Evaluation of Hypoxia-Conditioned MSC Exosomes’ Effects on GM-CSF and IL-1α Expression in a Rat Model of Androgenic Alopecia

Abstract

Objective

This study aims to evaluate the effects of hypoxia-conditioned mesenchymal stem cell exosomes (EH-MSCs) on granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-1α (IL-1α) expression in a dihydrotestosterone (DHT)-induced murine model of androgenetic alopecia.

Methods

This experimental study employed a randomized posttest-only control-group with 34 male C57BL/6 mice. These mice were divided into five groups: healthy control Group 1 (G1); DHT-induced control receiving saline (G2); DHT-induced mice treated with topical 5% minoxidil (G3); DHT-induced mice treated with EH-MSCs at 100 µg/kgBW (G4); and DHT-induced mice treated with EH-MSCs at 200 µg/kgBW (G5). The MSCs were isolated from umbilical cords, validated based on morphology, surface marker expression, and differentiation assays, and the exosomes were characterized by the presence of CD63 and CD9 markers. Gene expression of GM-CSF and IL-1α was quantified by real-time polymerase chain reaction. Statistical analyses included ANOVA or non-parametric equivalents, with p<0.05 being considered significant.

Results

EH-MSC treatment significantly modulated cytokine expression compared with controls. GM-CSF expression was upregulated in EH-MSC groups, with the highest level being found in G5 (3.42±0.31). This showed a dose-dependent effect (p<0.001). Conversely, IL-1α expression was downregulated, with the lowest level being found in G5 (0.95±0.33). This indicated potent anti-inflammatory activity (p=0.001). Finally, minoxidil did not significantly reduce IL-1α levels compared with saline.

Conclusions

EH-MSCs increased GM-CSF and decreased IL-1α expression in a dose-dependent manner, demonstrating dual regenerative and anti-inflammatory effects. These findings support the idea that EH-MSCs can serve as a promising therapeutic candidate for androgenetic alopecia.

INTRODUCTION

Androgenetic alopecia (AGA) is the most prevalent form of nonscarring hair loss. This is marked by progressive miniaturization of hair follicles and patterned alopecia. Its prevalence increases with age, affecting 30% of men in their 30s, 50% in their 50s, and up to 70% later in life¹. In Indonesia, AGA is responsible for 39.7% of alopecia cases, making it the most common hair disorder². To those affected by it, AGA significantly impairs their self-esteem and quality of life beyond its physical manifestations³.

Currently available treatments, including topical minoxidil, oral finasteride, platelet-rich plasma, and hair transplantation, often produce suboptimal results, require prolonged use, or carry risks and high costs⁴.

These limitations underscore the need for alternative therapeutic strategies.

Exosomes from mesenchymal stem cells (MSCs), enriched with bioactive proteins and microRNAs, are increasingly recognized as innovative tools for regeneration. When MSCs are subjected to hypoxic conditioning, MSC exosomes (EH-MSCs) show enhanced activities in promoting cell proliferation, modulating inflammation, and supporting tissue repair^{5, 6}. In alopecia research, exosomes have demonstrated the ability to promote hair regrowth in C57BL/6 mice through the rat sarcoma/extracellular signal-regulated kinase signalling pathway⁶ and to increase hair density in patients with AGA⁷. Despite these advantages, the immunomodulatory role of EH-MSCs in alopecia remains poorly explored. Inflammatory cytokines such as granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-1α (IL-1α) are central to hair follicle regulation. The former contributes to immune responses⁸. Meanwhile, the latter acts as a potent inhibitor of hair growth^{9, 10}.

The current study aims mainly to evaluate the effects of EH-MSC exosomes on GM-CSF and IL-1α expression in a dihydrotestosterone (DHT)-induced C57BL/6 mouse model of AGA. It is hypothesized that EH-MSC treatment reduces IL-1α expression and modulates GM-CSF activity. Thus, it exerts both anti-inflammatory and regenerative effects on hair follicles.

MATERIALS and METHOD

Study Design and Experimental Groups

This experimental in vivo study used a randomized post-test-only control-group design. Five groups of mice served as research subjects. They were a healthy control group without treatment (G1); a DHT-induced group receiving normal saline (G2); a DHT-induced group treated with topical minoxidil 5% (G3); and two DHT-induced groups treated with EH-MSCs at 100 µg/kgBW (G4) and 200 µg/kgBW (G5). To minimize bias, randomization was applied and alopecia induction was validated before treatment.

MSC Isolation and Hypoxia-Conditioned Exosome Preparation

MSCs were isolated from C57BL/6 mouse umbilical cords at 21 days’ gestation and cultured in Dulbecco’s Modified Eagle Medium supplemented with fetal bovine serum and antibiotics. The isolated cells displayed spindle-shaped, fibroblast-like morphology and adhered to the flask surface. To analyze these cells, flow cytometry was performed. Flow cytometry confirmed both positive and low expression for some clusters of differentiation (CD). positive expression was observed for CD29 (98.8%) and CD90 (98.7%). Low expression levels were observed for CD45 (0.36%) and CD31 (4.81%). Their multipotencies were verified by osteogenic and chondrogenic differentiation, as confirmed by Alizarin Red and Alcian Blue staining (Figure 1). After reaching 80% confluence, the MSCs were cultured under hypoxic conditions (5% O₂) for 24 h, and the conditioned medium was filtered by tangential flow filtration (TFF) using 100-500 kDa filters to concentrate extracellular vesicles in the exosome size range. This size-based TFF approach is in line with previously described protocols for MSC-derived exosome isolation using 100-300 kDa ultrafiltration/tangential flow membranes. Exosome protein concentration was determined using the bicinchoninic acid protein assay, and the final suspension used for treatment was adjusted to 80 µg/mL based on total protein content. Flow cytometry with CD63 and CD9 markers was then performed, following widely used surface-marker-based characterization methods for MSC exosomes (consistent with MISEV2018 recommendations), to phenotypically confirm the presence of exosomes (Figure 2) rather than to quantify their concentrations^{11, 12, 13}. Bead and isotype controls were used to define the negative region; a bead gate was then drawn on the FL3/FL4 dot plot and applied to subsequent histograms to assess marker expression. This isolation and characterization workflow is consistent with previously published protocols for MSC-derived exosomes, which commonly use ultrafiltration or TFF with membranes of ~100-300 kDa and identify vesicles by tetraspanin markers such as CD63 and CD9^{12, 13, 14}.

Figure 1.

Differentiation potential of MSCs. (A) Osteogenic differentiation confirmed by Alizarin Red staining showing calcium deposits (indicated by black arrow, 200× magnification). (B) Chondrogenic differentiation confirmed by Alcian Blue staining showing proteoglycan accumulation (indicated by black arrow, 100× magnification).

MSC: Mesenchymal stem cell exosomes

Figure 2.

Flow-cytometry characterization of EH-MSCs. Exosomes were captured on beads and stained with anti-CD63 and anti-CD9 antibodies conjugated to APC and PE, respectively. (A) Dot plot FL-3 PerCP-H vs FL-4 APC-H showing the bead gate (CD63+/ CD9+ events). (B) Histogram of PE-A (CD9-PE) for gated beads, showing a clear fluorescence shift compared to controls.

EH-MSC: Hypoxia-conditioned mesenchymal stem cell exosomes, CD: Cluster of differentiation, APC: Allophycocyanin, PE: Phycoerythrin, PerCP: Peridinin chlorophyll protein

Induction of AGA

Furthermore, AGA was induced in male C57BL/6 mice by daily intraperitoneal injection of DHT (1 mg/day) for 17 days. Intraperitoneal administration was chosen to ensure consistent systemic exposure of hair follicles to DHT via the circulation, to avoid repeated local dermal injections, and to employ a route and dose regimen reported to reliably induce DHT-mediated alopecia in C57BL/6 mice in previous preclinical studies^{15, 16}. They were then validated macroscopically (Figure 3). Compared with mice in the healthy control group, these mice showed reduced hair density in shaved areas. Microscopically, it indicated the absence of anagen-phase follicles on hematoxylin-eosin staining (Figure 4). Only mice that met these criteria were randomized into the experimental groups.

Figure 3.

Macroscopic validation of mice before and after treatment. Representative images of each group (G1-G5) on day 7 (after acclimatization) and day 25 (after 17 days of DHT induction).

DHT: Dihydrotestosterone

Figure 4.

Microscopic observation of hair follicles after DHT administration. Healthy mice (left) show normal hair follicle density and structure (red arrow), while DHT-induced mice (right) display miniaturized hair follicles characteristic of AGA (black arrow).

DHT: Dihydrotestosterone, AGA: Androgenetic alopecia

Experimental Animals, Sample Size, and Treatment Protocol

A total of 34 male C57BL/6 mice, aged 6-8 weeks and weighing 20-25 grams, were included. This sample size was calculated using the Federer formula, which required five mice per group. Moreover, one additional animal was added to each group to account for mortality, resulting in six mice per group. Finally, four additional mice were used for validation. These mice were housed under controlled conditions at 20-28°C with ad libitum access to food and water. After one week of acclimatization, DHT was induced from day 8 to day 24. On day 25, AGA was validated, and treatment was initiated. The healthy control group (G1) received no treatment. The DHT-induced control group (G2) received subcutaneous injections of normal saline (NaCl). Meanwhile, the positive control group (G3) received topical 5% minoxidil, applied twice daily for 14 consecutive days. Finally, the intervention groups received subcutaneous injections of EH-MSCs at 100 µg/kgBW (G4) and 200 µg/kgBW (G5) on days 25 and 32. All animals were euthanized on day 39 and their skin tissues were collected for analysis.

The low and high EH-MSC doses were set at 100 and 200 µg/mL based on prior in vivo work in a fluconazole-induced alopecia model, in which EH-MSCs at 100 and 200 µg/mL significantly reduced tumor necrosis factor-α and increased vascular endothelial growth factor (VEGF) levels without observable toxicity, indicating that this concentration range is biologically effective and well tolerated for hair-regrowth applications. In addition, MSC-derived exosomes administered at 100 µg/kg body weight effectively attenuated lipopolysaccharide-induced myocardial injury in mice^{17, 18}. Taken together, these data support 100 µg/kgBW as a biologically active in vivo dose for MSC exosomes and 200 µg/kgBW as a higher comparison dose within a safe experimental range.

RNA Extraction and Real-Time Polymerase Chain Reaction

These skin tissues were first homogenized. From them, RNA was extracted with Trizol and cDNA was synthesized by reverse transcription. The gene expression of GM-CSF and IL-1α was quantified by real-time polymerase chain reaction (PCR) with SYBR Green and specific primers, and normalized to internal controls.

Ethical Approval

All procedures performed on the animals in this study complied with institutional and national guidelines for the care and use of laboratory animals. The study protocol was reviewed and approved by the Institutional Ethical Committee of Universitas Islam Sultan Agung (approval no.: 396/VII/2025/Komisi Bioetik, date: 30.07.2025).

Statistical Analysis

All statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) for Windows, version 26.0 (IBM Corp., Armonk, NY, USA). Data are presented as mean ± standard deviation. Normality of each continuous variable was assessed using the Shapiro-Wilk test, and homogeneity of variance across groups was evaluated using Levene’s test. For variables that were normally distributed with homogeneous variances (Levene p≥0.05), intergroup differences were analyzed using one-way ANOVA followed by least significant difference post-hoc  test. For variables that were normally distributed but showed heterogeneity of variances (Levene’s test, p<0.05), Welch’s ANOVA was applied, and pairwise comparisons were conducted using Tamhane’s T2 post-hoc  test. Two-tailed p-values <0.05 were considered statistically significant. The reporting of statistical results was performed in line with international statistical reporting guidelines¹⁹.

RESULTS

Before evaluating treatment effects, the DHT-induced alopecia model was confirmed by comparing the healthy control group (G1) with the DHT-induced control group (G2). Macroscopically, G2 showed a visible reduction in hair density in the shaved dorsal area after 17 days of DHT induction compared with G1 (Figure 3). Microscopically, hematoxylin-eosin staining demonstrated a marked decrease in anagen hair follicles and pronounced follicular miniaturization in G2, whereas G1 retained normal follicular architecture (Figure 4). At sacrifice, G2 also exhibited significantly lower GM-CSF expression and higher IL-1α expression than G1 (GM-CSF: 0.44±0.12 vs 1.06±0.15, p<0.001; IL-1α: 2.35±0.93 vs 1.03±0.31, p<0.001; Tables 1 and 2), supporting successful establishment of a pro-inflammatory alopecic model.

Table 1. Descriptive statistics of GM-CSF and IL-1α relative expression across experimental groups.

Group	Mean ± SD	Shapiro-Wilk test (p-value)
GM-CSF	-	-
G1 (Healthy)	1.06±0.15	0.62†
G2 (NaCl)	0.44±0.12	0.28†
G3 (Minoxidil)	1.03±0.48	0.22†
G4 (EH-MSC 100 μg/kgBW)	2.91±1.01	0.50†
G5 (EH-MSC 200 μg/kgBW)	3.42±0.31	0.15†
Levene test (p-value)	<0.001
Welch’s ANOVA (p-value)	<0.001*
IL-1α	-	-
G1 (Healthy)	1.03±0.31	0.18†
G2 (NaCl)	2.35±0.93	0.14†
G3 (Minoxidil)	2.00±0.51	0.25†
G4 (EH-MSC 100 μg/kgBW)	1.28±0.48	0.84†
G5 (EH-MSC 200 μg/kgBW)	0.95±0.33	0.90†
Levene test (p-value)	0.11‡
One-way ANOVA (p-value)	0.001*

^†Normal (p>0.05), ^‡Homogenous (p>0.05), *Significant difference (p<0.05).

Data are summarized as mean ± standard deviation. Statistical analyses were performed using Shapiro-Wilk test for normality, Levene’s test for homogeneity of variance, and ANOVA (Welch’s ANOVA for GM-CSF and one-way ANOVA for IL-1α). G1: Healthy control, G2: NaCl control, G3: Minoxidil, G4: EH-MSC 100 µg/mL, G5: EH-MSC 200 µg/mL.

EH-MSC: Hypoxia-conditioned mesenchymal stem cell exosomes, GM-CSF: Granulocyte-macrophage colony-stimulating factor, SD: Standard deviation, IL-1α: Interleukin-1α

Table 2. Post-hoc comparison of GM-CSF and IL-1α expression between experimental groups.

Group	Group Comparison	GM-CSF	IL-1α
		p-value	p-value
G1 (Healthy)	G2 (NaCl) G3 (Minoxidil) G4 (EH-MSC 100 μg/kgBW) G5 (EH-MSC 200 μg/kgBW)	<0.001 1.000 0.063 <0.001	<0.001 0.007 0.450 0.796
G2 (NaCl)	G3 (Minoxidil) G4 (EH-MSC 100 μg/kgBW) G5 (EH-MSC 200 μg/kgBW)	0.258 0.018 <0.001	0.293 0.003 <0.001
G3 (Minoxidil)	G4 (EH-MSC 100 μg/kgBW) G5 (EH-MSC 200 μg/kgBW)	0.043 <0.001	0.038 0.004
G4 (EH-MSC 100 μg/kgBW)	G5 (EH-MSC 200 μg/kgBW)	0.964	0.314

*Significant difference (p<0.05).

Pairwise group comparisons were conducted using post-hoc tests following ANOVA (LSD for GM-CSF and Tamhane’s T2 for IL-1α). Results are presented as p-values. G1: Healthy control, G2: NaCl control, G3: Minoxidil, G4: EH-MSC 100 µg/mL, G5: EH-MSC 200 µg/mL.

EH-MSC: Hypoxia-conditioned mesenchymal stem cell exosomes, GM-CSF: Granulocyte-macrophage colony-stimulating factor, LSD: Least significant difference, IL-1α: Interleukin-1α

The EH-MSC administration significantly altered the expression of GM-CSF and IL-1α in DHT-induced C57BL/6 mice. GM-CSF expression differed markedly among groups, with the highest level observed in mice treated with EH-MSC200 µg/kgBW (G5), followed by the ones treated with EH-MSC 100 µg/kgBW (G4). Expression levels in both EH-MSC groups were significantly higher than those in the saline and the minoxidil groups. Our statistical analysis confirmed that these differences were significant (p<0.05), and the post-hoc testing indicated that the effect was dose-dependent (Tables 1 and 2 and Figure 5). These findings showed that EH-MSC treatment restored and enhanced regenerative cytokine activity in alopecia.

Figure 5.

Relative expression of GM-CSF in C57BL/6 mice after treatment. Bar graph showing mean ± SD of GM-CSF expression (fold change) across experimental groups (G1-G5). Statistical significance was determined using post-hoc analysis(*p<0.05, **p<0.01, ***p<0.001).

SD: Standard deviation, GM-CSF: Granulocyte-macrophage colony-stimulating factor

In contrast, IL-1α expression was reduced in EH-MSC-treated groups, with the lowest level recorded in the group receiving the 200 µg/kgBW dose. Both EH-MSC doses significantly decreased IL-1α expression when compared with the saline and minoxidil groups. The reduction was most pronounced at the higher dose, suggesting stronger anti-inflammatory activity (Tables 1 and 2 and Figure 6). The statistical analysis confirmed that the differences among the groups were significant (p<0.05), with the post-hoc comparisons showing consistent suppression of IL-1α by EH-MSCs.

Figure 6.

Relative expression of IL-1α in C57BL/6 mice after treatment. Bar graph showing mean ± SD of IL-1α expression (fold change) across experimental groups (G1-G5). Statistical significance was determined using post-hoc analysis (*p<0.05, **p<0.01, ***p<0.001).

IL-1α: Interleukin-1α, SD: Standard deviation

Taken together, the EH-MSC treatment increased GM-CSF expression and decreased IL-1α expression in a dose-dependent manner. These results indicate that EH-MSCs exert both regenerative and anti-inflammatory effects. This finding supported their therapeutic potential for AGA.

DISCUSSION

The EH-MSC treatment produced a dual effect by enhancing regenerative cytokine activity through GM-CSF upregulation and reducing inflammatory signaling through IL-1α downregulation. The elevation of GM-CSF in the intervention groups, especially at higher doses, was consistent with its known role in stimulating fibroblasts, keratinocytes, and endothelial cells to support tissue repair and angiogenesis^{8, 20}. In hair biology context, growth factors such as GM-CSF contributed to the prolongation of anagen phase of the follicular cycle²¹, suggesting that EH-MSCs created a favorable microenvironment for follicular regeneration. The combination of macroscopic hair loss, microscopic follicular miniaturization, and the shift toward lower GM-CSF and higher IL-1α, observed in the DHT-induced control group compared with healthy controls, further supports that our DHT protocol successfully established an androgen-driven, pro-inflammatory alopecic model.

The reduction of IL-1α following EH-MSC administration highlighted their anti-inflammatory and immunomodulatory properties. Known widely as the inhibitor of hair follicle growth and the driver of miniaturization in AGA^{9, 22}, it’s the suppression of IL-1α by EH-MSCs aligned with the evidence that MSC-derived exosomes carried anti-inflammatory proteins and microRNAs capable of downregulating proinflammatory pathways^{5, 23}. Interestingly, minoxidil failed to lower IL-1α expression to the same extent. This suggested that EH-MSCs provided broader mechanisms of action than conventional therapies.

In this study, topical minoxidil 5% was selected as a positive control because it represents the current first-line pharmacologic therapy for AGA and is widely used as a benchmark in both clinical and preclinical studies^{24, 25}. Minoxidil promotes hair regrowth mainly through vasodilation, opening of adenosine triphosphate-sensitive potassium channels, and upregulation of VEGF in dermal papilla cells, thereby improving perifollicular microcirculation and prolonging the anagen phase, rather than directly modulating inflammatory cytokines such as GM-CSF or IL-1α^{26, 27, 28}. Consequently, the absence of significant changes in these cytokines in the minoxidil group observed in our experiment is biologically plausible. By contrast, exosomes from hypoxia-conditioned MSCs are expected to exert broader paracrine and immunomodulatory effects, including regulation of pro-inflammatory mediators^{5, 29, 30}. Therefore, minoxidil serves as a clinically relevant comparator for hair regrowth efficacy and follicular outcomes, while highlighting the distinct mechanisms underlying conventional vasodilatory and cell-free regenerative therapies.

These findings are consistent with earlier reports that exosome-based therapies enhance hair regrowth and follicle stemness. The first report was by Mao et al.⁶ who demonstrated that exosomes derived from umbilical cord MSCs promoted hair regrowth in C57BL/6 mice by activating the rat sarcoma/extracellular signal-regulated kinase pathway. Meanwhile, Ersan et al.⁷ confirmed increased hair density in AGA patients following exosome treatment. The present results extend these findings by showing modulation of cytokine levels and providing molecular evidence for both regenerative and anti-inflammatory effects. However, a previous study had also noted that GM-CSF could play a dual role, exerting proinflammatory effects in autoimmune contexts⁸. This suggested that therapeutic safety depended on controlled dosing and standardized preparation of exosomes, as variability in MSC source and culture conditions could lead to inconsistent outcomes³¹.

Study Limitations

The study had several limitations. The first one was that the animal model might not fully replicate the pathophysiology of human alopecia. Secondly, despite the statistical sufficiency of sample size, it remained relatively small. Only two cytokines were analyzed, whereas alopecia involves complex signaling pathways, such as NF-κB and JAK-STAT^{32, 33}. In addition, the follow-up period was short. Therefore, the long-term effects of EH-MSCs remained unknown. It is recommended that future research expand molecular targets, include histological and macroscopic assessments of hair regrowth, and progress to human clinical trials to validate translational potential.

CONCLUSION

The EH-MSC treatment increased GM-CSF expression and decreased IL-1α expression in a dose-dependent manner. The findings indicated regenerative and anti-inflammatory effects that supported the potential of EH-MSCs as a therapeutic option for AGA.

Ethics

Ethics Committee Approval: The study protocol was reviewed and approved by the Institutional Ethical Committee of Universitas Islam Sultan Agung (approval no.: 396/VII/2025/Komisi Bioetik, date: 30.07.2025).

Informed Consent: This is a study conducted on animals.