Artificial skin is made by seeding human keratinocytes and fibroblasts onto a scaffold—collagen, fibrin, or a synthetic polymer—that is then matured in a bioreactor and, for permanent grafts, often cross-linked to resist breakdown. The result is a bilayer or single-layer construct that closes wounds, delivers cells, or serves as a test bed for drugs. It is used clinically for burns, chronic ulcers, and reconstructive surgery, and industrially for toxicity testing and cosmetic development.

The bilayer architecture and why two cell types are needed

Human skin is not one tissue. The epidermis is a stratified keratinocyte sheet that provides the water barrier. The dermis beneath it is a fibroblast-rich collagen matrix that provides mechanical strength and blood-vessel support. A graft that supplies only keratinocytes will close a wound but heal fragile and scarred. A graft that supplies only dermal matrix will not resurface the wound. The engineering problem is therefore to reproduce both layers, or to reproduce the dermal layer and let the patient's own keratinocytes migrate over it.

The most common permanent construct is a collagen-glycosaminoglycan sponge populated with allogeneic fibroblasts, covered by a silicone sheet that acts as a temporary epidermis. The fibroblasts remodel the collagen and secrete growth factors that encourage the patient's own cells to invade. The silicone is removed later and replaced by a thin epidermal graft. This is the design of a class of products that regulators classify as biologic or combination products, not simple devices.

Temporary constructs skip the cell-seeding step. They are acellular matrices—collagen, chitosan, or synthetic polymer—that act as a template for the patient's own cells. Acellular means no donor cells and no immune rejection of cells, but it also means the graft cannot deliver living cells to a wound that lacks them.

Scaffold chemistry: why collagen, fibrin, and polymers behave differently

Collagen is the default scaffold because it is the body's own structural protein. Type I collagen extracted from bovine or porcine tendon is purified, dissolved in acid, and freeze-dried to form a porous sponge. Pore size matters: pores smaller than roughly 50 micrometers exclude cells, while pores above about 200 micrometers allow vascular infiltration but weaken the matrix. Manufacturers tune pore size by controlling freezing rate and ice-crystal growth.

Fibrin is used when the graft must be injectable or sprayable. Fibrinogen and thrombin are combined at the point of use, forming a gel that polymerizes in minutes. The gel is soft and degrades quickly, so it is usually paired with a more durable matrix or used as a cell-delivery vehicle rather than a standalone graft.

Synthetic polymers—polycaprolactone, polylactic-co-glycolic acid, polyurethane—offer reproducible chemistry and tunable degradation. Their weakness is that they are not bioactive by default. Cells do not adhere to them well unless the surface is coated with fibronectin, collagen, or a peptide sequence such as RGD. A synthetic scaffold with an RGD coating can match collagen's cell-attachment performance at a fraction of the batch-to-batch variability.

The trade-off is consistent across all three: bioactivity versus reproducibility. Natural scaffolds are bioactive but vary between batches. Synthetic scaffolds are reproducible but need functionalization. Most commercial permanent grafts use a hybrid approach.

From biopsy to bioreactor: the manufacturing sequence

The process begins with a donor biopsy, typically a small full-thickness sample from a donor site. Keratinocytes and fibroblasts are separated by enzymatic digestion and cultured separately, because the two cell types have different growth requirements. Keratinocytes need feeder layers or defined media with low calcium to stay in a proliferative state; fibroblasts tolerate higher serum concentrations.

Cells are expanded over one to three weeks. Expansion is the bottleneck: keratinocytes divide slowly in early passages, and the number of doublings before senescence limits how much graft can be produced from one biopsy. A single biopsy can yield enough cells for many grafts, which is why allogeneic banks are feasible.

The expanded cells are then seeded onto the scaffold. Seeding density is a critical parameter; too few cells and the graft fails to establish, too many and the cells consume the scaffold's nutrients before vascularization. Seeded constructs are matured in a bioreactor that controls temperature, pH, and oxygen. Some designs use perfusion to deliver nutrients through the scaffold, mimicking the capillary flow the graft will encounter after implantation.

Maturation takes days to weeks depending on the product. The construct is then tested for sterility, cell viability, and mechanical strength before release. Regulatory oversight of the cell-culture steps is what makes these products expensive to bring to market; the manufacturing is closer to a pharmaceutical process than to a device line.

Quantified context: dimensions, timing, and the wound environment

Graft thickness is a design constraint, not a free parameter. Diffusion limits oxygen penetration, so a graft thicker than roughly 1 to 2 millimeters without a blood supply will develop a necrotic core. This is why thick dermal templates are designed to be vascularized quickly, and why thin epidermal sheets are used where the dermis is intact.

Wound closure timing drives construct choice. A graft placed on a clean, well-perfused wound bed vascularizes in about 7 to 14 days. A graft placed on an infected or ischemic bed fails regardless of construct quality, which is why wound-bed preparation precedes grafting. The graft does not fix a bad bed; it exploits a good one.

Cost is a practical constraint that shapes use. Permanent bilayer grafts are priced per unit area and are among the most expensive wound-care products. Acellular matrices are cheaper per area but deliver no cells. The cost difference is large enough that reimbursement decisions, not just clinical need, determine which products are used in which settings.

Construct type Cell content Typical role Key trade-off
Bilayer living skin equivalent Keratinocytes + fibroblasts Permanent burn and ulcer coverage Closest to native skin; highest cost and fragility
Dermal template with silicone Fibroblasts (allogeneic) Deep burns needing dermal replacement Requires later epidermal graft
Acellular collagen matrix None Template for host-cell ingrowth No cell delivery; relies on patient cells
Fibrin sealant with cells Keratinocytes or fibroblasts Cell delivery to irregular wounds Rapid degradation; often adjunctive
Synthetic polymer scaffold None or coated Research and select clinical uses Reproducible; needs biofunctionalization
Epidermal sheet (cultured) Keratinocytes only Resurfacing over intact dermis Fragile; no dermal component

Timeline: what happens to a graft after implantation

Phase Time window What occurs
Adherence 0–48 hours Graft attaches to wound bed via fibrin; no blood supply yet
Imbibition 1–4 days Graft absorbs wound fluid; nutrients arrive by diffusion
Inosculation 3–7 days Host capillaries connect to graft vessels; perfusion begins
Vascularization 7–14 days New vessels invade scaffold; graft stabilizes
Remodeling 2 weeks–months Fibroblasts reorganize matrix; scar matures
Silicone removal (if used) 2–3 weeks Temporary epidermis removed; epidermal graft applied
Long-term integration Months–years Graft contracts and remodels; final appearance settles

The timeline is a range, not a schedule. Infection, ischemia, or patient factors can delay each phase. A graft that has not vascularized by about two weeks is at high risk of failure and should be reassessed rather than simply observed.

Edge cases and the counterintuitive detail

The counterintuitive finding is that living cells in a graft are not always the active ingredient. In several dermal templates, the allogeneic fibroblasts are gradually replaced by the patient's own cells within weeks. The donor cells act as a temporary signaling population, secreting growth factors that recruit host cells, rather than as a permanent replacement. This means a graft can work even if most of its donor cells die, provided the matrix and the early signaling are intact. It also means that matching donor cells to the patient is less critical than matching the scaffold's mechanical and degradation properties.

Edge cases that change construct selection:

Burns over joints need grafts that tolerate movement; fragile epidermal sheets tear. Chronic ulcers with biofilm require debridement before any graft will take. Immunocompromised patients may reject allogeneic cells faster; acellular matrices avoid this. Pediatric patients have higher metabolic demand and faster remodeling; graft timing shifts. Grafts over exposed bone or tendon need a vascularized flap, not a skin substitute. Large-surface-area burns may exceed available donor biopsy expansion capacity; banking matters.

Misconceptions worth correcting

Artificial skin is not a permanent replacement for native skin. It is a scaffold and a cell delivery system that the body remodels. Final skin after grafting is usually scar tissue, not regenerated skin with hair follicles and sweat glands. Appendages do not regenerate from these constructs.

Artificial skin is not the same as a skin substitute dressing. Dressings are passive barriers; skin constructs contain cells or designed matrices intended to integrate. The distinction matters for coding, reimbursement, and expectations.

Artificial skin is not "printed" in most clinical products. Bioprinting exists in research, but the products in clinical use are typically cast, freeze-dried, or seeded scaffolds. The printing label is often marketing.

Artificial skin is not risk-free. Allogeneic cells carry a theoretical infectious-disease risk, which is why donor screening and inactivation steps are regulated. Acellular products avoid the cell risk but not the risk of scaffold failure or infection at the graft site.

Plain-language translation: a skin graft is a temporary scaffold that teaches the wound to heal itself, and it rarely restores perfect skin.

Practical actions: how to evaluate or use these products

For clinicians selecting a construct, the decision sequence is wound bed first, then construct. Debride and control infection before grafting; a graft on a bad bed wastes the product. Confirm perfusion; a graft on ischemic tissue will not vascularize. Match construct thickness and cell content to defect depth. Document the product's regulatory status and its approved indications, because off-label use shifts liability.

For researchers and buyers, verify the release criteria. Ask for cell viability at release, scaffold pore size, degradation time, and sterility testing method. A construct with no published release criteria is not a comparable product.

For patients and caregivers, ask what the graft is intended to do: close the wound, deliver cells, or template the dermis. Ask whether a second procedure (silicone removal, epidermal graft) will be needed. Ask what the expected final appearance is, because scar is the likely outcome, not normal skin.

Thresholds to reconcile: graft thickness limits (1–2 mm without vascular supply), vascularization timing (7–14 days), and silicone removal timing (2–3 weeks) are design and management guides, not rigid rules. They interact: a thicker graft needs faster vascularization, which needs a well-prepared bed, which needs infection control. Treat the numbers as a coupled system, not a checklist.

Limitations and what the evidence cannot settle

The field's central limitation is that graft success depends heavily on the wound bed and the patient, factors that construct design cannot control. Comparative trials between constructs are few, and many products are approved on mechanistic and safety grounds rather than on superiority. Outcomes such as scar quality, contraction, and long-term function are hard to measure and rarely reported consistently.

A second limitation is that the mechanisms described here are stable, but product-specific details change. Prices, approval status, and clinical recommendations vary by country and over time. Verify current labeling and reimbursement before making procurement or treatment decisions. The manufacturing principles—cell expansion, scaffold seeding, bioreactor maturation—are durable; the product landscape is not.

What can be concluded with confidence: artificial skin works by delivering cells, matrix, or both to a prepared wound bed, and its success is limited by vascularization and infection. What cannot be concluded from mechanism alone: which product is best for a given patient. That requires clinical evidence and wound-specific assessment, not a materials-science argument.