Hair Follicle Model: Promoting Growth and New Drug Insights

Laboratory-grown hair follicle models have become one of the most important tools for understanding why hair grows, why it stops, and how new drugs might reverse hair loss. These models range from simple cultures of individual plucked follicles to elaborate stem-cell-derived organoids that sprout pigmented hairs over several months in a dish. Their collective value is straightforward: they let researchers test candidate treatments on human tissue without waiting years for clinical trials, and they reveal biological mechanisms that animal studies alone cannot capture.

Why Human Hair Needs Its Own Models

Hair follicles are deceptively complex. Each one is a miniature organ involving cells from at least three developmental lineages: the outer skin layer, the underlying connective tissue, and neural-crest-derived cells like the melanocytes that produce pigment. These cell types constantly exchange chemical signals, and the specific signaling choreography determines whether a follicle grows a new hair, rests, or sheds. The Wnt signaling pathway, for example, plays distinct and tightly timed roles in both the skin surface and the deeper tissue layers during follicle formation, orchestrating when and where new follicles appear.¹

Mouse models have been the workhorse of hair biology for decades, but they have a fundamental translation problem. Human and mouse follicles differ in their growth cycles, their hormonal responses, and the behavior of the dermal papilla cells at the follicle’s base. A striking example: cultured rodent dermal papilla cells can spontaneously clump together and induce new hairs when transplanted onto rodent skin, but human dermal papilla cells do not do this on their own.² That single difference means a compound that regrows fur in mice may do nothing for a person. Building models from human cells and tissues closes that gap.

Ex Vivo Organ Culture

The most straightforward approach is to take actual human hair follicles, dissect them out of scalp tissue, and keep them alive in a nutrient bath. This technique, first refined in the 1990s, remains one of the best preclinical tools available. It preserves the three-dimensional architecture of the follicle and allows researchers to watch, in real time, how a compound affects growth over days or weeks.³

These cultured follicles respond to the same drugs that work in people, making them a reliable first-pass test. Adding a compound to the culture medium mimics what happens when you take a pill; it reaches the follicle through the surrounding fluid, much like a drug delivered by the bloodstream. More recent protocols have expanded the toolkit: instead of isolated follicles, researchers can now culture full-thickness pieces of human scalp skin. This lets them test topical formulations applied to the skin surface and intradermal injections, not just dissolved compounds.⁴ The practical advantage is obvious: if a company is developing a topical serum or a microneedle patch, they can test the actual delivery method, not just the active ingredient.

The main limitation of ex vivo culture is supply. Healthy human scalp tissue typically comes from cosmetic surgery, and as those procedures have become less invasive, the tissue samples have gotten smaller and harder to obtain. That bottleneck has pushed the field toward building follicles from scratch.

3D Spheroids and the Dermal Papilla

At the very base of every hair follicle sits a tiny cluster of specialized cells called the dermal papilla. These cells act as the follicle’s command center, sending signals that tell surrounding cells when to multiply and form a new hair shaft. When dermal papilla cells are grown flat on a laboratory dish, they quickly lose their identity and their ability to induce hair growth. They become generic fibroblasts. This has been one of the central frustrations in hair biology for years.

The breakthrough came from a deceptively simple idea: let the cells form three-dimensional clumps, called spheroids, instead of forcing them flat. When human dermal papilla cells are grown as spheroids, their gene activity shifts back toward what it looks like inside a living follicle, and they partially regain their ability to trigger new hair formation. In one landmark experiment, spheroid-cultured human dermal papilla cells were transplanted between layers of human skin and induced entirely new follicles to form, something flat-cultured cells could not do.⁵

Building on that principle, researchers have used 3D spheroid dermal papilla models as screening platforms for hair-growth compounds. A global gene analysis of cells grown in hanging-drop spheroids showed that the 3D format activated cell adhesion molecules and hair-growth-related signaling pathways far more than flat culture did.⁶ These spheroids respond to minoxidil, the active ingredient in Rogaine, in measurable ways, which gives researchers a benchmark: if a spheroid model does not respond to a known hair-growth drug, something is probably wrong with the model. That kind of internal validation is critical when you are trying to evaluate new, unproven molecules.

Skin Organoids Grown from Stem Cells

The most ambitious models skip tissue donation entirely and start from human pluripotent stem cells, the blank-slate cells that can become any tissue in the body. Over a carefully timed series of chemical cues spread across about two weeks, these stem cells are coaxed into becoming the two cell populations that give rise to skin: the surface ectoderm (which becomes the epidermis) and the cranial neural crest (which becomes the dermis). From there, the researchers essentially step back and let developmental biology take over.⁷

The results are remarkable. By about 60 days in culture, the organoids produce hair follicles. By roughly 130 days, they have developed stratified skin layers, pigmented hairs, sebaceous glands, touch-sensitive Merkel cells, and sensory neurons. The overall complexity resembles fetal skin at around 18 weeks of gestation.⁷ These organoids are not cosmetically useful hair, to be clear. They are small, spherical masses covered in tiny follicles. But they recapitulate the developmental process with enough fidelity that researchers can study how follicles form, why they sometimes fail, and how drugs influence each stage.

The timeline is slow compared to simpler models, but the payoff is a system where every cell type arose from a single, genetically defined starting population. That means you can create organoids from patients with specific genetic hair disorders and watch the disease unfold in a dish, comparing them against organoids from healthy donors.

Engineered Platforms and Bioprinting

Between the biological elegance of organoids and the practical simplicity of organ culture, a middle ground has emerged: engineered platforms that combine living cells with manufactured scaffolds. One approach uses 3D bioprinting to deposit dermal papilla cell spheroids at precise locations within a pre-formed layer of dermal tissue. When keratinocytes and melanocytes were added, the resulting constructs developed structures resembling native hair follicles, complete with the right cell types in roughly the right arrangement.⁸

Other groups have experimented with hydrogel-based bioinks that mimic the composition of real skin. One team formulated a gel from modified gelatin and hyaluronic acid, matching the collagen and glycosaminoglycan ratios found in natural skin. The gel maintained the hair-inductive properties of the embedded cells while also forming pore-like openings on the surface that resemble the openings hair shafts grow through.⁹

Microfluidic “hair-on-a-chip” devices take a different tack. Rather than printing tissue, they carve tiny channels and wells into a chip, then seed living hair tissue into them. One recent design used laser-drilled holes roughly 150 micrometers wide in a scaffold, sealed on one side with a layer of blood vessel cells. The setup allowed long-term culture of hair tissue without the contraction and collapse that plague traditional gel-based systems.¹⁰ Chips like these can be connected to pumps that flow nutrients past the tissue, simulating blood circulation, and they are small enough that dozens can run in parallel.

Modeling Hair Loss Diseases

One of the most practical applications of these models is recreating specific diseases in a controlled environment. Androgenetic alopecia, the progressive thinning that affects the majority of men and a substantial share of women, is driven by the hormone dihydrotestosterone (DHT). Researchers have built organoid-based models of this condition by exposing hair-bearing organoids to DHT, which reduces the number of pigmented hairs and suppresses growth-related biomarkers. When known protective agents were added alongside DHT, hair follicle numbers and biomarker expression recovered, confirming that the model faithfully mimics the disease and responds to treatment.¹¹

A separate group built a more detailed organ model of androgenetic alopecia and characterized the cellular damage in specific terms: DHT exposure caused about a 24% drop in the proliferation of matrix cells (the rapidly dividing cells that build the hair shaft) and a 31% rise in cell death. At the molecular level, key growth-promoting signals decreased while growth-inhibiting signals climbed.¹² Numbers like these give drug developers concrete targets: a successful treatment should reverse those specific shifts.

Alopecia areata, the autoimmune form of hair loss where the immune system attacks follicles, presents a different modeling challenge. The hallmark of the disease is the collapse of the follicle’s “immune privilege,” a protective shield that normally prevents immune cells from attacking the follicle’s inner compartments. Researchers have developed protocols to deliberately break down this immune privilege in cultured human follicles, creating a controllable version of the disease in a dish. The same system can then be used to screen compounds that might prevent or reverse the immune attack.¹³ This is particularly valuable because alopecia areata has been difficult to model in animals, and the approved treatments (JAK inhibitors like baricitinib and ritlecitinib) were developed largely through clinical observation rather than systematic preclinical screening. Better models could accelerate the search for next-generation therapies.

Scaling Up for Drug Screening

A beautiful model is useless for drug discovery if you can only make one at a time. The field has invested heavily in making these systems scalable. Recent high-throughput platforms use co-cultures of epithelial and mesenchymal cells in standardized well plates, producing dozens or hundreds of miniature follicle-like units under consistent conditions. These platforms allow precise, reproducible assessment of whether a compound promotes or inhibits hair growth, and they are designed to be simple enough for a pharmaceutical lab to run routinely.¹⁴

Another organoid system designed for screening emphasizes scalability and simplicity while maintaining enough biological complexity to reveal not just whether a molecule works, but how it works: what pathways it activates, what concentration is optimal, and whether combining it with other active ingredients produces synergy or cancellation.¹⁵ That last point matters more than it might sound. Many topical hair products contain multiple active ingredients, and without a reliable screening model, figuring out whether ingredient A helps or hinders ingredient B requires expensive trial and error in human volunteers.

Adding Blood Vessels and Immune Cells

Most current hair follicle models are missing two things that real follicles depend on: a blood supply and an immune system. In living skin, every follicle is surrounded by a network of tiny blood vessels, and immune cells patrol the tissue constantly, influencing the hair cycle in ways researchers are still mapping. The absence of these components limits what today’s models can tell us about inflammatory hair disorders or about how a drug’s effect on blood flow might contribute to hair growth (a major part of how minoxidil is thought to work).

Recent work has begun to address both gaps. One group used digital-light-processing 3D bioprinting to create a miniaturized skin-on-a-chip model, about the size of a standard clinical biopsy, that included engineered hair follicles surrounded by a superficial vascular plexus. The model was small enough to study immune cell trafficking through the tissue.¹⁶ A separate review has outlined the broader vision: moving from structurally focused organoids toward what the authors call “immune-competent hair follicle microphysiological systems,” platforms that incorporate immune cells, perfusion (flowing nutrients like blood would), and neural connections alongside the hair follicle itself.¹⁷

Building these features in is not just an academic exercise. Alopecia areata, as mentioned earlier, is an immune-mediated disease, and testing immune-modulating drugs in a model that lacks immune cells is inherently limited. Similarly, some promising hair-growth compounds fail in clinical trials not because they do not stimulate follicles, but because they cause inflammation or are cleared too quickly by the immune system. Models that include these layers of biology should catch more of those failures earlier.

The Extracellular Matrix as a Growth Signal

The tissue surrounding follicle cells is not inert scaffolding. The extracellular matrix, the mesh of proteins and sugars between cells, actively influences whether and when hair grows. Recent research has identified a specific subpopulation of fibroblasts, marked by a protein called Hmmr, that appears to sense mechanical properties of the surrounding matrix. When the matrix softens and becomes more viscoelastic in a particular region, these Hmmr-positive fibroblasts detect the change and relay a growth signal to dermal papilla cells. The researchers describe a three-part mechanical module: the matrix remodels (the trigger), the Hmmr fibroblasts detect it (the sensor), and the dermal papilla cells respond by initiating hair regeneration (the executor).¹⁸

This finding has immediate implications for model design. If the stiffness and elasticity of the scaffold material matter to hair induction, then models built on overly rigid or overly soft substrates may be giving misleading results. It also opens a new therapeutic angle: rather than targeting the follicle cells directly, future drugs might work by altering the mechanical environment of the skin to make it more permissive for hair growth. Microneedling, which is already used off-label for hair restoration, may derive part of its effect from exactly this kind of matrix disruption, though that connection remains speculative.

What These Models Cannot Yet Do

For all their progress, current hair follicle models have real limits that anyone evaluating new treatments should understand. The stem-cell-derived organoids produce follicles that resemble fetal rather than adult hair, and no one has yet coaxed them through a full adult hair cycle of growth, regression, rest, and regrowth in a dish. Ex vivo cultures eventually stop growing after a few weeks because they lack the hormonal and neural inputs that sustain cycling in a living person. Most engineered platforms still omit the sebaceous gland, the oil-producing structure that is integral to real follicle function and is implicated in conditions like acne and seborrheic dermatitis. And while immune-competent models are being developed, they remain early-stage prototypes rather than validated screening tools.

The species-translation problem, while improved, has not vanished. Even the best human models are tested in isolation from a living body. A compound may look effective in a dish but fail once it encounters human metabolism, blood-brain barrier restrictions, or off-target effects in organs that no skin model includes. Hair follicle models are powerful filters that eliminate bad candidates early; they are not replacements for clinical trials.

From Model to Medicine

The practical pipeline for a new hair-growth drug increasingly runs through these models in a defined sequence. Early-stage candidates are screened against high-throughput spheroid or organoid platforms to identify compounds that promote growth markers and suppress regression signals. Promising hits are then tested in ex vivo human follicle cultures, where researchers can measure actual hair shaft elongation over time. Disease-specific models, like the DHT-driven androgenetic alopecia systems, allow testing against the specific biology of the condition the drug is meant to treat. Only candidates that clear all these human-tissue hurdles advance to clinical trials in people.

This pipeline matters because the history of hair-loss treatment is littered with compounds that worked in mice but not in humans. Minoxidil and finasteride, the two FDA-approved treatments for pattern hair loss, were both discovered by accident rather than through systematic screening. The field’s hope is that purpose-built human models will shift drug discovery from serendipity to engineering, making it possible to develop treatments that are both more effective and more precisely targeted than what is currently available. Whether that hope materializes depends on whether the remaining biological gaps in these models, especially immune function, vascularization, and full hair-cycle control, can be closed in the coming years.