Human hair follicles are among the most information-dense mini-organs in dermatology. They orchestrate cycling, pigmentation, immune interactions, and tissue remodeling through tightly coordinated epithelial–mesenchymal crosstalk. Yet for many drug discovery and translational programs, a practical problem persists: we still lack reliable, compartment-resolved reference expression data for the human hair follicle that preserves spatial context.
The hairMONic atlas—a human hair follicle transcriptome database built from laser-capture microdissection (LCM)–based RNA-sequencing of defined human hair follicle compartments—recently published by QIMA Life Sciences/QIMA Monasterium in the Journal of Investigative Dermatology (2025) and publicly available as an interactive database, fills that gap. We spoke with first author Dr. Markus Fehrholz to understand why the team chose LCM over single-cell approaches, how the atlas was built, and how teams in pharma and biotech can use it to accelerate target validation, biomarker selection, and mechanism-of-action work in hair and scalp indications.
Markus, what was the central problem this study set out to solve, and why was it important enough to tackle now?
MF: We built the atlas because a fundamental bottleneck in human hair follicle research is knowing where genes are expressed at the compartment level – reliably, in human tissue, and with minimal processing artefacts. Hair follicles are highly structured: dermal papilla, dermal cup, connective tissue sheath, outer and inner root sheath, hair matrix compartments, each has distinct biology and different relevance for disease mechanisms and therapeutic intervention.
While single-cell RNA-seq has pushed the field forward, it often requires relatively large sample sizes and includes enzymatic and mechanical dissociation steps that can affect transcriptional profiles. For programs working with limited or precious material particularly diseased follicles, this becomes a real limitation. We wanted data that reflects what is actually happening inside the intact follicle, in each compartment, in its native physiological context. That is what laser-capture microdissection gives you.
The motivation from a drug discovery perspective is straightforward: if you want to intervene specifically in the dermal papilla to modulate hair inductivity, you need to know what that compartment expresses — not what a bulk homogenate of the whole follicle expresses. Compartment-specific target identification requires compartment-specific data.
Tell us about the hairMONic atlas itself: how is it built, and what it contains.
MF: The hairMONic atlas, which stands for “hair follicle MONasterium InteraCtive” is an interactive transcriptomic database derived from bulk RNA-seq on eight defined compartments of human occipital scalp terminal anagen VI hair follicles. The compartments are:
- bulge outer root sheath (bORS)
- suprabulbar outer root sheath (sbORS)
- inner root sheath (IRS)
- germinative hair matrix (gHM)
- precortical hair matrix (pcHM)
- dermal papilla (DP)
- lower dermal cup (DC)
- connective tissue sheath (CTS)
The database allows users to interrogate expression profiles in a compartment-resolved manner, including separation by male vs female donors. The idea is to provide a reference map that teams can use to rapidly evaluate target localisation, compartment specificity, and potential epithelial–mesenchymal communication patterns relevant to hair disorders.
“If your program targets the dermal papilla, you need dermal papilla data, not a whole-follicle average. The atlas gives you that precision for human tissue, publicly, for the first time.”
Dr. Markus Fehrholz
Walk us through the technical approach. Why did you choose laser-capture microdissection (LCM) rather than going straight to single-cell or spatial transcriptomics?
MF: Each method has strengths. Our rationale for LCM-based RNA-seq was driven by three needs:
- Minimal tissue requirement. LCM enables isolation of defined regions with very limited starting material, which matters when tissue is scarce or when you want to investigate diseased follicles where sample access is limited.
- Reduced manipulation artefacts. LCM allows profiling of cells in situ within their tissue compartment without enzymatic dissociation. Dissociation steps can introduce transcriptional changes and can distort signals in stress-responsive pathways.
- Physiological niche capture. Because we isolate intact compartments rather than dissociated single-cell clusters, we capture transcriptomic signatures that reflect the native niche—including signals from resident immune or endothelial cells present in mesenchymal regions. For many translational questions, this “native complexity” is a feature, not a bug.
We also complemented this by benchmarking our signatures against published microarray and scRNA-seq datasets, to show concordance while highlighting method-specific differences.
“The moment your dissociation protocol touches the cells, you have already changed their phenotype. LCM is the only approach that lets you interrogate the transcriptome of an intact physiological niche.”
Dr. Markus Fehrholz
How did you ensure that the compartments you dissected were truly specific and not cross-contaminated?
MF: This is central to LCM work. We built specificity into the workflow at multiple levels:
- Dissection strategy: during LCM we deliberately left a small margin between neighbouring regions to avoid cross-contamination, and we followed a defined order of isolation for compartments.
- Quality control: after sequencing, we excluded outliers and samples below mapping thresholds, and we used principal component analysis to confirm compartment-specific clustering and donor-independent clustering.
- Biological validation: we checked that established “signature genes” mapped to the expected compartments. For example, classic bulge-associated markers were enriched in bORS, dermal papilla inductivity markers peaked in DP/DC, proliferation markers enriched in gHM, and pigmentation-related markers localised to the expected matrix compartment.
- Orthogonal validation by in situ hybridisation: we validated selected, previously underappreciated markers by FISH in independent follicles. This step helps confirm that the transcriptomic localisation corresponds to actual tissue localisation.
What were the most scientifically significant findings, the results that genuinely surprised you?
MF: There were several. One of the most satisfying aspects was discovering novel compartment-specific markers, genes that, to our knowledge, had not been reported in human hair follicles in this context. We validated four of these by fluorescence in situ hybridisation, which is a stringent confirmation. PAPPA2, a metalloproteinase that enhances IGF bioavailability, was highest in the dermal papilla, consistent with its role in promoting dermal papilla cell proliferation seen in sheep and mice, but not previously demonstrated in human follicles. FOXM1, a transcriptional regulator of proliferation and stem cell maintenance, showed highest expression in the germinative hair matrix. Neither of these had been localised in human follicles at this resolution before.
For translational teams, these types of markers can help with compartment-specific readouts, sample QC, or mechanistic interpretation when testing novel targets.
The paper also includes a CellChat analysis. What does that add?
MF: CellChat is a computational framework to infer cell–cell communication networks from gene expression data. Although it’s commonly applied to scRNA-seq, we applied it here at the compartment level.
The analysis reinforced a biologically intuitive point: mesenchymal compartments (DP, DC, CTS) are major drivers of outgoing signalling, whereas outer root sheath compartments show higher predicted susceptibility to incoming signalling, based on receptor expression.
We also saw signalling patterns consistent with canonical hair follicle biology, including pathways such as Wnt, BMP, FGF, Notch, TGFβ, and VEGF. For drug discovery teams, this is helpful because it provides an additional layer to evaluate where ligand–receptor components are expressed and which compartment interactions may be relevant for mechanism-of-action hypotheses.
From a pharma perspective, how can teams use this atlas in practice?
MF: We see several immediate applications:
1) Target localisation and de-risking
If a target is proposed for alopecia areata, chemotherapy-induced alopecia, scarring alopecia, or hair pigmentation disorders, you want to know whether it is expressed in the relevant compartment (e.g., DP vs ORS vs matrix) and whether expression patterns suggest plausible biology.
2) Compartment-specific biomarker selection
When designing translational panels – qPCR arrays, RNA-seq signatures, protein readouts – you can use the atlas to select markers that are enriched where your mechanism is expected to act, improving signal-to-noise in ex vivo or clinical sample analyses.
3) Mechanism-of-action interpretation in human tissue models
For teams running ex vivo follicle studies or scalp tissue work, compartment-resolved expression helps interpret whether observed pathway changes are likely epithelial-driven, mesenchymal-driven, or reflect niche-level effects.
4) Hypothesis generation for epithelial–mesenchymal crosstalk
Hair disorders often involve disrupted signalling loops. The atlas plus communication analysis offers a structured way to explore candidate pathways rather than treating the follicle as a black box.
Why did you include male and female donors and what did you observe?
MF: We included both sexes because sex differences are relevant in hair biology, and reference datasets should make this visible rather than average it out. In our data, higher principal components showed sex-specific clustering, and, as expected, many of the strongest differences involve Y-chromosome–linked expression in male follicles, alongside a smaller set of non-Y genes that were differentially expressed by compartment.
In practical terms, this supports the idea that when programs study androgen-sensitive conditions or sex-biased phenotypes, reference expression context matters.
What makes this atlas different from existing datasets researchers may already be using?
MF: The key differentiator is compartment-resolved human follicle biology with minimal manipulation. Many datasets either focus on cultured dermal papilla cells, bulk skin, or dissociated single-cell clusters. Those are valuable, but they address different questions.
Our atlas provides a structured baseline across multiple follicular compartments in intact human tissue, with an interactive format so researchers can query targets and compare patterns quickly. It’s designed to be practical: something you can use early in a program to make better decisions about targets, markers, and model design.
How do you see the atlas evolving?
MF: The current atlas focuses on healthy anagen VI follicles. A major future direction is to extend this concept to diseased follicles or distinct hair cycle stages, and to deepen resolution where it adds value for translational work, always balancing feasibility and interpretability.
From a CRO standpoint, we’re also interested in integrating atlas-driven insights into study design: selecting endpoints and compartments more rationally for ex vivo programs and for mechanistic packages supporting therapeutic development. We are also seeking partners to complement and further develop the atlas.
For teams working on hair and scalp programs, what’s the best next step?
MF: A good first step is simply to explore the interactive atlas for your targets of interest and see whether the compartment distribution supports your current hypotheses. If you’re designing preclinical work in hair indications, we can also discuss how to translate atlas insights into a practical assay strategy: what tissue to use, what endpoints to measure, and how to build a robust mechanistic package in human-relevant models.
A final question: what would you say to a pharma or biotech R&D director who is wondering whether this kind of fundamental biology resource is actually relevant to their pipeline?
MF: Drug development in hair disorders has historically struggled with poor target selection and a disconnect between preclinical models and human biology. Part of that problem is that we have been working with inadequate maps. If you do not know where your target is expressed in the human follicle, which compartment, at what level, in what cellular context, your model selection, your endpoint design, and your interpretation of results are all compromised from the start.
The hairMONic atlas is not a curiosity. It is a reference resource that closes a real gap in the field, the equivalent of having a genome reference before you begin sequencing. The data is human, it is from unmanipulated tissue, it is sex-stratified, and it is publicly accessible. Any team working on hair loss, hair cycle disorders, folliculitis, or follicle-related inflammatory disease now has a resource they did not have before.
And for teams who want to go deeper, to build on the atlas data with custom studies, to profile diseased follicles, to map their specific target across compartments, or to establish co-development programs, QIMA Life Sciences is the team that built this atlas and knows this tissue better than anyone. We welcome that conversation.
You are developing a hair follicle or scalp program and want to discuss target localisation, compartment-specific biomarkers, or mechanism-of-action study design in human tissue models?
Acknowledgements and thanks to collaborators
This work was enabled by close scientific collaboration across academia, clinical partners, and industry. We thank all co-authors for their valuable contributions. We also thank Andrei Mardaryev, Leslie Ponce, Giammaria Giuliani, Fabio Rinaldi, Yoshikazu Uchida, and Tamas Biro for valuable scientific discussions and input.
We also acknowledge the contributions of our co-authors and partners, including colleagues at QIMA Life Sciences, QIMA Monasterium (Münster, Germany), Mediteknia Dermatology Clinic and University Fernando Pessoa Canarias (Las Palmas de Gran Canaria, Spain), Giuliani Pharma (Milan, Italy), and the University of Miami Miller School of Medicine (Miami, USA).
