2.1.

Culture and Osteogenic Differentiation of Human Mesenchymal Stem Cells

hMSCs isolated from the bone marrow aspirates of healthy donor were cultured in MesenPro medium consisting of 2% fetal bovine serum.²⁷ (Gibco BRL, Invitrogen, Grand Island, New York.) Osteogenic differentiation of hMSCs was performed as previously reported.^9,27 The procedure to obtain hMSCs is approved by Institutional Review Board of Taipei Veteran General Hospital and informed consent of the donors was obtained as previously reported.²² In briefly, hMSCs were first seeded at a density of $4000\mathrm{cells}/{\mathrm{cm}}^2$. Next, osteogenic induction was carried out at $\sim 50\%$ confluency of cultured hMSCs in osteogenic induction medium [serum-free IMDM, $0.1\mu \mathrm{M}$ dexamethasone, 0.2 mM ascorbic acid, 10 mM $\beta$-glycerol phosphate (Sigma-Aldrich, St. Louis, Missouri), and $100\text{units}/\mathrm{ml}$ penicillin and $100\mathrm{mg}/\mathrm{ml}$ streptomycin]. The induction medium was changed twice a week until the cells were harvested. In part of the experiments, cells were treated with $1\mu \mathrm{g}/\mathrm{ml}$ OA in the induction medium at the beginning of osteogenic induction to attenuate/inhibit oxidative metabolic activity. We carefully chose the dose of OA based on our previous study⁹ so that the cell survival rate remained the same as that of the controls but ATP decreased during differentiation.

2.2.

NADH Fluorescence Lifetime Image

NADH fluorescence lifetime images of hMSCs were obtained as previously reported.^22,28,29 In the beginning, hMSCs were seeded onto a 24-mm diameter glass coverslip (Paul Marienfeld GmbH & Co., Lauda- Konigshofen, Germany) and osteogenesis was induced as previously described. Prior to fluorescence lifetime imaging, the coverslip was washed twice using PBS and then transferred into an imaging chamber and 1-ml aliquot of 5 mM 4-(2- hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer (5 mM KCl, 140 mM NaCl, 2 mM ${\mathrm{CaCl}}_2$, 1 mM ${\mathrm{MgCl}}_2$, 10 mM glucose, pH 7.4) was added to maintain physiological state and avoid light absorption of the culture medium. hMSCs were imaged with a two-photon laser scanning microscope and with a $60\times /\mathrm{NA}$ 1.45 PlanApochromat oil objective lens (Olympus Corp., Tokyo, Japan). NADH fluorescence was excited at 740 nm by a Verdi pumped modelocked femtosecond Ti:sapphire laser (Coherent, Inc., Santa Clara, California) at 76 MHz and the emitted fluorescent light was detected at $447\pm 30\mathrm{nm}$, with the NADH fluorescence peaks detected by a bandpass filter (Semrock Inc., Rochester, New York). Fluorescence photons were detected by a photon-counting photomultiplier H7422P-40 (Hamamatsu Photonics K.K., Hamamatsu, Japan). All the images were taken at $256\times 256\text{pixels}$ resolution with an acquisition time of 900 s. Cells were incubated at 37°C and 5% ${\mathrm{CO}}_2$ during the image acquisition.

Time-resolved detection was conducted by the time-correlated single-photon counting SPC-830 board (Becker & Hickl GmbH, Berlin, Germany). Data were analyzed with the commercially available SPC Image v2.8 software (Becker & Hickl GmbH, Berlin, Germany) via a convolution of a two component exponential decay function, $F(t)=a_1{e}^{-t/\tau 1}+a_2{e}^{-t/\tau 2}$ and the instrument response function (IRF), and then were fitted to the actual data to extract lifetime parameters ${\tau }_1,{\tau }_2,a_1,a_2$, and ${\tau }_m$. ${\tau }_m$ is the mean lifetime defined as $(a_1{\tau }_1+a_2{\tau }_2)/(a_1+a_2)$. IRF was measured using a second-harmonic–generated signal from a periodically poled lithium niobate crystal as a standard procedure. The full width at half maximum of our IRF is $\sim 300\mathrm{ps}$, which is similar to other published results.^19,30 In addition, the value of NADH lifetime was calculated over the entire image. The field-of-view of each image is $100\times 100\mathrm{\mu m}$, which covered about one to two stem cells. The sample size per time point is five to eight independent samples, meaning five to eight cell culture dishes. We took two to three images from different sites per dish.

2.3.

Alkaline Phosphatase Staining and Activity Assay

For alkaline phosphatase (ALP) staining, cells grown in 35 mm dishes were fixed with 4% paraformaldehyde and then permeabilized with 0.1% Triton-X 100. Intracellular ALP was revealed by staining 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (Sigma-Aldrich). The assay of ALP activity used is similar to that previously reported.⁹ Cells grown in 48-well plates were disintegrated with 0.05% sodium dodecyl sulfate (SDS) and then incubated in assay solution containing 8 mM 4-nitrophenyl phosphate and 2 mM ${\mathrm{MgCl}}_2$ in 2-amino-2-1-propanol for 1 h. The reaction was inhibited by the addition of 0.02 N NaOH and the release of 4-nitrophenol was measured by an ELISA reader (Thermo Scientific, Waltham, Massachusetts) at 405 nm. Cell content was measured by a standard called alamarBlue assay. The results are presented as the normalized absorbance at 405 nm to cell content.

2.4.

Measurement of Oxygen Consumption Rate

The oxygen consumption rate (OCR) was measured by 782 Oxygen Meter (MitocellMT200A, Strathelvin Instrument, Motherwell, United Kingdom) as previously reported.^29,31 An aliquot of $4\times {10}^5$ cells was re-suspended in $330\mu \mathrm{l}$ of assay buffer (125 mM sucrose, 65 mM KCl, 2 mM ${\mathrm{MgCl}}_2$, 20 mM phosphate buffer, pH 7.2), and the cell suspension was transferred into the respiration chamber at 37°C with a water circulation system. An aliquot of 0.0004% digitonin was added to permeablilize the outer membrane of mitochondria. The substrate-supported OCR was measured after 17 mM glutamate and 17 mM malate were injected into the chamber for the requirement of the oxygen molecule. After recording the substrate-supported OCR, 1.5 mM adenosine diphosphate (ADP) was injected to measure the ADP-stimulated OCR, which assesses the efficiency of ATP synthesis and the integrity of mitochondrial respiratory chain.²⁷

2.5.

Measurement of Intracellular Adenosine Triphosphate Level

The intercellular ATP level was measured by the bioluminescent somatic cell ATP assay kit (Sigma-Aldrich). Cells were collected at different time points and an aliquot of $50\mu \mathrm{l}$ of viable cell suspension was mixed with $150\mu \mathrm{l}$ of somatic cell releasing buffer to release the intercellular ATP. Half of the mixture (i.e., $100\mu \mathrm{l}$) was then transferred into a black OptiPlate-96F 96-well plate containing $100\mu \mathrm{l}$) of ATP assay mix in each well. The luminescence intensity was measured using the Victor2 1420 Multilabel Counter (PerkinElmer, Waltham, Massachusetts). The reported luminescence intensity was normalized to the cell number.

2.6.

Measurement of Lactate Release Rate

The lactate production rate was measured by a Lactate Reagent kit (Trinity Biotech PLC, Bray, Ireland). Cells in 6-well plates were replenished with fresh medium and incubated for 8 h. An aliquot of $10\mu \mathrm{l}$ of medium was then transferred to a 96-well plate to mix with the Lactate Reagent, and absorbance at 540 nm was measured by an ELISA reader (Thermo Scientific). The lactate release rate was calculated as the absorbance normalized to the total cell number and then divided by the incubation time (h).

2.7.

Western Blot Analysis

An aliquot of 25 to $30\mu \mathrm{g}$ of proteins was separated on a 10% SDS polyacrylamide electrophoresis gel and transferred onto a polyvinylidene difluoride membrane (Amersham Biosciences). Nonspecific bindings were blocked by 5% skim milk in Tris-buffered saline Tween 20 buffer (50 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20, pH 7.4), and the membrane was blotted indicating the anti-Complex I-V subunit primary antibodies, which were all purchased from Molecular Probe (Invitrogen). After incubation with a horseradish peroxidase-conjugated secondary antibody, the protein intensity was determined by an enhanced chemilluminescence reagent (PerkinElmer Life and Analytical Sciences, Boston).

3.1.

Osteogenic Differentiation of hMSCs Without and With Oligomycin A Treatment

We measured the ALP activity, a widely used differentiation biomarker during osteogenesis, in normal osteogenic differentiation as controls and with $1\mu \mathrm{g}/\mathrm{ml}$ OA treatment to determine their osteogenic differentiation. After osteogenic induction, the morphology of hMSCs gradually became flattened and polygonal, and there was an increase in ALP activity staining [Fig. 1(A)]. Under OA treatment, we found that OA-treated hMSCs had weaker ALP staining at the same time point as compared with untreated controls [Fig. 1(A)]. We further measured ALP activity by spectrophotometric assay and found that the ALP activity under OA treatment increased slowly to show it started differentiation at week 2 with less activity as compared to normal osteogenic induction which started differentiation at week 1 [Fig. 1(B)]. These results confirmed that OA treatment inhibited osteogenic differentiation.

Fig. 1

(A) Alkaline phosphatase (ALP) stain of undifferentiated hMSCs (a) and differentiated hMSCs at 3 weeks without (ost3w_N) (b) or with OA treatment (ost3w_OA) (c). Note that the cell morphology became flattened and polygonal shapes after osteogenic induction. Scale bar is 100 μm. (B) ALP activity of hMSCs during osteogenic differentiation at 4 days (ost4d), 1 week (ost1w), 2 weeks (ost2w), 3 weeks (ost3w), and 4 weeks (ost4w). Black bars represent normal osteogenic induction and white bars represent osteogenic induction with OA treatment. Each data point was calculated based on data of 4 samples. Symbols * indicate that the value is significantly different ($p$-value $<0.05$, Student’s $t$-test) as compared with undifferentiated hMSCs (hMSC).

3.2.

NADH Fluorescence Lifetime of hMSCs Increased During Normal Osteogenic Induction, but did not Increase With OA Treatment

We have reported that NADH lifetime increased gradually during osteogenic differentiation for up to 3 weeks.²² To confirm this result, we repeated the experiment using hMSCs from a different donor. Figure 2(A) shows NADH fluorescence lifetime images of undifferentiated and differentiated hMSCs at 4 days and 1 to 4 weeks. Each pixel represents the mean lifetime (${\tau }_m$) and was color-coded between 500 (red) and 2000 (blue) ps. Similar to our previous study, the NADH fluorescence lifetime of hMSCs exhibited a shift from shorter (yellow) to longer lifetime (green to blue) during osteogenic differentiation [Fig. 2(Aa–Af)]. Figure 2(Ag–Ak) shows NADH fluorescence lifetime image of differentiated hMSCs from 4 days to 4 weeks under OA treatment. No significant lifetime change was observed. We acquired at least five different images of mean lifetime from five different cultural dishes at each time point and plotted the average result in Fig. 2(B). The NADH fluorescence lifetime of normal hMSCs increased gradually from $975.6\pm 33.7\mathrm{ps}$ to reach a peak value of $1179.7\pm 104.9\mathrm{ps}$ at week 3. The NADH fluorescence lifetime significantly increased ($p<0.05$) at week 1. The NADH fluorescence lifetime of differentiated hMSCs with OA treatment did not show any difference from that of undifferentiated hMSCs.

Fig. 2

(A) Two-photon fluorescence lifetime images of undifferentiated (hMSC) (a) and osteogenic differentiated hMSCs at 4 days (ost4d) (b), 1 week (ost1w) (c), 2 weeks (ost2w) (d), 3 weeks (ost3w) (e) and 4 weeks (ost4w) (f) after osteogenic induction. Osteogenic differentiated hMSCs with oligomycine A (OA) are shown from (g) to (k) at the same time points as those without OA. (B) The averaged results of NADH fluorescence lifetime were calculated from multiple images taken from different samples. Each data point was calculated based on data of at least five independent experiments. Symbols * indicate that the values are significantly different ($p$-value $<0.05$, Student’s $t$-test) as compared with undifferentiated hMSCs. Black bars represent normal osteogenic induction and white bars represent osteogenic induction with OA treatment.

3.3.

ATP Level and Oxygen Consumption Increased and Lactate Release Rate Decreased During Normal Osteogenic Differentiation of hMSCs

Because osteogenic differentiation of hMSCs is associated with metabolic change,^9,27 our observed NADH fluorescence lifetime increase is likely linked to this metabolic change during osteogenic differentiation. We measured cellular ATP level, oxygen consumption, and lactate release rate with or without OA treatment. Figure 3(a) shows that the cellular ATP level of hMSCs increased from $95.1\pm 18.1\mathrm{nmol}\text{e}/{10}^4$ cells gradually to a peak value of $147.1\pm 19.5\mathrm{nmol}\text{e}/{10}^4$ cells after differentiation for 3 weeks. The ATP level of differentiated hMSCs under OA treatment did not change, and even decreased to $65.1\pm 12.7\mathrm{nmol}\text{e}/{10}^4$ cells at 4 weeks after differentiation. The increasing trend in oxygen consumption of normal-differentiated hMSCs is similar to that of the ATP level [Fig. 3(b)]. The highest value of oxygen consumption is $2.40\pm 0.29\mathrm{nmol}\text{e}/\min /{10}^6$ cells at 3 weeks after differentiation, which is $\sim 2\text{-}\mathrm{fold}$ increase as compared with undifferentiated hMSCs ($1.24\pm 0.23\mathrm{nmol}\text{e}/\min /{10}^6\text{cells}$). Under OA treatment, the value of oxygen consumption did not change compared to the value of undifferentiated hMSCs and significantly decreased to $<1\mathrm{nmol}\text{e}/\min /{10}^6\mathrm{cells}$ at weeks 1 and 4 after differentiation. Lactate release [Fig. 3(c)], an index of anaerobic glycolysis, was found to drop significantly in the beginning of normal osteogenic differentiation and then gradually increased afterward. The lactate release rate increased to about $90\mathrm{pmol}\text{e}/\mathrm{h}/{10}^3\mathrm{cells}$ after 2 weeks of differentiation but was still significantly lower than the value of undifferentiated hMSCs ($129.5\pm 15.8\mathrm{pmol}\text{e}/\mathrm{h}/{10}^3\mathrm{cells}$). Under OA treatment, the lactate release rate was higher than that under normal osteogeneic differentiation at all-time points from day 4 to week 4. At weeks 2 to 4, the lactate release rate was the same as that of undifferentiated hMSCs. The above results similar to those reported previously,⁹ indicated that the metabolic change of hMSCs is from glycolysis to more effective oxidative phosphorylation during normal osteogenic differentiation. However, under the inhibition of OA, the metabolic transformation from glycolysis to oxidative phosphorylation is not clear during osteogenic differentiation because both ATP and oxygen consumption did not pick up and lactate release was not efficiently inhibited.

Fig. 3

Measurements of adenosine triphosphate (ATP) content (a), oxygen consumption (b), and lactate release (c) of undifferentiated hMSCs (hMSC) and differentiated hMSCs at 4 days (ost4d), 1 week (ost1w), 2 weeks (ost2w), 3 weeks (ost3w), and 4 weeks (ost4w). Black bars represent normal osteogenic induction and white bars represent osteogenic induction with OA treatment. Symbols * indicates that the values are significantly different ($p$-value $<0.05$, Student’s $t$-test) as compared with undifferentiated hMSCs.

3.4.

NADH Fluorescence Lifetime Correlated with Oxygen Consumption and ATP Level

Figure 4 plotted the relations of NADH fluorescence lifetime versus ATP level [Fig. 4(a)] and versus oxygen consumption [Fig. 4(b)] for normal osteogenic differentiation of hMSCs (solid squares). It shows that the NADH fluorescence lifetime correlated well with the ATP level and oxygen consumption ($R^2=0.83$ and 0.91, respectively) suggesting that the NADH fluorescence lifetime is a reliable biomarker to reflect mitochondrial function during normal osteogenic differentiation. We also plotted the data points under OA treatment in the same figures (open squares). The NADH fluorescence lifetime remained at the minimal range regardless the change of ATP and oxygen consumption. Some of these changes are toward to lower values as compared to undifferentiated hMSCs.

Fig. 4

The correlation plots of NADH fluorescence lifetime with ATP level (a) and with oxygen consumption (b) at different time points. Black solid squares and open squares represent normal differentiation and differentiation with OA treatment, respectively. The correlation curve and $R^2$ values were calculated for normal differentiation group (black solid squares) only.

3.5.

Expression of Respiratory Enzyme Complex Proteins Increased During Osteogenic Differentiation

Increased expression of respiration enzyme complex proteins was reported to correspond to increased oxidative phosphorylation.^9,27 These complex proteins are responsible for a series of electron transport reactions in producing ATPs. NADH was known to strongly associate with respiration enzyme Complex I.³² The free/bound NADH ratio was suggested to dominate the change of NADH fluorescence lifetime.^31,33 Therefore, our observed increase in fluorescence NADH lifetime during osteogenic differentiation of hMSCs was possibly attributed to NADH interacting with respiration enzymes complex proteins, particularly Complex I. We analyzed the expression levels of protein subunits of respiratory enzyme Complex I-V by Western blot analysis as shown in Fig. 5. We found a significant increase ($\sim 2\text{-}\mathrm{folds}$) in the protein levels of core I subunit of Complex III and $\alpha$ subunit of Complex V after osteogenic induction for 3 weeks. In addition, a slight increase ($\sim 1.5\text{-}\mathrm{folds}$) in the expression levels of NADH-ubiquinone-oxidoreductase subunit 6 of Complex I and cytochrome oxidase subunit 5 of Complex IV was also noted. The statistical result of complex proteins expression was listed in the bottom table of Fig. 5. Additionally, their association significance ($p$-value) with NADH fluorescence lifetime was listed in Table 1.

Fig. 5

The expression level of respiratory enzyme complex proteins. The lower panel table is the statistical results by measuring the expression from at least three independent experiments and the value is the normalized expression to the value of hMSC. Symbol * indicates that the expression level of respiratory enzyme is significantly higher than that of the hMSCs.

Table 1

The association between the parameters of NADH lifetime (τm, τ1, τ2, a1/a2) and Complex I-V expression after osteogenic differentiation for 4 weeks.