The hMSCs were obtained from the Scientific Research Centre of China-Japan Union Hospital, Jilin University (Changchun, China). A vascular endothelial growth factor (VEGF) enzyme-linked immunosorbent assay (ELISA) kit was obtained from Cosmo Bio Co., Ltd. (Tokyo, Japan). A mouse/rat insulin-like growth factor-1 (IGF-1) ELISA kit was obtained from R&D Systems (Minneapolis, MN, United States). Immunohistochemical antibodies against marker of proliferation Ki-67 (MKI67) and β-catenin (CTNNB) were obtained from Abcam Tec (Shanghai, China). Minoxidil was obtained from Shanghai yuanye Bio-Technology Co., Ltd., (Shanghai, China).

The hMSC membrane was prepared according to published protocols (Gao et al., 2016; Yao et al., 2023). Briefly, 4 × 10⁶ hMSCs were lysed with hypotonic solution, and purified stem cell membranes were obtained by freeze-thawing and differential centrifugation. Next, MXD was mixed with the hMSC membrane solution, placed in an ultrasonic cleaner for 10 min, and centrifuged at 10,000 g for 30 min. Then, the pellet was resuspended in phosphate-buffered saline (PBS), and the precipitate was the STCM-MXD-NPs.

The morphology of the nanoparticles was observed by transmission electron microscope (TEM; JEM-2100F, JEOL, Tokyo, Japan). The samples were dispersed directly into PBS. A drop of the STCM-MXD-NPs suspension was transferred to copper grid. After staining with 3% (w/v) phosphotungstic acid solution and drying at room temperature, the sample was observed by TEM.

The absorption spectra of hMSC membranes, MXD, and STCM-MXD-NPs were measured at the same concentration using a UV spectrophotometer at a 231 nm wavelength to verify whether the hMSC membranes were successfully encapsulated with MXD.

This analysis used an Agilent ZORBAX SB-C18 (4.6 × 250 mm; 5 μm) column, methanol-water (70:30) mobile phase, 1.0 mL/min flow rate, 285 nm detection wavelength, and 30°C column temperature. A 1 mL sample of MXD solution or stem cell membrane-loaded MXD nanoparticle solution was filtered through a 0.22 μm microporous filter membrane. Then, 20 μL of this renewed filtrate was injected into the high-performance liquid chromatography (HPLC) instrument. The samples’ peak times and shapes were compared to verify whether the hMSC membrane was successfully loaded with MXD (Wang et al., 2017).

The STCM-MXD-NPs were diluted with double-distilled water and then purified with a 220 nm filter membrane. Their particle distribution and zeta potential were measured with a laser diffraction size distribution analyzer and zeta potential meter.

The amount of unencapsulated MXD was determined at a 231 nm wavelength using a UV spectrophotometer. The encapsulation rate was calculated using the formula: encapsulation rate = (total drug amount − unencapsulated drug amount)/total drug amount.

Briefly, 1 mL of STCM-MXD-NPs at 1 mg/mL was placed in a dialysis bag in the in vitro release medium and stirred uniformly at 200 rpm and a constant temperature of 37°C ± 0.5°C. Next, 1 mL of the release solution was removed at different time points, and the original solution was supplemented with an equal amount of media buffer. Then, the MXD concentration in the medium was determined by UV spectrophotometry, the cumulative percentage of drug release was calculated, and the release curve was plotted.

The treated experimental pig skin was placed between the two-halves of the Franz diffusion cell and secured with clamps in a circulating water bath with a rotational speed of 200 rpm and 32°C ± 0.5°C. Next, 1 mL each of pig skin grinding solution, stem cell membrane solution, 1 mg/mL of MXD, and STCM-MXD-NPs containing 1 mg/mL of MXD were added to the supply chamber, which was then sealed with cling film to prevent water evaporation. Then, 1 mL samples were taken from the receiving cell at 2, 4, 6, 8, 10, 12, and 14 h and replaced with an equal volume of fresh receiving solution at the appropriate temperature. After filtration through a 0.22 μm filter membrane, the sample’s drug content was determined by HPLC.

At the end of the transdermal experiment, the pig skin was removed, rinsed 3–4 times with PBS buffer, and then cut into pieces. Next, 1 mL of saline was added, and the sample was homogenized. Then, MXD was extracted from the pig skin with anhydrous ethanol in three batches (1 mL was used each time). Next, the supernatant was centrifuged at 12,000 rpm for 30 min at a low temperature. Then, the drug concentration was determined using HPLC, and the amount of drug retention was calculated.

Forty specific-pathogen-free-grade C57BL/6J mice aged 7 weeks and weighing 18 ± 2 g were obtained from the Experimental Animal Centre of Jilin University. All mice were housed under standard conditions (room temperature: 25°C; light/dark time = 12/12 h). The experiments were conducted after 1 week of acclimatization, during which time the mice were supplied with adequate food and water, and the cages and bedding were cleaned every 2 days. The experimental protocol was approved by the Ethics Committee of Jilin University.

The mice were randomly divided into four groups of 10: vehicle (PBS), STCM, 5% MXD, and 5% STCM-MXD-NPs. The hair on the back of the mice was removed before the experiment. During the experiment, 200 μL of PBS, STCM, 5% MXD, or 5% STCM-MXD-NPs was evenly applied to the skin of the depilated area on the back of the mice over an area of about 2 cm × 2 cm. The treatment was administered once daily for 15 consecutive days.

Hair growth was monitored daily at 9:00 a.m. using a digital camera under the same light source. Changes in hair area were assessed using the ImageJ image analysis software. The area under the hair area-time curve (AUC) was calculated.

After 15 days, the mice were euthanized by injection of a lethal pentobarbital dose. The skin was taken from the hair removal site on their back, fixed with 4% paraformaldehyde, dehydrated, embedded in paraffin wax, and stained with hematoxylin for 3–5 min. Next, the sections were dehydrated with various ethanol concentrations for 5 min each and stained with eosin. Then, the sections were dehydrated in rows, made transparent, and sealed with neutral gum. The hematoxylin-eosin (HE)-stained sections were observed under an ordinary microscope. The morphology of hair follicles was observed, the total number of hair follicles in each group was counted, and the ratio of anagen/telogen (A/T) follicles was determined.

The mice were euthanized by injection of a lethal pentobarbital dose. Their hair was collected with forceps, and the hair bulb region was separated. Next, the hair bulb and the rest of the skin tissue were homogenized separately in ethanol, and the homogenate was centrifuged at 20,400 g for 20 min at 4°C. Then, the MXD content in the supernatant was measured by HPLC as previously described (Wang et al., 2017). In this study, the MXD contents in the skin tissue and hair bulb are expressed as per mg protein, with total protein measured using a Bio-Rad protein assay kit.

The mice were euthanized by injection of a lethal pentobarbital dose. The total RNA in their hairballs was extracted using the Trizol reagent. The reverse transcription and qPCR reactions were performed using an RNA PCR Kit and LightCycler FastStart DNA Master SYBR Green I Kit according to the manufacturer’s instructions. RNA concentration and purity were determined using a Nanodrop 2000. A 20 μL reverse transcription reaction system was prepared to synthesize the complementary DNA (cDNA). Then, the cDNA was subjected to qPCR, with three replicate wells per sample. The qPCR procedure comprised pre-denaturation for 30 s, followed by 40 cycles of denaturation at 95°C for 15 s and annealing/extension at 60°C for 30 s. The mRNA expression levels of Vegf and Igf-1 were calculated.

Mice were euthanized by injection of a lethal pentobarbital dose. Hairball samples were homogenized in purified water according to the manufacturer’s instructions and centrifuged at 20,400 g for 20 min at 4°C. Next, the homogenates were incubated in microtiter plate wells pre-packed with mouse monoclonal antibodies against IGF-1 and VEGF. Then, detection reagents were added, and the absorbance was measured at 450 nm using a microplate reader.

After paraffin sections were deparaffinized and washed with water, they were placed in a microwave oven for antigen repair before being left to cool naturally. Next, the sections were washed, incubated, and washed again for serum closure. Then, primary antibodies against MKI67 and CTNNB were added separately and incubated overnight at 4°C. Next, a secondary antibody was added. Then, the section was washed and developed using 3,3′-diaminobenzidine, with color development terminated by rinsing the sections with running water; positive cells stained brownish yellow. The sections were counterstained with hematoxylin and washed with distilled water. Finally, the sections were dehydrated to make them transparent, sealed, and observed and imaged under a light microscope.

Data were statistically analyzed using SPSS software (version 19.0). Data are expressed as mean ± standard deviation (SD) and compared between groups using one-way analysis of variance (ANOVA). A p < 0.05 was considered statistically significant. The Origin software was used to create the graphs. Key: *, **, and *** represent p < 0.05, p < 0.01, and p < 0.001, respectively.

The morphology of STCM-MXD-NPs captured under transmission electron microscope was shown in Figures 2A–C. The MXD standard had a characteristic absorption peak at 231 nm. Its standard curve of absorbance at 231 nm as a function of concentration is shown in Figure 2D (y = 0.1479x + 0.0166; R ² = 0.9994), which shows a good linear relationship. The STCM-MXD-NP solution showed an absorption peak near 231 nm. The peak times of the STCM-MXD-NP and MXD solutions were the same, both at about 13 min, and the peaks had the same shape (Figures 2E, F), showing that the hMSC membranes were successfully encapsulated with MXD. In addition, we examined the zeta potential of the STCM-MXD-NPs. The newly prepared STCM-MXD-NPs were diluted and placed in a nanoparticle-size zeta potential meter. Their particle size was detected to be 181.4 ± 5.98 nm (Figure 2G), and their zeta potential was −31.6 mV.

The STCM-MXD-NPs prepared with different concentrations of MXD and hMSC membrane solutions showed different encapsulation efficiencies. The encapsulation rate first increased and then decreased with increasing MXD concentration, and the encapsulation rate reached about 62.56% at 1–9 mg/mL (Supplementary Data S1).

Then, we assessed MXD solubility by dissolving MXD in ethanol solution or PBS. MXD seemed insoluble in PBS, with white crystals gradually deposited at the bottom of the tube. However, it was soluble in ethanol. The hMSC membrane dissolved in PBS was a colorless transparent liquid. The STCM-MXD-NPs prepared by ultrasonication and low-temperature differential centrifugation formed a white precipitate. STCM-MXD-NPs dissolved in PBS formed a white suspension, indicating that hMSC membranes loaded with MXD NPs improved MXD solubility (Supplementary Data S2, S3).

The in vitro release experiments of STCM-MXD-NPs are shown in Figure 3A. MXD was slowly released from the hMSC membranes, with a cumulative release percentage of about 48.15% within 72 h, and it continued to be slowly released after 72 h.

Figures 3B, C show that the 14 h cumulative amount of MXD transmitting through the skin was 41.73 ± 3.57 and 59.03 ± 6.14 μg/cm² for MXD and STCM-MXD-NPs, respectively, showing sustained MXD transmission through the experimental pig skins. Therefore, the STCM-MXD-NPs have a more potent transdermal ability than MXD.

Figure 3D shows that after 14 h of transdermal absorption, MXD retention in the skin was 12.21 ± 1.56 μg with MXD and 50.46 ± 5.48 μg withSTCM-MXD-NPs. The intradermal MXD retention was higher in the STCM-MXD-NP group than in the MXD group at the same drug concentration.

Addressing how to deliver more MXD to the hair bulb is vital in treating alopecia. In order to compare the utilization of MXD and STCM-MXD-NPs, we examined the amount of MXD in skin tissues and hair bulbs. The amount of MXD in skin tissues of the STCM-MXD-NP group was 71.72% that of the MXD group. However, the amount of MXD in hair bulbs was 1.53 times higher in the STCM-MXD-NP group than in the MXD group (Figures 3E, F).

In order to evaluate the effects of STCM-MXD-NPs on hair development and the number of hair follicles in mice, we subjected the back skin of mice to depilation treatment. Figure 4 shows that the skin on the back of mice in all groups was pink, with no redness or swelling and apparent no rupture. On days 1–7, the skin color of the back of the mice did not change significantly in the Vehicle group, was light grey in the STCM group, was grey in the MXD group, and mostly gradually changed from pink to dark grey in the STCM-MXD-NPs group.

On days 8–11, the skin on the back of mice only turned grey in the Vehicle group, with no hair outgrowth. In contrast, only a tiny amount of hair outgrowth was visible on the backs of mice in the STCM group, and a small amount of hair outgrowth was visible on the backs of mice in the MXD group. However, the backs of the mice in the STCM-MXD-NP group were darker in color, and most of the hairs were outgrown.

On day 15, only a small amount of hair growth was evident in the Vehicle group, while hair growth was seen in the STCM group. Much hair growth was evident in the MXD and STCM-MXD-NP groups. However, the hair was lighter in color in the MXD group than in the STCM-MXD-NP group, where the hair was mainly black and glossy. There were no symptoms of skin irritation, such as redness, swelling, desquamation, and ulceration, at the site of the back of the skin.

In addition, we evaluated the hair growth rate of mice in each group (Figures 5A–C). The hair growth areas of mice in the STCM, MXD, and STCM-MXD-NP groups were 1.25, 1.57, and 1.90 times larger than that of the Vehicle group, respectively, with that of the STCM-MXD-NP group also being 1.21 times larger than that of the MXD group. In addition, the STCM-MXD-NP group showed a significant increase in hair growth scores after day 7 compared with the remaining groups. The STCM-MXD-NPs showed significantly better efficacy than MXD, significantly promoting hair growth.

Then, we assessed the number of hair follicles and the cycle of hair follicles in the dorsal skin of the mice. After 15 days of treatment, the number of hair follicles was significantly higher in the STCM-MXD-NP group than in the other groups. In addition, their hair length was longer, their hair follicles and the endogenous sheaths were well-developed and enriched with melanin particles, their hair follicles grew to the deeper layers of the subcutaneous tissues and sebaceous glands, and their hair follicles were in the anagen phase more often. The follicles of mice in the MXD group had fewer melanin granules than those in the STCM-MXD-NP group. Hair follicles and inner root sheaths were generally developed in the STCM group, and the bulb of hair follicles was more shallowly located. In addition, the number of hair follicles and the A/T hair follicle ratio were significantly higher in the STCM-MXD-NP group than in the MXD group (Figures 5D–G).

We evaluated the effects of STCM-MXD-NPs on the mRNA and protein expression of VEGF and IGF-1 in the hairballs of mice (Figures 6A–D). Their mRNA and protein levels were significantly higher in the STCM-MXD-NP group than in the other groups. The Vegf and Igf-1 mRNA levels were 2.22 and 1.64 times higher in the STCM-MXD-NP group than in the MXD group, and the VEGF and IGF-1 protein levels were 1.31 and 1.80 times higher in the STCM-MXD-NP group than in the MXD group.

The more vigorous the cell proliferation, the more pronounced the MKI67 expression and the deeper the hair follicle staining. After paraffin sections of mouse skin tissues from all groups were immunohistochemically stained for MKI67, mice in the STCM-MXD-NP group showed the deepest hair follicle staining and the highest MKI67 expression. The MKI67 expression in the STCM-MXD-NP group was 1.25-fold higher than in the MXD group, 1.93-fold higher than in the STCM group, and 2.74-fold higher than in the Vehicle group (Figures 6E, G).

CTNNB is important in promoting hair follicle cell proliferation and growth. It is strongly expressed in the anagen phase of hair follicles and weakly expressed in the catagen and telogen phases. Immunohistochemical staining for CTNNB in paraffin sections of mouse skin tissues from all groups showed that the hair follicles of mice in the STCM-MXD-NP group stained darker (i.e., had the highest CTNNB expression), and more hair follicles were in the anagen phase. The CTNNB expression in the STCM-MXD-NP group was 1.35-fold higher than in the MXD group, 2.61-fold higher than in the STCM group, and 4.42-fold higher than in the Vehicle group (Figures 6F, H).