Mitochondrial aerobic respiration is activated during hair follicle stem cell differentiation, and its dysfunction retards hair regeneration

Experimental animals

Eight-week-old C57BL/6 mice were used in all experiments except the old mice group (aged two-year old). All experiments were repeated at least three times with 3–5 mice per experiment. All animals received humane care, maintained in seperate cages with general rodent diet between 22 °C–24 °C.

The study was approved by the Ethics Committee of the Center, Scientific Research Center with Animal Models, Xiangya Hospital, Central South University (No: 2011-01-05). All procedures on animals followed the guidelines for humane treatment set by the Ethics Committee of the Center, Scientific Research Center with Animal Models, Xiangya Hospital, Central South University.

Preparation of tissue samples

Ketamine (80 mg/kg per mice) and Xylazine (5 mg/kg per mice) were injected i.p. before tissue preparation. After anesthesia, the skin samples with different phases of hairs were obtained from the back of mice after anesthesia, and the wound was sewed afterwards. Then the skin samples were incubated in 0.25% solution of Dispase (Dispase I, Sigma-Aldrich Co. LLC) in Hanks’ balanced salt solution (HBSS, Life technologies, Thermo fisher Scientific Inc.,Grand Island, NY) at 4 °C overnight. Based on previous research, hair follicles represent grey or black in anagen phase, while showing pink with no pigment during telogen phase (Maksim & Plikus, 2009). And the epidermis with telogen hair follicles or anagen hair follicles was separated with forceps under a binocular light microscope according to skin color and morphology.

MitoTracker

The telogen hair follicles with epidermis and the whole anagen hair follicles were incubated in 50 nM MitoTracker media (MitoTracker^® Red CMXRos; Life Technologies, Thermo Fisher Scientific Inc., Waltham, MA, USA) for 30 min at 37 °C. After incubation and washing, the tissues were incubated with 3 uM DAPI (DAPI, 4′, 6-Diamidino-2-Phenylindole, Dilactate, Life technologies, Thermo fisher Scientific Inc., Waltham, MA, USA) in PBS for 10 min at room temperature. Then the samples were mounted on the glass slide covered with glycerol and observed with a con-focal microscope.

Transmission electron microscope (TEM)

Immediately after removal of the mouse skin, tissues were sliced into small pieces (1 mm³) and fixed in 3% buffered glutaraldehyde (Glutaraldehyde 25% solution; Sigma-Aldrich Co. LLC) for 4 h at 4 °C. Tissue specimens were then fixed in 1% osmium tetroxide (OsO⁴, ReagentPlus^® , 99.8%; Sigma-Aldrich Co. LLC) for 90 min. Fixed tissue was dehydrated using ascending grades of ethanol and transferred into the resin via propylene oxide. After impregnation with pure resin, specimens were embedded in the same resin mixture. Ultra-thin sections of silver shades (60–70 nm) were cut using an ultra-microtome (Leica Rotary Microtome RM2255; Leica, Wetzlar, Germany) equipped with a diamond knife; sections were then placed on copper grids and stained with uranyl acetate (20 min) and lead citrate (5 min). Stained sections were observed with a TEM (JEOL JEM-1011) operating at 80 kV.

Detection of ROS

The telogen hair follicles with epidermis and the whole anagen hair follicles were washed with PBS and treated with 10 µM DCFDA (29,79-dichlorofluorescein diacetate, DCFDA-Cellular Reactive Oxygen Species Detection Assay Kit; AbcamInc., Cambridge, MA, USA) in DMEM (Dulbecco’s Modified Eagle Medium, Life Technologies, Thermo Fisher Scientific Inc.) for 20 min at 37 °C in the dark. The samples were washed with PBS 4 times and then mounted on a glass slide covered with glycerol and observed with a confocal microscopy.

Immunohistochemical Staining and Immunofluorescence staining

First, the skin samples were fixed in 4% Paraformaldehyde overnight at 4 °C and embedded with paraffin. After deparaffin and hydration, the samples sections were treated in boiling 0.01 M Tri-Sodium Citrate buffer (pH 6.0) for 20 min in water bath for antigen retrieval. And the samples were then incubated in 3% H₂O₂ at room temperature (22 °C –24 °C) for 10 min to quench endogenous peroxidase. Immunostaining procedure was carried out according to the manufacturer’s instructions for the M.O.M kit (Cat No. PK-220; Vector Laboratories Inc., Burlingame, CA, USA). The samples were incubated with primary antibodies for rabbit anti-SOD1 (1:200, Abcam Inc., Cambridge, MA) or mouse anti-SOD2 (1:200; Abcam Inc., Cambridge, MA, USA) overnight at 4 °C. The DAB substrate kit (Abcam Inc., Cambridge, MA, USA) was used for color development.

Early anagen hair follicles with epidermis were fixed in 100% methanol for 1 h. Fixed samples were treated with 0.5% triton X-100 for 15 min at room temperature (22 °C –24 °C) and blocked with 5% bovine serum albumin for 1 h at 37 °C. After rinsing with PBS, the samples were incubated at 4 °C overnight with PDK or PDH antibodies (1:200; Santa Cruz Biotechnology Inc., Dallas, TX, USA), and then incubated in antibodies against K15 and Ki67, respectively, at 37 °C for 1 h. Samples were rinsed with PBS four times (5 min each time) and incubated in the dark for 1 h at 37 °C with two appropriate fluorescence-labeled secondary antibodies. After rinsing with PBS 4 times (5 min each time), the samples were incubated with 3 µM DAPI in PBS and then were mounted on glass slides covered with glycerol and observed with con-focal microscopy.

Length of mitochondria was measured through Image pro plus 6.0. Integrated optical density (IOD) and area of figures were evaluated by Image pro plus 6.0, and mean density was calculated as: Mean density = IOD/area. A two-tailed student’s t test was used for comparison.

Drug preparation

Antimycin A stock (Sigma-Aldrich Co. LLC) was made by dissolving in DMSO, and the working solution was made by diluting the stock solution with DMEM to a final concentration of 10 µM prior to use.

Hair regeneration in vivo

Synchronous anagen was induced by depilation in the back skin of mice with all dorsal skin HFs in telogen stage of the hair cycle as described by Muller-Rover et al. (Porter, 2003). After HFs switched from anagen to telogen, we injected 100 µl antimycin A (experiment group) and DMSO (control group) intracutaneously on respective side of the mouse back for 10 days and 200 hairs were plucked at the drug treated sites at the 3rd day of treatment (Day 3). Mice were separated in different cages (1 mice per cage) and under close observation everyday. Pictures were taken on the studied location on mouse back every day and the time when the hair grows out was recorded.

Statistical analysis

Non-parametric Mann Whitney test was performed for comparison through GraphPad Prism 6.0 software. A value of P < 0.05 was considered statistically significant.

Change of mitochondrial ultrastructure during hair follicle bulge cells differentiation

During telogen phase, the inferior part of the hair follicle consists of mostly bulge stem cells and secondary hair germ (Mayumi Ito, Kenji & Kazuto, 2004). The bulge stem cells differentiate into proliferating matrix cells, entering the anagen phase (Wilson et al., 1994; Hideo Oshima et al., 2001). Hence in this paper, telogen phase hair follicle bulge cells (abbreviated as telogen bulge cells) and anagen phase proliferating hair follicle matrix cells (abbreviated as anagen matrix cells) were used as representative of HFSCs and differentiated HF cells respectively to detect the change in mitochondrial morphology and function during HFSCs differentiation.

Mitochondrial morphology was observed with electron microscopy. In the ultrastructure, more mitochondria were observed in anagen matrix cells (Fig. 1A) and they were elongated with organized cristae (Fig. 1C). However, mitochondria were discrete and spherical in telogen bulge cells (Fig. 1B) with less cristae (Fig. 1D). And the average length of mitochondria in anagen matrix cells was significantly increased than that in telogen bulge cells (Fig. 1E). Accordingly, the mitochondria became more sophisticated in ultrastructure after differentiation, implying that differentiated matrix cells have a higher energetic potential.

Alteration of mitochondrial activity in HFSCs differentiation

It was previously observed that iPSCs typically have glycolytic energy production at pluripotent phase, whereas oxidative phosphorylation is essential during cell proliferation and differentiation. In addition to the reduction of energy metabolism, iPSCs also have less mitochondria and lower mitochondrial activity than that in differentiated cells. The alteration of the mitochondrial morphology and function is a crucial marker of iPSCs differentiation (Varum et al., 2011). However, the change of mitochondrial activity during HFSCs differentiation has not been elucidated.

Thus, mitochondrial activity was assessed using Mito Tracker Red. The result indicates that, fluorescence intensity was significantly increased in anagen matrix cells compared with telogen bulge cells (Fig. 2A). Keratin 15 (K15) is known as a marker for stem cells while Ki67 symbolizes proliferating matrix cells (Eisinger et al., 2010). To locate HFSCs and proliferating HF matrix cells precisely, K15 and Ki67 were detected as shown in Fig. 2B. And there is an approximately three times increase in the fluorescence intensity of Mitotracker Red in Ki67⁺ proliferating cells than that of K15⁺ stem cells (Fig. 2B), suggesting an enhancement in mitochondrial activity during HFSCs differentiation, which is in accordance with the feature of embryonic stem cells (Chung et al., 2007). In addition, mitochondrial activity in young mice (8 week-old) does not differ from that of the old mice (2 year-old) (Fig. 2C), indicating that ageing does not have an significant influence on mitochondrial activity during stem cell differentiation.

Redox balance was sustained through enhancing SOD2 expression

As mitochondrion are the major generator of endogenous ROS in cells and electrons that leak out from the electron transport chain contribute to the production of ROS (Chen et al., 2008). H2DCFDA (2′, 7′-dichlorodihydrofluorescein diacetate) immunofluorescence was used to measure ROS levels in HFSCs. Unexpectedly, ROS levels are almost identical in these two stages of cell types (Fig. 3A). Nonetheless, expression of respiratory enzyme SOD2 was significantly improved in anagen matrix cells compared with that in the telogen bulge cells (Fig. 3B). We speculate that SOD2 levels are upregulated in anagen matrix cells to clear ROS during the differentiation process for redox homeostasis. Expression of SOD1, another respiratory enzyme, was also estimated during HFSC differentiation, but showed no significant difference (data not shown).

Change of respiratory enzymes expression during HFSCs differentiation

To determine the type of metabolism used by HFSCs, we again measured respiratory enzymes in K15⁺ stem cells and in Ki67⁺ proliferating cells. As shown in Fig. 4, PDK was highly expressed in HFSCs (Figs. 4A) while PDH was highly expressed in differentiated cells (Fig. 4B), indicating that anaerobic mitochondrial metabolism plays a dominant role in HFSCs, whereas aerobic metabolism is essential in differentiated cells.

Suppression of mitochondrial oxidative phosphorylation delays hair regeneration

Oxidative phosphorylation increases during HFSCs differentiation, which is supplied mainly via the respiratory pathway (Armstrong et al., 2010). It is temping to speculate that disrupting mitochondrial oxidative phosphorylation might inhibit the differentiation and proliferation of hair stem cells and retard hair regeneration. Hence, a respiratory inhibitor, antimycin A [complex III inhibitor], was injected on one side of mouse dorsal skin subcutaneously to prohibit mitochondrial activity. DMSO was treated on the contralateral side as the control. The treatment process is summarized in Fig. 5A. 200 hairs were plucked after three days of drug treatment. After plucking, hair regrowth was recorded in these treated regions and the appearance of neonatal hair was monitored by taking photographs each day (Fig. 5B). The antimycin A group showed significant delays (9.6 ± 0.9 days) in hair growth compared with the control group (6.7 ± 0.7 days), as shown in Fig. 5C (P < 0.05). Accordingly, disruption of mitochondrial respiration leads to the delay of hair follicle regrowth, revealing that alteration of mitochondrial respiratory function might be essential in HFSCs differentiation.