Organotypic culture of human amnion cells in air-liquid interface as a potential substitute for skin regeneration
SIMAT SITI FATIMAH1,2, KIENHUI CHUA3, GEOK CHIN TAN1,
TENGKU IBRAHIM AZMI4, AY EENG TAN5& HAYATI ABDUL RAHMAN1,6
1Department of Pathology and5Department of Obstetrics & Gynecology, Faculty of Medicine, Universiti Kebangsaan Malaysia, Jalan Yaacob Latif, Bandar Tun Razak, Cheras, Kuala Lumpur, Malaysia,2Biotechnology Research Institute, Universiti Malaysia Sabah, Jalan UMS, Kota Kinabalu, Sabah, Malaysia,3Department of Physiology, Faculty of Medicine, Universiti Kebangsaan Malaysia, Jalan Raja Muda Abdul Aziz, Kuala Lumpur, Malaysia,
4Department of Veterinary Preclinical Sciences, Faculty of Veterinary Medicine, Universiti Putra Malaysia, UPM Serdang, Selangor Darul Ehsan, Malaysia, and6Department of Medical Sciences II, Faculty of Medicine & Health Sciences, Universiti Sains Islam Malaysia, Jalan Pandan Utama, Kuala Lumpur, Malaysia
Abstract
Background aims.The aim of the present study was to evaluate the effects of air-liquid interface on the differentiation po- tential of human amnion epithelial cells (HAECs) to skin-like substitute in organotypic culture.Methods.HAECs at passage 1e2 were seeded onto afibrin layer populated with human amnion mesenchymal cells to form the organotypic cultures. The organotypic HAECs were then cultured for 7, 14 and 21 d in two types of culture system: the submerged culture and the air- liquid interface culture. Cell morphogenesis was examined under the light and electron microscopes (transmission and scanning) and analyzed by immunohistochemistry.Results.Organotypic HAECs formed a single layer epithelium after 3 wk in submerged as well as air-liquid interface cultures. Ultrastructurally, desmosomes were observed in organotypic HAECs cultured in the air-liquid interface but not in the submerged culture. The presence of desmosomes marked the onset of early epidermal differentiation. Organotypic HAECs were positive against anti-CK18 and anti-CK14 in both the submerged and the air-liquid interface cultures. The co-expression of CK14 and CK18 suggested that differentiation of HAECs into skin may follow the process of embryonic skin development. However, weak expression of CK14 was observed after 2 and 3 wk of culture in air-liquid interface. CK10, involucrin, type IV collagen and laminin-5 expression was absent in organotypic HAECs. This observation reflects the initial process of embryonic epidermal differentiation and stratification.Conclusions.
Results from the present study suggest that the air-liquid interface could stimulate early differentiation of organotypic HAECs to epidermal cells, with a potential use for skin regeneration.
Key Words: air-liquid interface, epithelial stem cells,fibrin, human amnionederived stem cells, organotypic culture, skin regeneration
Introduction
Skin that is seriously injured, for example, by burns, trauma or chronic ulcers, requires immediate replacement to facilitate repair to restore function.
Autologous skin grafting is the gold standard for skin replacement whereby a normal skin area is separated and then transplanted onto the recipient area, where it receives a new blood supply from the same indi- vidual. However, these autograft skin substitutes are limited in terms of their availability. At the same time, an injury is inflicted onto the donor part of the patient. Ex vivo-cultured autologous keratinocytes could be an alternative choice. However, a long
duration of time (2e3 wk) is required to grow confluent epithelial sheetsin vitro, during which the patient would be exposed to potential infection (1).
On the other hand, the development of allogeneic skin substitutes is still limited by the availability of suitable source of cells, the problems with graft rejection, the possible transmission of diseases and the high cost of the procedure.
Human amnion epithelial cells (HAECs) have several remarkable features that make these cells attractive as an alternative source of epithelial stem cells. Benefits to skin tissue engineering with HAECs
Correspondence: Prof DrHayati Abdul Rahman, MBChB, DCP, FAM, Department of Pathology, Faculty of Medicine, Universiti Kebangsaan Malaysia, Jalan Yaacob Latif, Bandar Tun Razak, Cheras, 56000 Kuala Lumpur, Malaysia. E-mail:[email protected]
Cytotherapy, 2013; 15: 1030e1041
include (i) stem cellelike characteristics with multi- lineage differentiation potential(2e5); (ii) high self- renewal capacity that does not form tumors on transplantation (6,7); (iii) immunoprivileged char- acteristics that may reduce the risks of rejection on transplantation (6,7); (iv) abundance, easy access and ready availability from healthy mothers as the cells are discarded following delivery; (v) ease of isolation with high yield through the use of enzymatic techniques (3); (vi) noninvasive procedures during harvesting that are ethically free from controversial issues associated with clinical use of the cells. Both HAECs and fetal epidermis are derived from the embryonic ectoderm that expresses K18 (8). At embryonic day 9.5 (E9.5), which is before the onset of stratification, the embryonic epidermis also ex- presses K5, K14 and p63. These proteins are also expressed in the basal layer of adult epidermis (8e10). At E10.5, after thefirst stage of epidermal differentiation, the periderm is formed(11). Further epidermal maturation takes place between E15.5 and E18.5. At E15.5 of maturation, an intermediate layer is formed between the basal layer and the periderm, which corresponds to the adult spinous layer and is marked by K1 and K10 expression (12). At E16.5 and E17.5, the formation of granular layer and stratum corneum create a water-impermeable barrier in which involucrin, loricrin andfilaggrin (cornified cell envelope) arefirst expressed(12e14). At E18.5, when the epidermis is fully stratified, the periderm is shed off. Thus, HAECs are a promising source of cells for skin regeneration.
The air-liquid interface was developed to produce skin cell culture sheets for transplantation. It en- hances epidermal cell differentiation, stratification and cornification with intercellular lamellar body deposition, thus displaying strong similarity to skin tissue (15e17). The air-liquid interface creates a microenvironment very similar to skin microenvi- ronmentin vivo. In view of the fact that HAECs have stem cellelike characteristics and have the potential to differentiate into skin, we therefore used the air-liquid interface approach to study the differentiation po- tential of HAECs to skin epidermal-like cells. Fibrin was used as the scaffold and seeded with human amnion mesenchymal cells (HAMCs) to form the dermal layer. The construction of the organotypic HAECs is a novel approach, which makes this study very different from other previous studies(18). Thus, the aim of the present study was to evaluate the effects of the air-liquid interface culture system on HAEC differentiation toward epidermal cells. These orga- notypic HAECs were also cultured in a submerged system for comparison. Organotypic keratinocyte seeded ontofibrin-populated dermalfibroblasts was used as a positive control for the organotypic HAECs
under the same culture conditions. Histological and immunohistochemical analyses and ultrastructural studies with the use of scanning and transmission electron microscopes were used in this study.
Methods
This study was approved by the Universiti Kebang- saan Malaysia Research and Ethics Committee (Approval Project Code: FF-279-2009). Written consent was obtained from donors before collection of human placentas. Human term placentas were obtained from uncomplicated, elective caesarean sections on healthy mothers of ages ranging from 20e40 years. Under the same ethical approval, dis- carded circumcised human foreskins were obtained from male children ranging from 8e12 years of age, with written consent from their parents or guardians.
Isolation and culture of amnion-derived cells
The amnion layer was mechanically peeled off the chorion and washed several times with phosphate- buffered saline (PBS, Gibco-Invitrogen, Grand Island, NY, USA) to remove red blood cells. It was then cut into 22-cm pieces and processed as pre- viously described(3). In brief, to release HAECs, 20 pieces of human amnion (22 cm) were digested in 10 mL of 0.05% trypsin containing 0.2 g/L of ethylene diamine tetraacetic acid (EDTA) (Gibco-Invitrogen) and incubated at 37C with constant agitation.
Trypsin was inactivated by addition of 10 mL of F12:Dulbecco’s modified Eagle’s medium (DMEM) (1:1) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco-Invitrogen; FD þ 10%
FBS). The suspension from thefirst 10 min was dis- carded to exclude cell debris. The cells from the second and third 30 min of digestion were collected and washed with PBS. HAECs were then plated on six-well plates at 20,000 cells/cm2seeding density, in culture medium F12:DMEM (1:1) supplemented with 10% FBS, 10 ng/mL epidermal growth factor (Peprotech, USA), 1% antibiotic-antimycotic (Gibco-Invitrogen), 1% Glutamax (Gibco-Invitrogen) and 1% vitamin C (Merck, Darmstadt, Germany).
The culture plates were incubated in a humidified incubator at 37C with 5% CO2. HAECs from the initial culture (P0) were subcultured to passage 1 (P1) with a split ratio of 1:3 once the culture reached 80e90% confluence and then subcultured to subse- quent passages with a split ratio of 1:2.
The denuded amnion was washed with PBS (Gibco-Invitrogen) and subjected to 0.3% collagenase type I (Worthington, Lakewood, NJ, USA) digestion for 4 h at 37C with constant agitation. The cell sus- pension was then centrifuged at 600gfor 5 min at room
temperature. The resulting cell pellet was washed two to three times before the cells were cultured. The HAMCs were cultured in Ham F12: DMEM (1:1,1, Gibco-Invitrogen) supplemented with 10%
FBS (Gibco-Invitrogen), 1% antibiotic-antimycotic (Gibco-Invitrogen), 1% glutamine (Gibco-Invitrogen) and 1% vitamin C (Merck) in a humidified incubator of 37C with 5% CO2. The culture medium was changed every 3 d. The primary culture of HAMCs (passage 0, P0) was subcultured once the culture reached 80e90% confluence with a split ratio of 1:4, with the use of 0.05% trypsin-EDTA. These cells were subsequently passaged until passage 5 (P5), when the cultures reached 80e90% confluence and cultured under the same conditions described above.
Isolation and culture of skin keratinocytes and dermal fibroblasts
The circumcised foreskin obtained was rinsed with 70% isopropanol and then placed in PBS containing 1% antibiotic-antimycotic (Gibco-Invitrogen). The skin was then cut into small pieces (approximately 2 mm2) before digestion in 0.3% collagenase type I (Worthington) overnight in an incubator shaker at 37C. The cell suspension was then centrifuged at 600gfor 5 min at room temperature and washed with PBS (1); 10 mL of 0.05% trypsin-EDTA was added to the cell pellet and the epidermis. These mixtures were placed in the incubator for 5 min at 37C with constant agitation to isolate skin kerati- nocytes; 10 mL of trypsin inhibitor (Cascade Bi- ologics, Gibco-Invitrogen) with a concentration of 10 mg/mL was added to inhibit the trypsin-EDTA activity before the mixtures werefiltered with the use of a cell strainer (BD, Bioscience, San Jose, CA, USA) to separate the undigested dermis and epi- dermis. The cell suspension, which contained skin keratinocytes and dermalfibroblasts, was centrifuged and washed several times before these cells were counted with the use of a hemocytometer and trypan blue exclusion dye (Gibco-Invitrogen). The cells were then cultured in a media containing Epilife:Ham F12:DMEM (2:1:1) (Cascade Biologics Gibco-Invitrogen) supplemented with 5% FBS with a cell density of 2e3 105 cells per well in a six- well plate (Nunc, Langenselbold, Germany). Epilife medium (Cascade Biologics Gibco-Invitrogen) is a serum-free medium that supports the growth of keratinocytes that have been supplemented with Human Keratinocyte Growth Supplement (HKGS, Cascade Biologics, Gibco-Invitrogen) and 1%
antibiotic-antimycotic (Gibco-Invitrogen).
Dermal fibroblasts were isolated once the co- culture of keratinocyte-fibroblasts reached a conflu- ence of 80e90%, with the use of 2 mL of 0.05%
trypsin-EDTA for 5 min at 37C. The trypsin-EDTA activity was inhibited by adding 2 mL of trypsin in- hibitor (concentration of 10 mg/mL). Cell suspen- sions were centrifuged, washed and counted with the use of a hemocytometer and trypan blue exclusion dye. The isolated cells (dermal fibroblasts) were plated in a six-well plate with a cell density of 1105 cells per well. Thefibroblasts were then cultured in a growth media Ham F12:DMEM (1:1) supplemented with 10% FBS, 1% antibiotic-antimycotic (Gibco- Invitrogen), 1% L-glutamine (Gibco-Invitrogen) (FDþ10% FBS). Thefibroblasts were subcultured with a split ratio of 1:4 on attaining confluence.
After dermal fibroblast isolation, the remaining cells in the six-well plate were washed with PBS (1).
These cells were skin keratinocytes and they were cultured in Epilife medium supplemented with HKGS and 1% antibiotic-antimycotic (Gibco-Invitrogen).
Keratinocytes were subcultured once they reached 80e90% confluence with a split ratio of 1:3.
Preparation of human fibrin as a biomaterial
Whole blood was drawn from a human healthy donor into 9-mL plasma collection tubes containing 3.2%
sodium citrate (Bio-One, Greiner, Frickenhausen, Germany). The blood sample was centrifuged at 600gfor 5 min to collect the plasma. The plasma was filtered with the use of a 0.2-
m
m sterile syringefilter (Sartorius AG, Goettingen, Germany) to remove the cell debris to avoid clotting of the plasma. The plasma was kept at 20C.Organotypic culture of human amnion-derived cells HAECs at passages 1 and 2 and HAMCs at passage 5 were used in organotypic culture, whereas skin keratinocytes at passages 1 to 4 and dermal fibro- blasts at passages 3e6 were used as the control.
Organotypic culture of HAECs was achieved with the use of a combination of cell culture techniques, which was a modification of the method described by Bell et al.1983(19).
HAMCs and dermal fibroblasts (control) were trypsinized and mixed with 1 mL of human plasma with a cell density of 5 105cells/mL and plated into a six-well plate culture insert (polyethylene terephthalate, 0.4-
m
m pore size, BD Falcon, NJ, USA); 100m
L of aprotinin (Merck) at a concentra- tion of 2 mg/mL was added to the admixture of plasma-HAMCs and plasma-fibroblasts to reduce the degradation rate of thefibrin-HAMCs andfibrin- fibroblasts in the culture. Calcium chloride (1 mol/L, Merck) was added to the admixture of plasma- HAMCs and plasma-fibroblasts at a volume ratio of 1:40 (CaCl2: admixture plasma-HAMCs andplasma-fibroblasts) to initiate the fibrin polymeriza- tion. After the fibrin-HAMCs and fibrin-fibroblast formation, the samples were cultured overnight in a medium containing F12:DMEM, (1:1) supple- mented with 10% FBS (FDþ 10% FBS). HAECs and keratinocytes were trypsinized and seeded onto the surface of the fibrin-HAMCs and fibrin- fibroblasts, respectively. The substitutes formed with HAECs and keratinocytes were cultured in a media Epilife:F12:DMEM (2:1:1) þ5% FBS overnight in a submerged culture before they were exposed to the air-liquid interface. Epilife used here was pre-supplemented with HKGS and 1% antibiotic- antimycotic. Six organotypic cultures remained in submerged cultures and six of the others were cultured in the air-liquid interface. Substitutes in submerged and air-liquid interface were cultured for 7, 14 and 21 d; they were fixed in 10% formalin (Merck) for light microscopy and 2.5% or 4%
glutaraldehyde (Merck) for electron microscopy.
Histological analysis
After fixation in 10% formalin for 24 h, the tissues were processed for paraffin embedding. Three- to 4-
m
m-thick tissue sections were then cut, deparaffi- nized and stained with hematoxylin and eosin.Stained sections were examined under the light microscope.
Immunohistochemical analysis
Three- to -micrometer-thick sections of the tissues were deparaffinized for immunohistochemical anal- ysis according to the manufacturer’s protocol, DAKO REAL EnVision Detection System (DAKO, Glostrup, DA, Denmark). Tissue sections were pre- treated with proteinase K (DAKO) at 37C for 6 min or cooked in citrate buffer at pH 6.0 (DAKO) or Tris-EDTA buffer at pH 9.0 (Lab Vision, Fremont, CA, USA). The tissue sections were then cooled and rinsed three times with tris-buffered saline (TBS, Lab Vision). Sections were then incubated in 0.3%
hydrogen peroxide for 6 min to inhibit the endoge- nous peroxidase activity, followed by rinsing in TBS.
The samples were then incubated in each of the following primary mouse anti-human monoclonal antibodies: anti-CK10 (1:200, DE-K10, Lab Vision), anti-CK14 (1:200, LL002, Lab Vision), anti-CK18 (1:50, DC 10, DAKO), anti-p63 (1:200, 4A4 þY4A3, Lab Vision), anti-collagen type IV (1:33.3, CIV 22, DAKO), anti-involucrin (1:200, SY-5, Lab Vision) and antielaminin 5 (1:37.5, 4G1, DAKO); for 30 min at room temperature. Samples were then rinsed with TBS and incubated for 30 min with horseradish peroxidaseelabeled with polymer
goat secondary antibody molecules against mouse immunoglobulin (DAKO). To develop the signal, 3,30-diaminobenzidine (DAKO) was added, fol- lowed by hematoxylin for nuclear counterstaining.
Scanning electron microscopy
The samples were fixed in 4% glutaraldehyde in PBS for 24 h at 4C, washed with PBS (1, pH 7.2) and dehydrated stepwise in a series of ethanol (30e100%). The samples were then dried with the use of a critical point dryer (Baltec CPD030, Switzerland), mounted onto the stubs before sputtered-coating with gold palladium with the use of the Sputter Polaran E-5100 SEM Coating Coater. Samples were examined with the use of a scanning electron microscope (FE-SEM) LEO SUPRA 55VP (Germany) and (SEM) Philips XL30 (The Netherlands).
Transmission electron microscopy
The samples were diced into small pieces (1 mm2) andfixed in 2.5% glutaraldehyde in PBS for 24 h at 4C. Samples were then washed with PBS (1, pH 7.2) before post-fixation in 1% osmium tetroxide for 1e2 h at room temperature. Samples were then washed three times in distilled water before block staining with 0.5% uranyl acetate for 1 h at room temperature. Samples were then washed with distilled water and dehydrated stepwise in a series of ethanol (25e100%), followed by 100% propylene oxide (electron microscope grade, EMS, USA). Samples were subsequently infiltrated with a mixture of resin and propylene oxide at the respective ratio of 1:2 for 30 min, 1:1 for 1 h, 2:1 for 2-4 h and 100% resin overnight, polymerized in an oven for 24 h at 60C.
The Agar 100 Resin Kit was used in this study (Agar Scientific, UK). Ultrathin sections obtained with the use of the ultramicrotome (LEICA ULTRACUT, UCT, Austria) were placed onto a G200-Cu Glider copper grid (Agar Scientific). These ultrathin sections were stained with lead/uranyl acetate and examined under the transmission electron microscope, LEO 912AB ENERGY FILTER TEM (Germany).
Results
Morphological and histological analysis
Organotypic HAECs remained as a single epithelial layer after 1, 2 and 3 wk in both the submerged and the air-liquid interface cultures as shown inFigure 1.
HAECs that adhered to thefibrin matrix were either cuboidal or squamous in shape. As for the organo- typic keratinocytes, which was the positive control, it
remained as one to three layers throughout the 3 wk in the submerged culture (Figure 1A,E,I). However, in the air-liquid interface culture, a remarkably well-stratified epithelial layer consisting of basal, spinous, granular and cornified strata was formed (Figure 1B,F,J). The organotypic keratinocytes formed one to three layers of cells after 1 wk, followed by the formation of rete ridges and keratin layers after 2 wk in the air-liquid interface culture (Figure 1F).
Subsequently, the epidermis displayed four to eight layers of stratified keratinocytes with progressive flattening of the cells from the basal layer toward the surface (Figure 1F). Stratification was further enhanced after 3 wk, forming six to 15 layers of keratinocytes (Figure 1J).
Detection of cytokeratin expression, epithelial differentiation markers and basement membrane components by immunohistochemical analysis
CK18 was observed throughout the entire organo- typic HAECs in both the submerged and the air- liquid interface throughout the 3 wk culture but was negative in organotypic keratinocytes. CK14 was expressed in HAECs throughout the 3 wk in sub- merged culture. In the air-liquid interface, however, HAECs only partially expressed CK14 after 2 and 3 wk. CK14 was homogenously positive throughout the layers of the organotypic keratinocytes in both the
submerged and the air-liquid interface cultures.
Organotypic HAECs showed negative staining against p63. However, p63 was detected in organo- typic keratinocytes in the submerged and the air- liquid interface culture. In the submerged culture, p63 was detected throughout the epithelium. A different pattern was observed in the air-liquid interface culture, where p63 was detected in all the cell layers after 1 wk of culture but was more local- ized to the basal and suprabasal layers of the orga- notypic keratinocytes after 2 and 3 wk (Figure 2, Supplementary Figure 1, Supplementary Figure 2, Table I).
In the present study, organotypic HAECs were characterized through the use of epidermal differen- tiation markers such as CK10 and involucrin. Both organotypic HAECs and keratinocytes showed negative staining against CK10 in both the sub- merged and the air-liquid interface. Involucrin was also not detected in HAECs, but it was detected at the upper basal layer of organotypic keratinocytes for both the submerged and the air-liquid interface after 1, 2 and 3 wk of culture (Figure 2, Supplementary Figure 1,Supplementary Figure 2,Table I).
Expression of basement membrane proteins, such as collagen type IV and laminin-5, at the junc- tional zone between the basal layer of epithelial cells and fibrin matrix, was also determined. Collagen type IV was negative in the basement membrane
Figure 1. Hematoxylin and eosin staining of organotypic keratinocytes and organotypic HAECs after 1, 2 and 3 wk in the submerged culture and the air-liquid interface culture. Magnification200 (AeI, K, L);400 (J).
of organotypic HAECs after 1, 2 and 3 wk in sub- merged culture but, in the air-liquid interface cul- ture, there was weak positive staining only after 3 wk.
However, HAMCs stained positively for collagen type IV. On the other hand, collagen type IV was negative in the organotypic keratinocytes after 1 wk in both the submerged and the air-liquid interface.
After 2 and 3 wk, the organotypic keratinocytes in submerged culture showed weak positive staining,
whereas those in the air-liquid interface increased in their staining intensity, including the dermal fibro- blasts. Laminin-5 expression in organotypic HAECs was negative after 1, 2 and 3 wk of culture in both the submerged and the air-liquid interface. In the orga- notypic keratinocytes, weak, heterogenous staining were seen in both cultures after 1 wk, with increased intensity being observed in the air-liquid interface after a prolonged culture. Laminin-5 stained the
Figure 2. Immunohistochemistry of cytokeratin 18 (CK18; AeD), cytokeratin 14 (CK14; EeH), p63 (IeL), involucrin (Inv; MeP), collagen type IV (Col IV; QeT) and laminin-5 (UeX) for organotypic keratinocytes and organotypic HAECs after 3 wk in the submerged culture and the air-liquid interface culture. Magnification200 (AeE, GeI, KeM, OeQ, SeU, WeX );400 (F, J, N, R, V).
basement membrane as well as the basal cells of the keratinocyte layer (Figure 2, Supplementary Figure 1, Supplementary Figure 2,Table I).
Ultrastructural analysis with the use of electron microscopy analysis
Scanning electron microscopy. Scanning electron mi- croscopy showed squamous epithelial cells for both organotypic HAECs and keratinocytes. Both cell types showed the presence of microvilli when cultured in the submerged system. The microvilli flattened and disappeared when these cells were exposed to the air- liquid interface culture (Figure 3AeD). SEM also showed the spindle-shaped HAMCs and dermal fibroblasts embedded in the fibrin scaffold (Figure 3E,F).
Transmission electron microscopy (TEM).After 3 wk in submerged culture, organotypic HAECs were squa- mous in shape with deeply indented nuclei (Figure 4A). Within the cytoplasm, rough endo- plasmic reticulum, Golgi apparatus, ribosomes, lipid droplets and multivesicular bodies were distinct.
Keratin filaments were also observed in the cell cytoplasm (Figure 4A,B). The cell surface had numerous microvilli (Figure 4A,B). Along the cell membrane were numerous invaginations, and within the vicinity of these invaginations were numerous vesicles (Figure 4B). Adherent tight junctions were also present between the adjacent cells (Figure 4C);
however, no desmosomes were observed.
Similar characteristics were also observed with organotypic HAECs cultured in the air-liquid inter- face after 3 wk, except for some differences in which
organotypic HAECs showed a more squamous or polygonal shape (Figure 4D). The indentation of the cell nucleus was less pronounced. They were devoid of microvilli either at the free cell surface or contig- uous with the fibrin layer (Figure 4E). Similarly, invaginations of the cell membrane and vesicles were also absent. Desmosomes were observed between adjacent cell membranes (Figure 4F).
After 3 wk in submerged culture, organotypic keratinocytes showedflattened, layered keratinocytes with apical microvilli (Figure 5A). These cells were in contact with each other intercellularly through desmosomes (Figure 5A,C). These cells contained keratin filaments in the cytoplasm that were irregu- larly arranged, and the superficial cells contained a few keratohyaline granules (Figure 5A,C). Hemi- desmosomes were observed along the interface be- tween the basal cells and the dermal substitute (Figure 5B). Thinfibrils were also present along the basement membrane (Figure 5B).
In organotypic keratinocytes cultured in the air- liquid interface, no microvilli were present on the apical surface (Figure 5F). The basal cells were columnar or cuboidal, and subsequent layers showed progressive flattening of the epithelial cells (Figure 5DeF). These cells were connected inter- cellularly by desmosomes. Mitochondria were highly abundant in the cytoplasm of the basal cells and the suprabasal cells (Figure 5DeF). The incidence of mitochondria decreased as the cells became more differentiated to become corneocytes. Only keratino- cytes at the stratum granulosum contained kerato- hyaline granules in the cytoplasm (Figure 5F). The plasma membranes of the basal cells at the epidermal- dermal junctions were convoluted with frequent, more
Table I. Expression of differentiation proteins in organotypic HAECs and keratinocytes after 1, 2 and 3 weeks cultured in submerged and air-liquid interface.
CK18 CK14 p63 Involucrin CK10 Collagen IV Laminin-5
Submerged
KC 1st wk e Eb Eb USL e e BLa
2nd wk e Eb Eb USL e BMa BLa
3rd wk e Eb Eb USL e BMa BLa
HAECs 1st wk Eb E e e e e(but HAMCsb) e
2nd wk Eb E e e e e(but HAMCsb) e
3rd wk Eb E e e e e(but HAMCsb) e
Air-liquid interface
KC 1st wk e Eb Eb USL e e BLa
2nd wk e Eb BL/SLb USL e BM/D BLa
3rd wk e Eb BL/SLb USL e BM/D BLa
HAECs 1st wk Eb E e e e e(but HAMCsb) e
2nd wk Eb Ea e e e e(but HAMCsb) e
3rd wk Eb Ea e e e BMa(HAMCsb) e
KC, keratinocytes; BM, basement membrane; D, dermis; E, whole epithelium; BL, basal layer; SL, suprabasal layer; USL, upper suprabasal layer.
aWeak staining.
bStrong staining.
distinct hemidesmosomes (Figure 5H). However, no distinct lamina densa and lamina lucida were observed at the epidermal-dermal junction in the organotypic keratinocytes culture (Figure 5H).
Discussion
An air-liquid interface was developed to mimic the natural microenvironment of the skin that enabled the differentiation of organotypic HAECs into functional, specialized cells. It further supported and enhanced epidermal differentiation and mor- phogenesis of organotypic keratinocytes, which concurred with previous studies (15e17). In this study, the air-liquid interface could enhance skin keratinocytes stratification of up to 15 layers, with progressiveflattening of keratinocytes from the basal to the superficial layer, hence forming the keratin layers after 3 wk as compared with a submerged
culture. In the latter, the epidermis could only differentiate and stratify up to three layers. The ex- pressions of early- and late-stage epidermal markers were more distinct and organized similar to that of native skin when organotypic keratinocytes were brought to the air-liquid interface. The air-liquid interface culture also increased the basement mem- brane component protein expression in organotypic keratinocytes. The ultrastructural features of orga- notypic keratinocytes are similar to that of native human skin. Thus, organotypic keratinocytes cultured in the air-liquid interface in our culture system closely resembled native human skin(15e17).
The results of the present study showed that organotypic HAECs remained as a single layer of epithelium, as demonstrated by light and ultra- structural and immunostaining analyses. Scanning and transmission electron microscope analyses of organotypic HAECs in submerged or air-liquid
Figure 3. Scanning electron microscopy (SEM) of human amnion epithelial cells (HAECs) on organotypic HAECs (B, D) and keratinocytes (control) on organotypic keratinocytes (A, C) after 3 wk cultured in the submerged (A, B) and the air-liquid interface (C, D). Both cell types showed the presence of microvilli when cultured submerged, but the microvilliflattened and disappeared when exposed to the air-liquid interface. Mesenchymal cells in organotypic HAECs (F) and dermalfibroblasts in organotypic keratinocytes (E) were analyzed with the use of SEM after 3 wk in the submerged culture and the air-liquid interface culture.
interface culture revealed the ultrastructure of these cells to have resembled that of native human amnion (20,21). Such ultrastructural features include the presence of zonula occludens, numerous microvilli, Golgi complex and endoplasmic reticulum. These cells were active secretory cells with abundant cyto- plasmic vesicles and cell membrane invaginations of the cytoplasm, which were also typical features of HAECsin vivo(20,21). The secretory properties of organotypic HAECs could be beneficial in facili- tating wound closure if organotypic HAECs were used for skin regeneration. Previous studies have shown that amnion-derived cellular cytokine solution improves acute and chronic wound-healing by pro- moting the migration of keratinocytes andfibroblasts in wound-healing (22,23). Desmosomes were also observed in organotypic HAECs cultured in the air- liquid interface, which were absent in submerged culture. This also suggests that the air-liquid inter- face promotes the differentiation of organotypic HAECs to resemble that of native human amnion.
The presence of desmosomes in HAECs in the air- liquid interface lends support to the potential dif- ferentiation of human amnion to skin. As observed, native human amnion epithelium contains regions of stratified plague with evidence of cornification (or keratinization) and the co-expression of epidermal cytokeratin (CK1 and CK10) and simple epithelium cytokeratin (CK8 and CK18)(24). Because HAECs were derived from ectoderm, their differentiation to epidermal-like cells may be related to the develop- ment of embryonic skin. Furthermore, desmosomal
glycoproteins expressions were upregulated with the onset of epidermal stratification, and they may be involved in regulating epidermal differentiation dur- ing early embryogenesis(25). Thus, results from the present study suggest that the air-liquid interface culture could induce early differentiation of organo- typic HAECs to epidermal-like cells.
Immunostaining analysis revealed that organo- typic HAECs was positive for CK18, a marker for simple epithelium. CK18 was also expressed by a single-layered surface ectoderm. Additionally, HAECs highly expressed CK14 in submerged culture.
HAECs also highly expressed CK14 after 1 wk in the air-liquid interface. However, CK14 expression decreased gradually after 2 and 3 wk in the air-liquid interface. CK14 is a marker of stratified epithelium, expressed by cells at the basal layer. In the skin, CK14 expression decreases as the cells differentiate and migrate from the basal to suprabasal layer(26). Dur- ing embryonic skin development, expression of CK14 marks the onset of commitment of the surface ecto- derm to stratification when the developing epidermis is still single-layered, expressing CK18(8). Therefore, the gradual decrease in CK14 expression and consis- tent expression of CK18 in organotypic HAECs reflect the initial process of epidermal differentiation and stratification during embryogenesis(8).
In the present study, immunostaining of orga- notypic HAECs was also negative for p63; p63 is associated with the initiation of epithelial stratifica- tion and is expressed early during embryogenesis (27). The negative staining of HAECs for p63 could
Figure 4. Ultrastructure of organotypic HAECs after 3 wk cultured in the submerged (AeC) and the air-liquid interface (DeF) showed single-layered epithelium. Organotypic HAECs in submerged culture showed numerous microvilli, numerous cell membrane invaginations and numerous vesicles (B) and showed the presence of adherent tight junctions (C), but no desmosome was observed. Organotypic HAECs in the air-liquid interface have characteristics similar to those in the submerged culture, except for the lack of microvilli, cell membrane invagination and vesicles (E). Desmosome was observed in the air-liquid interface culture (F). D, desmosome; F,filament; GC, Golgi complex; Mt, mitochondria; MV, microvilli; N, nucleus; V, cytoplasmic vesicle; ATJ, adherent tight junction.
be due to the very low level of their gene expression for p63 that could not be detected by immuno- staining, as shown in the previous monolayer culture study (3). In addition, CK10, involucrin, type IV collagen and laminin-5 were also stained negative.
These late-stage markers of differentiation and cornification (involucrin and CK10) and basement membrane component proteins (type IV collagen and laminin-5) are usually expressed at later stages of embryonic skin development(12e14,28). Thus, the absence of these markers together with the co- expression of CK14 and CK18 and the presence of desmosomes of organotypic HAECs further confirmed that the air-liquid interface stimulated the early-stage differentiation of HAECs to epidermal- like cells compared with submerged culture.
Contrary to the presentfinding, Koikeet al.(18) proved that organotypic HAECs could differentiate and stratify up to 10 layers after 3 wk of culture in the air-liquid interface. They also showed that CK10 was positively expressed in the suprabasal layer of strati- fied HAECs, whereas CK19 (keratinocyte stem cell marker) was expressed throughout the entire
epithelium. Furthermore, occludin and zonula occluden-1 were expressed throughout the intercel- lular junctions of the epithelial layers (18). These discrepancies could be due to epigenetic factors caused by different culture conditions used. Differ- ences in conditions include the use of HAMCs, fibrin as biological scaffold for the organotypic cul- ture and serum with low calcium concentration in our culture system.
Previous studies reported that epithelial- mesenchymal interactions were crucial and re- quired for epidermal morphogenesis in vitro (29,30). In the present study, the absence of dermal fibroblasts in the organotypic keratinocytes did not enhance epidermal differentiation and stratification as compared with the presence offibroblasts in the air-liquid interface culture system, in which orga- notypic HAECs still remained as a single-layered epithelium in the absence of HAMCs (unpublished data). Hence, in organotypic HAECs, HAMCs could act as feeder layer cells by secreting intrinsic cytokines such as keratinocyte growth factor, which may promote growth and differentiation of HAECs
Figure 5. Ultrastructure of organotypic keratinocytes after 3 wk cultured in the submerged culture show three layers of keratinocytes (AeC) and in the air-liquid interface show up to 16 layers of keratinocytes, with cells at the stratum basale (D) and stratum spinosum (E) showing numerous mitochondria and intercellularly connected by desmosomes; stratum granulosum (F) shows keratohyaline granules and reduced mitochondria in the cytoplasm. Intercellular connections between cells at the stratum spinosum show desmosomes and keratohyaline granules (G). Dermal-epidermal junctions show more distinct hemidesmosomes (H). B, basal cell; D, desmosome; H, hemidesmosome; KF, keratin filament; KG, keratohyaline granule; Mt, mitochondria; MV, microvilli; N, nucleus; TF, thinfibril. Arrow indicates keratinfilament.
in a paracrine manner (31). The present investiga- tion also takes the advantage of HAMCs having stem cellelike characteristics that could differen- tiate to various types of cells while retaining the self- renewal capability with their angiogenic properties (32). HAMCs expressed angiogenic growth factors and could spontaneously differentiate into endo- thelial cells and form capillary-like structures in in vitromatrigel assay (33). The presence of HAMCs in the organotypic HAECs culture system is there- fore promising in their future clinical application, which could promote neovascularization and improve graft uptake of skin-like substitutes.
Human plasma provides a less costly autologous biomaterial supply of fibrin that can be easily ob- tained. It is important for wound-healing because of its properties. It has its roles in hemostasis, chemo- taxis and mitogenesis(34). Fibrin has been used in the construction of skin substitutes and retains the differentiation properties of keratinocytes (34e36).
Its clinical importance has been proven to improve wound-healing by enhancing graft uptake and to reduce transplantation rejection (35,36). Therefore, fibrin is a promising biomaterial for organotypic HAECs, which is applied in this study.
Therefore, we believe that the major contributions to the differences in the results obtained between our experiment (early-stage differentiation of HAECs) and that of Koikeet al.(18)(later-stage differentiation of HAECs) could be due to the constituents of the culture medium rather than the use offibrin and the presence of HAMCs embedded in fibrin, which are more beneficial to organotypic HAEC differentiation and the future clinical application. First, we designed a culture medium so that it could support the growth and differentiation of HAECs or skin keratinocytes and the growth of HAMCs or dermal fibroblasts in the air-liquid interface. The culture medium formu- lation was in accordance with that previously reported by Mazlyzamet al. (37). Although the calcium con- centration in this medium was not increased to 1.8 mmol/L, the medium was proven to support keratinocyte growth and differentiation(37). Second, the presence of serum (5% FBS) in the medium used could probably influence HAEC differentiation. Em- bryonic stem cells cultured in 5% serum and 10%
serum could differentiate to epidermal-like cells, with the expression of CK14(38,39). On the basis of the presentfinding, our culture system could support the growth of skin-derived cells and epidermal morpho- genesis but falls short of inducing the later stage in epidermal histogenesis of organotypic HAECs. Thus, the early differentiation stage of organotypic HAECs is attributed to the air-liquid interface, in contrast to the submerged culture of the same culture condition.
Optimization of the culture medium is hence required
to enable the late-stage differentiation of organotypic HAECs in air-liquid interface in the near future. To further assess their differentiation potential, ideally, the organotypic HAECs should be transplanted onto athymic mice, examined and evaluated. Wound- healing could also be monitored objectively in vivo through the use of 2D luminescence imaging of sur- rogate parameters, namely, pH and pO2(40,41).
Conclusions
In conclusion, the air-liquid interface culture could stimulate early differentiation of organotypic HAECs to epidermal histogenesis as compared with the submerged culture system. Organotypic HAECs cultured in the air-liquid interface have a promising future therapeutic application for skin regeneration as the result of the beneficial properties aforemen- tioned. Therefore, further studies are required to optimize the culture medium to promote later-stage differentiation of organotypic HAECs in the air- liquid interface and to transplant them to assess their differentiation potential and roles in wound-healing.
Acknowledgments
This study was funded by the Ministry of Science, Technology and Innovations, Grant No.
02e01e02eSF0289. Fatimah Simat Siti is our PhD student in Universiti Kebangsaan Malaysia and is a National Science Fellowship scholar of Ministry of Science, Technology and Innovations. We thank Megat Radzi Megat A. Rani, Rafiuz Zaman Haroun and Mohamad Hasnul Naim Abd Hamid for tech- nical expertise in electron microscopy. We also thank the Electron Microscopy Unit, Faculty of Science and Technology, UKM and Microscopy Unit, Institute of Bioscience, UPM.
Disclosure of interests: The authors have no commercial, proprietary, or financial interest in the products or companies described in this article.
References
1. Metcalfe AD, Ferguson MW. Tissue engineering of replace- ment skin: the crossroads of biomaterials, wound healing, embryonic development, stem cells and regeneration. J R Soc Interface. 2007;4:413e37.
2. Miki T, Lehmann T, Cai H, Stolz DB, Strom SC. Stem cells characteristics of amniotic epithelial cells. Stem Cells. 2005;
23:1549e59.
3. Fatimah SS, Ng SL, Chua KH, Hayati AR, Tan AE, Tan GC.
Value of human amniotic epithelial cells in tissue engineering for cornea. Hum Cell. 2010;23:141e51.
4. Simat SF, Chua KH, Abdul Rahman H, Tan AE, Tan GC.
The stemness gene expression of cultured human amniotic epithelial cells in serial passages. Med J Malaysia. 2008;
63(Suppl A):53e4.
5. Tan GC, Simat SF, Abdul Rahman H, Tan AE, Chua KH.
Quantitative RT PCR approach to evaluate the neurogenic and gliagenic gene expression of cultured human amniotic epithelial cells. Med J Malaysia. 2008;63(Suppl A):51e2.
6. Kubo M, Sonoda Y, Muramatsu R, Usui M. Immunogenicity of human amniotic membrane in experimental xeno- transplantation. Invest Ophthalmol Vis Sci. 2001;42:1539e46.
7. Akle CA, Adinolfi M, Welsh KI, Leibowitz S, McColl I.
Immunogenicity of human amniotic epithelial cells after transplantation into volunteers. Lancet. 1981;2:1003e5.
8. Byrne C, Tainsky M, Fuchs E. Programming gene expression in developing epidermis. Development. 1994;120:2369e83.
9. Yang A, Schweitzer R, Sun D, Kaghad M, Walker N, Bronson RT, et al. p63 is essential for regenerative prolifera- tion in limb, craniofacial and epithelial development. Nature.
1999;398:714e8.
10. Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR, Bradley A.
p63 is a p53 homologue required for limb and epidermal morphogenesis. Nature. 1999;398:708e13.
11. M’Boneko V, Merker HJ. Development and morphology of the periderm of mouse embryos (days 9e12 of gestation).
Acta Anat (Basel). 1998;133:325e36.
12. Bickenbach JR, Greer JM, Bundman DS, Rothnagel JA, Roop DR. Loricrin expression is coordinated with other epidermal proteins and the appearance of lipid lamellar gran- ules in development. J Invest Dermatol. 1995;104:405e10.
13. Mehrel T, Hohl D, Rothnagel JA, Longley MA, Bundman D, Cheng C, et al. Identification of a major keratinocyte cell envelope protein, loricrin. Cell. 1990;61:1103e12.
14. Akiyama M, Smith LT, Yoneda K, Holbrook KA, Hohl D, Shimizu H. Periderm cell form cornified cell envelope in their regression process during human epidermal development.
J Invest Dermatol. 1999;112:903e9.
15. Prunieras M, Regnier M, Woodley D. Methods for cultivation of keratinocytes with an air liquid interface. J Invest Dermatol.
1983;81(1 Suppl):28se33s.
16. Williams ML, Brown BE, Monger DJ, Grayson S, Elias PM.
Lipid content and metabolism of human keratinocyte cultures grown at the air-medium interface. J Cell Physiol. 1988;136:
103e10.
17. Nolte CJ, Oleson MA, Bilbo PR, Parenteau NL. Develop- ment of a stratum corneum and barrier function in an orga- notypic skin culture. Arch Dermatol Res. 1993;285:466e74.
18. Koike T, Yasuo M, Shimane T, Kobayashi H, Nikaido T, Kurita H. Cultured epithelial grafting using human amniotic membrane: the potential for using human amniotic epithelial cells as a cultured oral epithelium sheet. Arch Oral Biol. 2011;
56:1170e6.
19. Bell E, Sher S, Hull B, Merrill C, Rosen S, Chamson A, et al.
The reconstitution of living skin. J Invest Dermatol. 1983;
81(1 Suppl):2se10s.
20. Thomas CE. The ultrastructure of human amnion epithelium.
J Ultrastruc Res. 1965;13:65e83.
21. Lister UM. Ultrastructure of human amnion, chorion and fetal skin. J Obstet Gynaecol Br Commonw. 1968;75:327e41.
22. Uberti MG, Pierpont YN, Ko F, Wright TE, Smith CA, Cruse CW, et al. Amnion-derived cellular cytokine solution (ACCS) promotes migration of keratinocytes andfibroblasts.
Ann Plast Surg. 2010;64:632e5.
23. Franz MG, Payne WG, Xing L, Naidu DK, Salas RE, Marshall VS, et al. The use of amnion-derived cellular cyto- kine solution to improve wound healing in acute and chronic wound models. Eplasty. 2008;8:188e99.
24. Regauer S, Franke WW, Virtanen I. Intermediatefilament of cytoskeleton of amnion epithelium and cultured amnion epithelial cells: expression of epidermal cytokeratins in cells of a simple epithelium. J Cell Biol. 1985;100:997e1009.
25. Garrod D, Chidgey M, North A. Desmosomes: Differentia- tion, development, dynamics and disease. Curr Opin Cell Biol. 1996;8:670e8.
26. Sun TT, Eichner R, Nelson WG, Tseng SC, Weiss RA, Jarvinen M, et al. Keratin classes: molecular markers for different types of epithelial differentiation. J Invest Dermatol.
1983;81(Suppl 1):109se15s.
27. Koster MI, Kim S, Mills AA, DeMayo FJ, Roop DR. p63 is the molecular switch for initiation of an epithelial stratification program. Genes Dev. 2004;18:126e31.
28. Smith LT, Sakai LY, Burgeson RE, Holbrook KA. Ontogeny of structural components in the dermal-epidermal junctions in human embryonic and fetal skin: the appearance of anchoring fibrils and type VII collagen. J Invest Dermatol. 1988;90:480e5.
29. El-Ghalbzouri A, Gibbs S, Lamme E, Van Blitterswijk CA, Ponec M. Effect offibroblasts on epidermal regeneration. Br J Dermatol. 2002;147:230e43.
30. El Ghalbzouri A, Lamme E, Ponec M. Crucial role offibro- blasts in regulating epidermal morphogenesis. Cell Tissue Res. 2002;310:189e99.
31. Fatimah SS, Tan GC, Chua K, Tan AE, Nur Azurah AG, Hayati AR. Effects of keratinocyte growth factor on skin epithelial differentiation of human amnion epithelial cells. Burns. 2012.
http://dx.doi.org/doi:10.1016/j.burns.2012.10.019.
32. Fatimah SS, Tan GC, Chua K, Fariha MM, Tan AE, Hayati AR. Stemness and angiogenic gene expression changes of serial-passage human amnion mesenchymal cells. Micro- vasc Res. 2013;86:21e9.
33. Alviano F, Fossati V, Marchionni C, Arpinati M, Bonsi L, Franchina M, et al. Term amniotic membrane is a high throughput source for multipotent mesenchymal stem cells with the ability to differentiate into endothelial cells in vitro.
BMC Dev Biol. 2007;7:11.
34. Currie LJ, Sharpe JR, Martin R. The use offibrin glue in skin grafts and tissue-engineered skin replacements: a review. Plast Reconstr Surg. 2001;108:1713e26.
35. Llames SG, Del Rio M, Larcher F, García E, García M, Escamez MJ, et al. Human plasma as a dermal scaffold for the generation of a completely autologous bioengineered skin.
Transplantation. 2004;77:350e5.
36. Ronfard V, Rives JM, Neveux Y, Carsin H, Barrandon Y.
Long term regeneration of human epidermis on third degree burns transplanted with autologous cultured epithelium grown on afibrin matrix. Transplantation. 2000;70:1588e98.
37. Mazlyzam AL, Aminuddin BS, Fuzina NH, Norhayati MM, Fauziah O, Isa MR, et al. Reconstruction of living bilayer human skin equivalent utilizing human fibrin as a scaffold.
Burns. 2007;33:355e63.
38. Huang HJ, Gao QS, Tao BF, Jiang SW. Long-term culture of keratinocyte-like cells derived from mouse embryonic stem cells. In Vitro Cell Dev Biol Anim. 2008;44:193e203.
39. Haase I, Knaup R, Wartenberg M, Sauer H, Hescheler J, Mahrle G. In vitro differentiation of murine embryonic stem cells into keratinocyte-like cells. Eur J Cell Biol. 2007;86:801e5.
40. Schreml S, Meier RJ, Wolfbeis OS, Maisch T, Szeimies RM, Landthaler M, et al. 2D luminescence imaging of physiolog- ical wound oxygenation. Exp Dermatol. 2011;20:550e4.
41. Schreml S, Meier RJ, Weiß KT, Cattani J, Flittner D, Gehmert S, Wolfbeis OS, Landthaler M, Babilas P.
A sprayable luminescent pH sensor and its use for wound imaging in vivo. Exp Dermatol. 2012;21:951e3.
Supplementary data
Supplementary data related to this article can be found online athttp://dx.doi.org/10.1016/j.jcyt.2013.05.003