Development and validation of a comprehensive in vitro organ model for androgenetic alopecia

Introduction

AGA causes hair thinning and baldness. The mechanisms remain unknown as many factors combine to progress the condition, including aging, genetics, lifestyle, androgen levels and other environmental variables. The frontal hair follicles (HFs) are the main area affected with occipital HFs mostly remaining normal. Current therapies offer only minimal improvements with limited longevity and only delayed progression in some patients.

In this study( Liu et al. BMC Biotechnology (2026) 26:76 )the authors developed an AGA organ model based on the use of HFs from the frontal region and optimised the dihydrotestosterone (DHT) concentration in the media to replicate the AGA-specific mechanisms of perturbed hair growth. They characterised this model using immunohistochemistry (IHC), Immunofluorescence, qPCR and RNA-seq with pathway analysis.

Main Points

• Optimal concentrations of DHT to inhibit growth in AGA hair follicles were identified, that replicate the effect of androgens, in 7-day organ culture. A concentration of 5×10⁶ mol/L DHT inhibited elongation of frontal HFs but not occipital hair follicles. Frontal HFs also demonstrated an increase in catagen phase and a higher mean hair cycle score (anagen follicles having the lowest score), whereas occipital HFs had no changes in catagen phase or mean hair cycle score.
• HFs from 10 AGA patients were tested with this concentration of DHT. Frontal HFs from all 10 donors demonstrated inhibitory effects when exposed to DHT. This varied from 12% to 35% compared to controls at day 7.
• H&E staining indicated that HFs from frontal and occipital regions had typical morphology, with intact bulbs and developed papilla. HFs from frontal regions had a smaller diameter compared to occipital HFs. Treatment with DHT did not affect the diameter of HFs, regardless of region.
• TUNEL and ki-67 staining revealed decreased proliferation and increased apoptosis in frontal HFs treated with DHT after 3 days. This effect was mostly observed in the keratinocytes. There was no effect on proliferation or apoptosis in occipital HFs.
• b-catenin expression was decreased 45% in the AGA model, b-catenin was unaffected in occipital HFs.
• The WNT inhibitor DKK1 had 4-fold higher expression in day 1 cultures of frontal HFs compared to occipital. Frontal HFs also had a 58% increase in DKK1 expression on DHT treatment. No change was observed in occipital HFs. However, there was an increase in DKK1 expression at day 3 compared to day 1 for HFs from both regions.
• Expression of TGFb1, a reported inhibitor of HF proliferation, was increased by 76% in the AGA model and was localised to the outer root sheath. TGFb1 mRNA expression was also increased in frontal HFs but not occipital. Similarly, p21, which has been associated with cellular aging, was increased in frontal HFs but not occipital 3 days after DHT treatment.
• Immunofluorescent staining indicated that androgen receptor (AR) expression was higher in the dermal papillae of frontal HFs compared to occipital HFs. AR expression was also increased by DHT in treated frontal HFs compared to controls.
• Frontal HFs treated with AR agonists minoxidil and bicalutamide in combination with DHT demonstrated increased hair regrowth compared to DHT treatment alone but not as fast as control HFs.
• RNA-seq and pathway analysis revealed differential expression of genes involved in WNT/ b-catenin signaling via decreased WNT5a expression and increased DKK1 expression. Upregulated TGFb signaling was also identified in the AGA model.

Conclusion

Current models of AGA, such as animal models, in vitro cellular models and human tissue models fail to give complete pictures of the progression of the condition. The authors have described a frontal HF-based model with an optimised concentration of DHT that correlated with reported effects of AGA on hair growth. Effects on proliferation, apoptosis, WNT signaling and AR expression were consistent with AGA progression. RNA-seq and pathway analysis confirmed the effects on WNT and TGF signaling and also highlighted inflammatory pathways activated in the AGA model. The authors conclude that this organ model could provide further insight into AGA research and development of more effective therapeutics. They also highlighted the broader applicability of using patient-derived hair follicles as an in vitro model for disease.

EPISTEM SERVICES

Biomarker Discovery Platform.

The Epistem plucked hair biomarker platform is the ideal solution for minimally invasive repeat sampling in biomarker discovery and clinical evaluation.

Preclinical ex vivo maintenance of plucked hairs has been used to identify biomarker signatures in the PI3K, RAF/MAPK, NOTCH, hair growth and DNA damage pathways.

We have also confirmed and identified mRNA biomarkers from a diverse range of clinical trial samples. We have also optimised a variety of protein biomarkers in hair via IHC at pre-clinical and clinical stages, these include b-catenin and AR, along with various proliferation and apoptotic markers. IHC biomarkers are also used for a number of phosphor-antigens and DNA damage markers.

Our biomarker platform is GCLP-accredited, regularly supporting clinical trials, and includes full bioinformatic support.

Gene Expression.

We offer GCLP-accredited NGS and qPCR services for gene expression analysis and have experience extracting RNA from most tissue types. We specialise in extracting quality RNA from challenging tissues. In situ hybridisation (ISH) is also available.

Histology and IHC

Epistem has a highly experienced histology team with over 15 years of expertise in validation and protocol optimisation. Our histology services can be offered independently or integrated as needed, all conducted within a GCLP-Accredited and HTA Registered Laboratory.

Epistem offers a fully automated IHC and ISH service utilising the Ventana Discovery Ultra. Protocols may be directly transferred or developed in-house which are tailored to your biomarkers of interest.

Bioinformatics.

All of our models are supported by bioinformatics analysis, including differential expression and pathway analysis. We also offer biomarker signature discovery and analysis using our proprietary software in clinical and preclinical samples.