed by the cells being exposed for the next 7 days to no signaling molecules or half the Ca2+ and/or Pi concentrations (i.e., Ca16 and/or P8) they were originally cultured in. (b) Alizarin red (ALZ) staining was used as an indirect measure of mineralization over 14 days of cell culture. MSCs were exposed to no signaling molecules (Ctrl), Ca32, P16, and/or NaSH1, and combination of these molecules (Ca32/P16, Ca32/NaSH1, P16/NaSH1, Ca32/P16/NaSH1) for the first 7 days followed by the aforementioned decrease in Ca2+ and/or Pi concentrations for the respective groups. Both ALP activity and ALZ mineralization data were normalized on a per cell basis. Statistical analysis of the data is available in the supplementary information (Table S3 and Table S4).
In comparison the early-stage activity of ALP, matrix mineralization is a late-stage MSC osteogenic marker. In this research, acellular experiments were conducted in parallel using the same conditions allowing their results to be subtracted from the complementary cellular groups to eliminate solution-mediated mineralization. These cell-based mineralization results were then normalized by cell number to evaluate improved osteoinductivity by NaSH1 independent of its influence on proliferation. Evaluation of MSC mineralization revealed that Ca32/NaSH1, P16/NaSH1, and Ca32/P16/NaSH1 induced statistically significantly greater mineral deposition over time (Table S4) as well as greater than cells exposed to Ctrl and NaSH1 treatments (Fig. 3B and Table S3). Ca32, P16, and Ca32/P16 were found to limit mineralization instead of enhancing it over the Ctrl group over time (Fig. 3B and Table S3 and S4) likely due to the toxicity of these treatments. The mineralization results were in agreement with the ALP activity data supporting the beneficial effects of H2S when supplementing high concentrations of osteoinductive ions.
H2S Release from GluSH
In order to investigate the cytoprotective effect of sustained H2S delivery, GluSH was synthesized and tested for its ability to controllably release H2S over time. GluSH is a highly water soluble product that releases a sulfanyl ion (HS-) in the presence of bicarbonate by the chemical reaction outlined in Fig. 4 which can then be protonated in water to yield H2S. HS- is only produced from GluSH in the presence of bicarbonate, which is a molecule that can be readily found in the human body.39 Therefore, bicarbonate-supplemented pH and osmotic balancing HBSS was used to investigate H2S release from GluSH over time. GluSH quantity used in the experiment was varied to determine what concentration would allow the maximum H2S concentration release to not exceed the previously NaHS-determined cytotoxic level (i.e., 4 mM). Fig. 5 summarizes the results of H2S production from 32 mM GluSH over 7 days. As shown in Fig. 5A, GluSH released around 50% of its total H2S generating payload within the first day. H2S continued to be generated through day 5 after which little more was alleviated from GluSH. When the data was converted to H2S concentration, GluSH was found to controllably release its payload within the non-cytotoxic range (i.e., 0.25 - 4 mM) for 7 days (Fig. 5B).
Sulfanyl ion (HS-) release mechanism from GluSH in the presence of bicarbonate. The binding of the carboxyl group to the amine creates an unstable complex, which results in cyclizing of the α-amino acid groups into an N-carboyxanhydride releasing HS-. Both reaction steps are reversible and the progression towards HS- release only occurs in the presence of a sufficient bicarbonate concentration.
GluSH hydrogen sulfide release. (a) The cumulative release of H2S alleviated from 32 mM GluSH immersed in bicarbonate-supplemented Hank’s Balanced Salt Solution (HBSS) was measured over 7 days at 37 °C. (b) The solution H2S concentration was determined from this release data.
GluSH Mediated Stem Cell Proliferation and Differentiation
To evaluate GluSH bioactivity, MSCs were exposed to GluSH32 supplemented with toxic level of Ca2+ and Pi. To distinguish the difference between the effect of H2S release from the rest of GluSH molecule, glutamic acid (Glu) supplemented media was also investigated as a control. The proliferation and viability of cells exposed to media supplemented with excess Ca2+ and Pi with or without Glu or GluSH is summarized in Fig. 6. Proliferation of MSCs exposed to Ca2+ and/or Pi in the presence of GluSH (i.e., Ca32/GluSH32, P16/GluSH32, and Ca32/P16/GluSH32) were statistically significantly greater than comparative ones supplemented with Ca2+ and/or Pi alone or with Glu (i.e., Ca32, P16, Ca32/P16, Ca32/Glu32, P16/Glu32, and Ca32/P16/Glu32) and lower than media-only control and GluSH32 (Fig. 6A and Table S5). These results are due to the synergistic effect of H2S moderating high concentration Ca2+ and Pi cytotoxicity while allowing these molecules to likely carry out their osteoinductive effects. Interestingly, cells exposed to GluSH32 alone proliferated to 8 times of their initial seeding concentration indicating the biocompatibility of our novel molecule. However, MSCs exposed to Glu32 alone did not proliferate over the course of 14 days. This results suggest that the acidic nature of glutamic acid could potentially cause some cytotoxicity.40 In contrast, GluSH releasing HS- produces the byproduct glutamate N-carboxyanhydride (Fig. 5), which generates a less acidic solution that glutamic acid while the presence of H2S would also serve as a cytoprotectant. The viability of the MSCs exposed to different combinations of signaling molecules reveals that the cells exposed to Ca32 and P16 as well as Glu were less than 30% metabolically active as those in the control group by day 14 (Fig. 6B and Table S5). Supplementing high Ca2+ and/or Pi concentration media with GluSH mitigated any cytotoxic effects increasing cell viability to more than 100% as compared to the control group.
Proliferation and viability of MSCs exposed to different combinations of Ca2+, Pi, and/or GluSH supplemented media. (a) Cell count was measured by the Quanti-iT PicoGreen Assay over 14 days of cell culture in media supplemented with no signaling molecules (Ctrl), 32 mM Ca2+ (Ca32), 16 mM Pi (P16), 32 mM glutamic acid (Glu32), 32 mM thioglutamic acid (GluSH32), or combination of these molecules (i.e., Ca32/P16, Ca32/Glu32, P16/Glu32, Ca32/P16/Glu32, Ca32/GluSH32, P16/GluSH32, Ca32/P16/GluSH32) for the first 7 days. This was followed by the cells being exposed for the next 7 days to no signaling molecules or half the Ca2+ and/or Pi concentrations (i.e., Ca16 and/or P8) they were originally cultured in. The black line indicates the original cell seeding number (i.e., 30,000). (b) An indirect measure of cell metabolism and viability was determined by NAD(P)H activity over 14 days using an MTS assay. MSCs were exposed to Ca32 and/or P16 with or without Glu32 or GluSH32 for the first 7 days followed by the aforementioned decrease in Ca2+ and/or Pi concentrations for the respective groups. All data were normalized against the results determined for MSCs given non-supplemented media. Statistical analysis of the data is available in the supplementary information (Table S5 and Table S6).
The ALP activity and cell-based mineralization of MSCs cultured with Ca32 and/or P16 with or without GluSH32 or Glu32 is described in Fig. 7. MSCs exposed to non-supplemented media (Ctrl) or media supplemented with Glu32 or GluSH32 showed background levels of ALP expression while cells subjected to Ca32, P16, Ca32/P16, Ca32/GluSH32, P16/GluSH32, and Ca32/P16/GluSH32 showed statistically significant increased levels of ALP expression compared to control (Fig. 7A, Table S5). Cells cultured with Ca32 and P16 with GluSH32 (i.e., Ca32/GluSH32, P16/GluSH32, and Ca32/P16/GluSH32) showed statistically significantly higher levels of ALP expression compared to those without GluSH32 (i.e., Ca32, P16, and Ca32/P16) that also increased over time (Fig. 7A, Table S5, and Table S6). Assessment of MSC mineralization revealed that Ca32/GluSH32, P16/GluSH32, and Ca32/P16/GluSH32 induced statistically significant mineral deposition over time (Table S6) that was greater than all other treatments from Day 3 on (Fig. 7B and Table S5). MSCs exposed to Ca32, P16, or Ca32/P16 showed limited changes over time (Table S6) which while possessing greater mineralization than Ctrl was statistically insignificant different than cells exposed to Glu32 (Fig. 7B and Table S5).
ALP activity and cell based mineralization of MSCs exposed to different combinations of Ca2+, Pi, and/or GluSH supplemented media. (a) Maturation enzyme activity was analyzed by an ALP Assay Kit over 14 days of cell culture in media supplemented with no signaling molecules (Ctrl), 32 mM Ca2+ (Ca32), 16 mM Pi (P16), 32 mM glutamic acid (Glu32), 32 mM thioglutamic acid (GluSH32), or combination of these molecules (i.e., Ca32/P16, Ca32/Glu32, P16/Glu32, Ca32/P16/Glu32, Ca32/GluSH32, P16/GluSH32, and Ca32/P16/GluSH32) for the first 7 days. This was followed by the cells being exposed for the next 7 days to no signaling molecules or half the Ca2+ and/or Pi concentrations (i.e., Ca16 and/or P8) they were originally cultured in. (b) Alizarin red (ALZ) staining was used as an indirect measure of mineralization over 14 days of cell culture MSCs were exposed to Ca32 and/or P16 with or without Glu32 or GluSH32 for the first 7 days followed by the aforementioned decrease in Ca2+ and/or Pi concentrations for the respective groups. Both ALP activity and ALZ mineralization data were normalized on a per cell basis. Statistical analysis of the data is available in the supplementary information (Table S5 and Table S6).
CONCLUSIONS
This research aimed to evaluate the cytoprotective effect of H2S on MSCs experiencing ion overload similar to what can occur in a bone defect site microenvironment. It was determined that an effective therapeutic range exists for H2S that can improve cell proliferation and differentiation even in the presence of cytotoxic levels of calcium and phosphate ions. Furthermore, a novel compound was generated through glutamic acid modification that was capable of sustained H2S release in the therapeutic window for 7 days, which proved effective in mitigating ion-mediated cytotoxicity. These results support the considerable promise of thioglutamic acid as a useful product in helping to facilitate cell survival in critical sized bone defects.
CONFLICT OF INTERESTS
There are no conflicts of interest to report.
ACKNOWLEDGMENTS
The authors gratefully acknowledge support from start-up funds as well as a College of Engineering Incentive Fund Grant and a University of Missouri Research Council Grant, all kindly provided by the University of Missouri.
REFERENCES
