The goal of this protocol is to build up a three-dimensional full thickness skin equivalent, which resembles natural skin. With a specifically constructed automated wounding device, precise and reproducible wounds can be generated under maintenance of sterility.
In vitro models are a cost effective and ethical alternative to study cutaneous wound healing processes. Moreover, by using human cells, these models reflect the human wound situation better than animal models. Although two-dimensional models are widely used to investigate processes such as cellular migration and proliferation, models that are more complex are required to gain a deeper knowledge about wound healing. Besides a suitable model system, the generation of precise and reproducible wounds is crucial to ensure comparable results between different test runs. In this study, the generation of a three-dimensional full thickness skin equivalent to study wound healing is shown. The dermal part of the models is comprised of human dermal fibroblast embedded in a rat-tail collagen type I hydrogel. Following the inoculation with human epidermal keratinocytes and consequent culture at the air-liquid interface, a multilayered epidermis is formed on top of the models. To study the wound healing process, we additionally developed an automated wounding device, which generates standardized wounds in a sterile atmosphere.
The skin is the largest organ of the body. It creates a barrier between the external environment and the internal organs. Moreover, the skin protects the body from fluid loss, environmental influences, injuries and infections and helps to regulate body temperature1. Due to its exposed location, the skin is often affected by mechanical, thermal or chemical trauma. Though the skin is generally capable of self-repair, multiple local factors such as infection, oxygenation, and venous sufficiency can lead to impaired wound healing. Wound healing can also be interfered by systemic factors as obesity, alcoholism, smoking, medication, nutrition and diseases such as diabetes.2
The process of wound healing can be divided into 3 phases: (i) the inflammatory phase, (ii) the proliferative and the (iii) remodeling phase. Upon injury to the skin, a complex signal-cascade starts, leading to the closure of the wound.3 After injury, the wound is bleeding and a blood clot is formed. Fibroblasts move into the blood clot and replace it with new tissue which is subsequently remodeled over years.
The current understanding of the biological processes underlying cutaneous repair is limited. Small animal and pig models have been used to study wound healing. However, these results cannot be directly transferred to humans due to species-specific differences. In addition to these in vivo models, some aspects of wound healing can be studied by simulating a wound situation via scratching of in vitro monolayer cultures based on immortalized cell lines or primary cells.4 These scratching models are highly standardized but do not sufficiently reflect the complex in vivo physiology.5 Besides two-dimensional models, three-dimensional human skin equivalents have been developed for dermatological research. The dermal part of these models are generated using various scaffolds including decellularized dermis,6 collagen hydrogels,7,8 glycosaminoglycans9 or synthetic materials.10 Employing these skin equivalents, the role of epithelial-mesenchymal interactions11, the re-epithelialization, the cellular crosstalk between fibroblasts and keratinocytes and the influence of different growth factors can be studied. Moreover, these models are useful to gain new knowledge about how fibroblasts migrate into the wounded area and how chemotactic factors influence tissue regeneration.12
Not only generation of the wound healing model itself is challenging, but also to establish a highly standardized wound in a model is problematic. Common techniques to create wounds are scratch tests,13 burns,14 tape abrasion,15 thermal injury,16 suction blisters,17 liquid nitrogen,18 lasers,19 scalpels,18 meshers6 and biopsy punches.20 Most of these methods have the same pitfalls. Injuries implemented manually are difficult to standardize and to reproduce between multiple tests. Size, shape and depth of the wound vary between studies and thus impair the quality of research data. The use of laser for defined skin wounding can be relatively easily standardized but leads to a situation mimicking burn wounds. Heat applied by laser can cause protein denaturation, platelets aggregation or vessels constriction, which can lead to necrotic tissue.
In an alternative approach, we developed an automated wounding device (aWD), which allows us to generate defined and precise cutaneous wounds under sterile conditions. Wounding parameters, like depth and speed of penetration as well as revolutions of the drill head can be regulated. In this study, we combined the aWD with in house developed full thickness skin equivalents (ftSE) that are comparable to a protocol published by Gangatirkar et al. 8 The dermal layer of the skin equivalent is composed of human dermal fibroblasts (hDF), which are embedded in a collagen I hydrogel. On the dermal layer, human epidermal keratinocytes (hEK) are seeded. Within two weeks at the air-liquid interface the hEK build up an epidermis composed of several vital cell layers and a stratum corneum. Besides the generation of this model, this study shows the use of the aWD to create defined and precise wounds in ftSE.
NOTE: The protocol is designed for the production of 24 full thickness skin equivalents. Human dermal fibroblasts and epidermal keratinocytes were isolated from skin biopsies according to a previously published protocol.21,22 Informed consent was obtained beforehand and the study was approved by the institutional ethics committee on human research of the Julius-Maximilians-University Würzburg (vote 182/10).
1. Production of Dermal Component
2. Addition of Human Epidermal Keratinocytes (hEK)
3. Culture of Full Thickness Skin Equivalents (ftSE)
4. Histological Analysis
5. Injury by Use of the Automated Wounding Device (aWD)
The isolated hDF and hEK differ both in morphology and in expression of typical markers. The hDF showed typical spindle-shape morphology, whereas morphology of hEK can be described by a cobblestone morphology. The cells were characterized by immunohistochemical staining (Figure 1) before using them for ftSE. The hDF are positive for vimentin (Figure 1A), a marker for fibroblasts. Primary hEK highly express early differentiation protein cytokeratin-14 (Figure 1B) but nearly no late keratinocyte differentiation protein cytokeratin-10 (Figure 1C).
Following primary culture, we detached the cells from the cell culture flasks and used them for the generation of ftSE. The ftSE were cultured for 7 days under submers medium conditions (Figure 2A) followed by a 14 days culture at the air-liquid interface (Figure 2B). During this process, the ftSE contract significantly. Within 21 days the hEK are proliferating (Figure 2C-E). By changing the medium level in such way that the surface of the models is exposed to air and by increasing calcium concentration, the hEK are stimulated to differentiate into a multi-cellular epidermis that is composed of several vital cell layers of keratinocytes in different differentiation states and a stratum corneum (Figure 2E).
Comparing the Haematoxylin & Eosin staining of ftSE (Figure 2E) with native human skin (Figure 2F) it becomes obvious that ftSE mimics histological architecture of the skin. This can be further demonstrated by immunohistochemical stainings (Figure 3). The hDF in the collagen I hydrogel can be stained with primary antibodies against vimentin (Figure 3A, E). Keratinocytes form an epidermal layer composed of cytokeratin-14 positive cells (Figure 3B, F) and the late differentiation marker cytokeratin-10 is expressed in the supra-basal layers (Figure 3C, G). Furthermore, the stratum corneum is positive for filaggrin (Figure 3D, H).
Following 21 day culture period, the aWD was used to generate cutaneous wounds in the ftSE. As shown in Figure 4A, the aWD is constructed as a closed box-system to ensure sterile conditions. In Figure 4B, a schematic drawing of the wounding process is demonstrated. The ftSE are fixed on a sample carrier plate. The whole wounding process, including the setting of parameters such as lowering speed and revolutions of the drill head (Figure 4D) as well as the depth of penetration is controlled computer assisted. The wounding procedure can be monitored optically via a front window or by a high resolution camera that records all steps during the wounding procedure. Size, shape and depth of the wound injury can be adjusted individually (Figure 4C).
Figure 1. Characterization of cells used for the construction of full thickness skin equivalents. Immunohistochemical staining of primary human dermal fibroblasts for vimentin (A) and of epidermal keratinocytes for the early differentiation keratin filament cytokeratin-14 (B) and the late differentiation filament cytokeratin-10 (C). Positive staining is shown by brown color. Scale bars indicate 100 μm. Please click here to view a larger version of this figure.
Figure 2. Changes of culture conditions and maturation of full thickness skin equivalents. Macroscopic pictures of full thickness skin equivalents cultured for 7 days under submers conditions (A) and 14 days at the air-liquid interface (B). Comparison of Haematoxylin & Eosin staining of full thickness skin equivalents in different developmental stages: After 7 days (C), 14 days (7 days at the air-liquid interface, D) and 21 days (14 days at the air-liquid interface, E). Haematoxylin & Eosin staining of native human skin (F). Scale bars indicate 100 μm. Please click here to view a larger version of this figure.
Figure 3. Characterization of full thickness skin equivalents. Comparison of immunohistochemical staining of full thickness skin equivalents (A–D) and native human skin (E–H) for vimentin (A, E), cytokeratin-14 (B, F), cytokeratin-10 (C, G) and filaggrin (D, H). Scale bars indicate 100 μm. Please click here to view a larger version of this figure.
Figure 4. Automatic Wounding Device for controlled wounding on full-thickness skin equivalents. The automated wounding device (aWD) provides sterile conditions during the controlled wounding procedure (A). An attached computer (CPU) controls the aWD. The full thickness skin equivalents (ftSEs) are placed on the sample carrier plate (SCP), deviations in sample morphology are equalized by a light barrier (LB), which triggers the wounding procedure. The wound shape depends on the used drilling head (DH). The diameter ranges from 1 mm (C, upper image) to 2 mm (C, lower image). The wounding procedure is monitored and can be recorded for documentation and quality management (D). Hematoxylin & Eosin stained cross section of a full thickness skin equivalent wounded with the automated wounding device using a 1 mm drilling head (E). All scale bars indicate 2 mm. Please click here to view a larger version of this figure.
In vitro cells are usually expanded in two-dimensional cell cultures, in which cells adhere to plastic surfaces. However, these culture conditions do not reflect the physiological three-dimensional conditions in which cells grow in vivo. Under three-dimensional conditions, cells can form natural cell-cell and cell-matrix attachments and migrate in three-dimensions. Especially in the cutaneous wound healing the resemblance of the in vivo situation is pivotal to generate meaningful data, as cell migration and matrix generation are key elements of wound healing.
In order to mimic these conditions we developed ftSE that are composed of human primary cells and reflect both the dermal and the epidermal layer of the skin. During two weeks of culture at the air-liquid interface, the hEK form an epidermis composed of several layers of keratinocytes in different differentiation phases and a stratum corneum. Furthermore, the dermal layer is composed of hDF, which are embedded in a collagen I hydrogel. During the culture time, the hDF remodel the dermal component, which results in a significant shrinkage of the models (Figure 2B).
To prepare reproducible and standardized wounds, we developed an aWD. The computer controlled machine can set defined and precise cutaneous injuries (Figure 4). Depending on the intended research, depth, shape and size of the injury can be adapted. However, due to the technical specifications of the aWD and the morphology of the ftSE, it is recommended to set a minimal penetration of 100 µm reproducibility. The aWD is operating under sterile conditions in a closed box-system and all components can be sterilized by autoclaving, disinfection solutions or ultraviolet light. In preparation for the wounding procedure the samples need to be transferred to a sample carrier plate. This plate facilitates an exact position of the samples. Furthermore the adhesion between sample and the sample carrier plate is sufficient to keep the samples in place during the wounding procedure. The wounding procedure itself can be monitored optically via a front window or by a high resolution camera that records all steps from a close-up position. Unlike other systems that use lasers to excavate tissue to generate a wound, the aWD employs a rotating drill. Thus, no heat is applied during the procedure, which is a vital perquisite to investigate mechanical wounds. Additionally, specific shape of the drilling head ensures a material removal from the wound channel. Using empirically determined wounding parameters mentioned above the wounding procedure can be adjusted to meet specific physiological properties of either the ftSE or native skin samples. This way any undesirable side effects like unintentional detachment of epidermal layers can be minimized and reproducible wounding performed with a success rate of 90%.
This study demonstrates that the developed ftSE partly mimic the in vivo situation of skin accurately and thus can be employed as a model for wound healing studies. Though important aspects of the wound healing process such as the effect of the vasculature, the blood clot and the immune system are lacking in the model, the epidermal-mesenchymal and cell-matrix interaction is recapitulated in a three-dimensional highly standardized environment. In addition, we could show that the aWD generates standardized wounds in skin models. In combination, both technologies can be used to investigate wound healing and the effect of substances and drugs that support the healing process. In order to mimic the in vivo situation more closely, the ftSE can be extended by other cells such as melanocytes, skin tumor cells or micro-vascular endothelial cells.
The authors have nothing to disclose.
The authors thank the Fraunhofer ISC for the collaboration concerning the construction of the automated wounding device. The project was founded by Fraunhofer internal project “Märkte von Übermorgen” (SkinHeal).
Name of Material/ Equipment | Company | Catalog Number | Comments/Description |
Trypsin EDTA (1:250) 0.5 % in DPBS | PAA | L11-003 | 0.05% |
Dulbecco’s Phosphate Buffered Saline | Sigma | D8537 | |
Collagen (6 mg/ml in 0,1 % acetic acid) | Produced in house | ||
Fibronectin Human Protein, Plasma (50 µg/ml) | Life Technologies | 33016-015 | |
Inserts | Nunc | 140627 | |
6 well plate | Nunc | 140685 | |
24 well plate | Nunc | 142485 | |
Microscope slides | R. Langenbrinck | 03-0070 | |
Fibroblasts culturing (500 ml): | |||
DMEM, high glucose | Life Technologies | 11965-092 | 89% (445 ml) |
Fetal bovine serum | Bio & Sell | FCS.ADD.0500 | 10% (50 ml) |
Penicillin-Streptomycin (10,000 U/mL) | Life Technologies | 15140-122 | 1% (5 ml) |
Keratinozyten culture medium (500 ml): | |||
Keratinocyte Growth Medium 2 | Promocell | C-20111 | 89% (445 ml) |
Keratinocyte Growth Medium 2 SupplementPack | Promocell | C-39011 | |
Penicillin-Streptomycin (10,000 U/mL) | Life Technologies | 15140-122 | 1% (5 ml) |
Gel neutralization solution (250 ml): | |||
Dulbecco’s Modified Eagle Medium, high Glucose Powder with L-Glutamine | PAA | G0001,3010 | 93% (232,5 ml) |
Chondroitin sulfate sodium salt from shark cartilage | Sigma | C4384-1g | 1% (2,5 ml) |
Fetal bovine serum | Bio & Sell | FCS.ADD.0500 | 3% (7,5 ml) |
HEPES | Sigma | H3375-1kg | 3% (7,5 ml) |
Skin model submers medium (500ml): | |||
Keratinocyte Growth Medium 2 | Promocell | C-20111 | |
Keratinocyte Growth Medium 2 SupplementPack | Promocell | C-39011 | |
Fetal calf serum | Bio & Sell | FCS.ADD.0500 | 5%-2% (25 ml-10 ml) |
Penicillin-Streptomycin (10,000 U/mL) | Life Technologies | 15140-122 | 1% (5 ml) |
Skin model air-liquid interface medium (500 ml): | |||
Keratinocyte Growth Medium 2 | |||
Keratinocyte Growth Medium 2 Supplement Pack | Promocell | C-39011 | adding only supplements: Insulin, Hydrocortisone, Epinephrine, Transferrin, CaCl2 |
Penicillin-Streptomycin (10,000 U/mL) | Life Technologies | 15140-122 | 1% (5 ml) |
CaCl2 (300 mM) | Sigma | C7902-500g | 0,62% (3,1 ml) |
Histology: | |||
IHC-Kit DCS SuperVision 2 HRP | DCS | PD000KIT | |
Vimentin antibody | Abcam | ab92547 | |
CK14 antibody | Sigma | HPA023040-100µl | |
CK10 antibody | Dako | M7002 | |
Filaggrin antibody | Abcam | ab81468 | |
H&E staining | |||
Mayer´s Haemalaun | AppliChem | A0884,2500 | |
Xylol | Sigma Aldrich | 296325-4X2L | |
Ethanol | Sigma Aldrich | 32205-4X2.5L | |
HCl | Sigma Aldrich | H1758-500ML | |
Eosin | Sigma Aldrich | E4009-5G | |
2-Propanolol | Sigma Aldrich | I9516-500ML | |
Mounting Medium | Sigma Aldrich | M1289-10ML |