- Email: [email protected]
A method of isolation and culture of microvascular endothelial cells from mouse skin
Microvascular Research 70 (2005) 198 – 204 www.elsevier.com/locate/ymvre
Short Communication
A method of isolation and culture of microvascular endothelial cells from mouse skin Sung Tae Cha a,1, Dodanim Talavera a,1, Erhan Demir a,1, Anjali K. Nath b, M. Rocio Sierra-Honigmann a,* a
Department of Surgery, Division of Plastic and Reconstructive Surgery, Engineered Wound Repair Laboratory at Davis Building, Cedars Sinai Medical Center, 8700 Beverly Blvd., Los Angeles, CA 90048, USA b Department of Pathology, Yale University School of Medicine, New Haven, CT 06520, USA Received 11 March 2005; revised 5 July 2005; accepted 4 August 2005 Available online 26 September 2005
Abstract Objectives: The study of isolated microvascular endothelial cells from mice has long been impeded due to the many difficulties encountered in isolating and culturing these cells. We focused on developing a method to isolate microvascular endothelial cells from the skin fragments of newborn mice. We also aimed at establishing optimal culture conditions to sustain the growth of these cells. Methods and results: Isolation of murine dermal microvascular endothelial cells (mDMEC) from P3 newborn mice was based first on enzymatic separation of the skin epidermal layer from the dermis using dispase and then on disaggregating dermal cellular elements using collagenase. The cells obtained from the dermis were subjected to a continuous density gradient centrifugation. Cells situated between densities 1.033 and 1.047 were then cultured on collagen IV-coated culture flasks using optimized growth culture conditions. Cells were characterized by endothelial appearance and by the presence and genetic expression of endothelial markers like CD31, NOS3, VEGFR-2 and Tie-2. Uptake of acetylated lowdensity lipoprotein (Ac-LDL) was used as a functional assay. Conclusions: The methodology described herein for isolation and culture of murine microvascular endothelium offers a distinctive advantage for those using mouse models to study endothelial cell biology. D 2005 Elsevier Inc. All rights reserved. Keywords: Endothelial cell; Isolation; Mouse; Dermal; Microvascular
Introduction The dermal microvascular endothelium with its strategic location plays a central role in physiological and pathological processes in the skin including anti-thrombogenicity, blood vessel permeability and pressure control, metabolism of lipoproteins, tissue aging, antigen presentation, tumor metastasis and angiogenesis during wound healing. Studies involving murine microvascular endothelial cells have been generally delayed by the inherent difficulties in developing primary cultures of these cells. Only recently, in vitro cultures of murine macrovascular endothelial cells have been achieved and are already proving to be useful in investiga* Corresponding author. Fax: +1 310 423 2325. E-mail address: [email protected] (M.R. Sierra-Honigmann). 1 Contributed equally to this work. 0026-2862/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mvr.2005.08.002
tions of some aspects of vascular biology (Lincoln-li et al., 2003; Huang et al., 2003; Magid et al., 2003). However, studies of cultured macrovascular endothelium may not be valid to investigate specific microvascular behavior and function (King et al., 1983). Microvascular endothelial cell cultures have been done using a variety of isolation- and culture-methods, for example human dermal microvascular endothelial cells isolated from neonatal foreskin, adult human skin or tissue derived from vascular tumors (Karasek, 1989). Methods based on continuous gradient centrifugation of cells obtained from human dermis after enzymatic and mechanical disaggregating of tissue fragments have been described over the last two decades (Marks et al., 1985; Imcke et al., 1991). However, murine models on microvascular cell isolation are still limited and there is no current description of microvascular segment isolation from the mouse skin (Dong et al., 1997; Marelli-Berg et al., 2000; Li et al., 2001).
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
199
Here, we describe a method for isolation of murine dermal microvascular endothelial cells (mDMEC) based on enzymatic disaggregating and continuous density gradient separation. We also describe optimized cultured conditions for these cells. The availability of mouse microvascular endothelial cultures would therefore offer an important asset for those attempting to answer questions related to endothelial biology through the use of murine models. Materials and methods Endothelial cell isolation and culture All animal experiments were done under an approved protocol of the Cedars-Sinai Animal Care and Use Committee. Murine dermal microvascular endothelial cells were derived from the skin of 3-day-old BALB/c mice (Jackson Laboratory, Bar Harbor, ME). Fig. 1 shows a schematic diagram of the procedure used to isolate these cells. After decapitation of the mice, the entire skin flap was dissected and placed immediately in Hank’s balanced salt solution (HBSS) without calcium and magnesium (Invitrogen Corp, Carlsbad, CA), containing 100 U/ml penicillin and 100 Ag/ml streptomycin (Invitrogen Corp, Carlsbad, CA), 2.5 Ag/ml fungiozone, 205 Ag/ml sodium deoxycholate and 10 Ag/ml gentamycin (Omega Scientific, Tarzana, CA). After washing the skin with this solution three times for 5 min each, the tissue flap was then incubated for 45 min at 37-C in 5 mg/ml dispase solution (Invitrogen Corp, Carlsbad, CA) with continuous agitation using an orbital mixer. Then, the epidermis was separated from the dermis by gentle mechanical
Fig. 2. Phase-contrast microscopy of subconfluent and confluent mDMEC monolayer. (A) mDMEC 2 h after plating. (B) Subconfluent endothelial cells 72 h after plating that have proliferated and started making contact to each other. (C) Confluent mDMEC monolayer exhibited a typical ‘‘cobblestone’’ morphology several days after plating. Scale bar = 50 Am.
Fig. 1. Schematic representation of the procedure used to isolate and culture mDMEC. The skin of 3 day-old mice was removed and the dermis was separated from epidermis after enzymatic treatment with dispase. The dermal fragments were subsequently treated with collagenase and the cell suspension was filtered and layered onto a gradient of Percoll. After a continuous density gradient centrifugation, the cells were plated with complete media on collagen IV-coated culture flasks. In several days, they reached confluence and exhibited ‘‘cobblestone’’ morphology.
dissociation with Adson forceps (Fine Science Tools Inc., Foster City, CA). The entire epidermis was removed as a single layer, leaving the intact dermis for further processing. The residual dermal fragments were then incubated on the orbital mixer overnight at 4-C in a 4% type 1A collagenase solution (Worthington, Lakewood, NJ) with 4% bovine serum albumin (BSA) in PBS. Shorter incubation times such as 1 h at 37-C were tested with similar results. At the end of the incubation, the dermal tissue fragments were passed through 100 Am nylon mesh cell strainer (BD Labware, Franklin Lakes, NJ), and the cell suspension obtained using 20 volumes of HBBS (Invitrogen Corp, Carlsbad, CA), and centrifuged for 5 min at 400 g. The cell pellet was resuspended in 200 Al of PBS and set aside on ice until the gradient was ready. Concurrently, a sterile gradient of Percoll was separately prepared by centrifuging 8 ml of a 35% Percoll solution (Amersham Biosciences, Uppsala, Sweden), in oak ridge polycarbonate centrifuge tubes (Nalgene, Rochester, NY) at 30,000 g for 15 min at 4-C. The formation of the gradient was done using maximum acceleration without brake on a fixedangle rotor (AM 10.17 angle rotor 10 10 ml) (Jouan Inc., Winchester, VA). Then, the resuspended cells were layered onto the gradient and centrifuged for 10 min at 400 g at room temperature. In parallel to any gradient tubes containing the cells, an additional gradient tube was used as density reference by substituting the sample with 200 Al of a mixture of colored density
200
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
Fig. 3. Expression of endothelial cell markers in mDMEC. Positive immunoreactivity for anti-CD31 (A), anti-NOS3 (B) and anti-vWF (C) antibodies (red) is shown, as well as uptake of DiI-Ac-LDL by mDMEC (D) (red). The nuclei were stained with DAPI for contrast (blue). Scale bar = 10 Am.
marker beads (Amersham Pharmacia Biotech, Uppsala, Sweden). Endothelial cells lay in the gradient between densities 1.033 g/ml and 1.047 g/ml, whereas nonendothelial cells have a density greater than 1.065 g/ml (Ruszczak, 1996). The cell band at the appropriate density was then collected, washed in HBSS (Invitrogen Corp, Carlsbad, CA), counted and plated at 7500 cells/cm2 on PRIMARIAi tissue culture plastic plates (BD Biosciences, Bedford, MA). The culture surface was coated with 3% type IV collagen (BD Biosciences, Bedford, MA), dissolved in 0.05 N HCl. The growth medium composition consisted of MDCB 131 medium (Invitrogen Corp, Carlsbad, CA) supplemented with 10% FBS (Omega Scientific, Tarzana, CA), 100 U/ml penicillin and 100 Ag/ml streptomycin, 10 mM HEPES, 10 mM l-glutamine (Invitrogen Corp, Carlsbad, CA), 30 Ag/ml endothelial cell growth supplement (ECGS) (Upstate, Lake Placid, NY), 15 U/ml heparin, 1 Ag/ml hydrocortisone acetate, 325 Ag/ml glutathione, 0.33 mM iso-butyl-methyl-xanthene (IBMX), 0.05 mM N6,2V-O-dibutyryladenosine 3V,5V-cyclic monophosphate (dbcAMP), 5 Ag/ml insulin, 5 Ag/ml transferrin, 0.01 mM forskolin, 5 AM 2-mercaptoethanol (Sigma, St. Louis, MO) and 5 ng/ml recombinant human vascular endothelial growth factor (rhVEGF, a gift from Genentech Inc., San Francisco, CA).
Immunofluorescence staining Endothelial cells were cultured on glass coverslips and allowed to grow to confluence. Cells were fixed with methanol/acetone at 20-C and washed three times in TBS-Tween (Sigma, St Louis, MO). Cells were rehydrated with PBS with 1% BSA (Sigma, St Louis, MO) solution for 30 min at room temperature, washed and incubated with primary antibody for 1 h at room temperature, then washed three times and incubated for 30 min at room temperature with a secondary antibody. After another washing in PBS, the cells on the coverslips were mounted onto slides using GEL/MOUNTi (Biomeda, Foster City, CA). Three antibodies were used for this study: rat anti-mouse CD31 (PECAM-1) monoclonal antibody (BD Pharmigen, San Diego, CA); anti-NOS3 rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and anti-von Willebrand Factor polyclonal antibody (anti-vWF) (Calbiochem, San Diego, CA). Goat anti-rat IgG and donkey anti-rabbit IgG antibody conjugated to Texas red (Jackson ImmnunoResearch, West Grove, PA) were used as secondary antibodies.
DiI-Ac-LDL uptake Confluent cell monolayers were used for uptake of acetylated low-density lipoprotein labeled with 1,1V-dioctadecyl-3,3,3V,3V-tetramethylindo-carbocynamina perchlorate (DiI-Ac-LDL) (Molecular Probes, Eugene, OR) at a final concentration of 20 Ag/ml and incubated for 6 h at 37-C. For microscopic visualization, the cells were washed and fixed with Cytospray (Streck Laboratories, La Vista, NE). For nuclear counterstaining, 4V-6V-diamidino-2phenylindole dihydrochloride (DAPI) was used and cells were mounted and observed under a fluorescence microscope.
Flow cytometric analysis Flow cytometric analysis was done to verify endothelial phenotype. Confluent monolayers were treated with Dil-Ac-LDL for 6 h as mentioned above and detached using Versene 0.05% (Invitrogen Corp, Carlsbad, CA) to be analyzed by FACS Scan (Becton Dickinson, San Jose, CA). Other monolayers of endothelial cells without any treatment were detached with Versene 0.05% (Invitrogen Corp, Carlsbad, CA) and then resuspended with 1% PBS – BSA (Sigma, St. Louis, MO). These cells were then incubated with rat anti-mouse CD31 (PECAM-1) (BD Pharmigen, San Diego, CA), CD34 (Novus Biologicals, Inc, Littleton, CO) and CD144 (BD Biosciences, San Jose, CA) monoclonal antibodies for overnight at 4-C, washed three times in PBS-1% BSA (Sigma, St Louis, MO) and finally stained with fluorescein isothiocyanate (FITC) conjugated goat anti-rat antibody (Jackson ImmnunoResearch, West Grove, PA) for 1 h at room temperature. Cells were subsequently washed in PBS three times and fixed with 2% paraformaldehyde. These groups of cells were also analyzed using flow cytometry.
Gene expression using real-time RT-PCR Total RNA was isolated from mDMEC in two consecutive extractions using TrizolR (Invitrogen, Carsbald, CA) to ensure RNA purity. Before cDNA synthesis, the samples were digested with DNase I to eliminate any residual genomic DNA contamination (Ambion, Austin, TX). The cDNA synthesis was carried out using SuperScripti II (Invitrogen, Carlsbad, CA). Real-time PCR
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
201
Fig. 4. Expression of endothelial cell markers CD31 (PECAM-1), CD34, CD144 (VE-Cadherin) and DiI-Ac-LDL uptake evaluated by flow cytometric analysis. The histograms in each graph show the increase in fluorescence for each endothelial cell marker and for DiI-Ac-LDL endothelial uptake (thick line) compared with the expression of an isotype-matched irrelevant mAb and no DiI-Ac-LDL treated cells used as control (thin line). (qPCR) was performed employing the iCycler iQ real-time PCR machine (BioRad, Hercules, CA) and using HotStarTaqi DNA polymerase (Qiagen, Valencia, CA). TaqmanR probe-based chemistry was chosen for accurate realtime PCR analysis. The TaqmanR used had dual-labeled fluorogenic probe (prb) 5V-FAM/3V-BHQ (Biosearch Technologies, Novato, CA), and flanking forward (fwd) and reverse (rev) primers. The sequences used were as follows (all sequences are 5V– 3V): CD31: prb-CCAAGCTGGGATCCTGTCCG, fwdCAGGACCACGTGTTAGTGTT, rev-ACTCCTGATGGGTTCTGACT; NOS3: prb-CGTGCACAGGCGGAAGATGT, fwd-AGAGCCTGCAATTACTACCA, rev-GTGGATTTGCTGCTCTGTAG; VEGFR2: prb-TCCTGGGACTGTGGCGAAGATG, fwd- GAAGATTGTAAACCGGGATG, rev-TTGGQ TCACTCTTGGTCACAC; and Tie2: prb-CGAGAGGCGATCCCTGCAAA, fwd-TGACTTGGCAACCGATATTT, rev-AGGCACTTTGATGTTCTGCT. All the probes and primers were designed to be specific for murine sequences.
Results mDMEC isolation and culture The skin of 3-day old mice was selected as the optimal tissue sample trough a series of trial and error experiments. Comparing skin from newborn, day 3, day 10 and adult mice (data not shown), the 3-day old mice skin was found to have the optimal strength, size and manageability for the various purification steps. The initial incubation with dispase was
found to be critical because it allowed for a clean separation of dermal and epidermal layers. Dispase targets primarily the basement membrane that separates these two layers of skin. Therefore, cell isolation was possible by manipulation of the skin dermis with virtually no keratinocyte contamination. The subsequent collagenase digestion step released endothelial cells and vascular fragments from the dermis. The gradient separation procedure after overnight collagenase digestion separated microvascular cells from nonendothelial cells such as dendritic cells, melanocytes, fibroblasts and pericytes. The continuos gradient of Percoll was prepared from freshly made solution of 35% Percoll in PBS. Once the gradient was formed, the tube was kept firmly upright and the cells were carefully layered using a steady flow from the side of the tube to avoid disturbing the gradient. Assisted by a parallel tube with colorcoded density marker beads, the endothelial layer was recovered between the densities of 1.033 and 1.047. The final purified cell yielded an average of 0.5 million cells per mouse. The collagen IV-coated culture surfaces allowed adequate endothelial cell attachment within 2 h. After 24 h, cells debris and nonadherent cells were removed and fresh complete medium was added. The plated cells typically reached confluence by day 5 and then they were subcultured. At
202
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
confluence, the average cell density was approximately 4 105 cells/cm2. The culture conditions described here are similar to those previously used for the growth of human dermal
microvascular endothelial cells (Ruszczak, 1996). These culture conditions discourage the growth of nonendothelial cells and stimulate the endothelial cell proliferation. A salient
Fig. 5. Gene expression of different endothelial-cell membrane-associated proteins. Total RNA from mDMEC was used for reverse transcriptase (RT) followed by real-time PCR reaction using murine-specific TaqmanR probe-primer sets. Expression of CD31, NOS3, VEGFR-2 and Tie-2 was determined with this method. The horizontal record in each graph corresponds to a negative control where no RT was used.
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
difference between the culture conditions described here and those used by other authors is the addition of rhVEGF described herein. Morphological and phenotypic characterization of mDMEC Cultured mDMEC monolayers showed morphological characteristics that closely resembled endothelial cell cultures from other species. Figs. 2A –C show endothelial cells at low density, subconfluent and full confluent ‘‘cobblestone’’, respectively. The cells typically maintain their uniform appearance up to 3 serial passages suggesting a homogeneous cell population. The cellular identity of mDMEC was confirmed by assessment of structural and functional parameters: the isolated mDMEC cultures had positive immmunoreactivity to anti-CD31 (PECAM) (Figs. 3A and 4). In addition, mDMEC also displayed the characteristic immunoreactivity to NOS3 in the Golgivesicular compartment (Fig. 3B). The cells also reacted to antibodies directed against von Willebrand Factor (anti-vWF), which is commonly used for endothelial cell identification (Fig. 3C). A functional endothelial assay was performed on subconfluent mDMEC cultures and demonstrated uptake of DiI-AcLDL (Figs. 3D and 4). Flow cytometric analysis To further address the endothelial phenotype and determine the degree of purity of the cell preparation, flow cytometry analysis was done using anti-CD31, CD34 and CD144 antibodies. The labeled fraction of the samples showed a significant increase in the fluorescence respect to controls (Fig. 4). The histograms shown reveled greater than 90% purity of the cell preparation (CD31, 91%; CD34, 93%; CD144, 92% and DiI-AcLDL, 99.8%). Gene expression studies Gene expression of endothelial-specific markers was analyzed by real-time RT-PCR. Fig. 5 shows the expression for four different endothelial markers (CD31, NOS3, VEGFR-2 and Tie-2). The presence of mRNA for each markers further confirmed the identity of the isolated cells. Discussion Murine microvascular endothelial cells have been difficult to culture for several reasons. Primary cultures are difficult to prepare and are often contaminated by fibroblast and other stromal cells. Furthermore, the available compositions of endothelial culture media do not fully satisfy the growth requirements of murine cells (Karasek, 1989; Imcke et al., 1991; Davison et al., 1980, 1983). The most common attempts to isolate murine microvascular endothelial cells have involved incubation of tissues in digestive enzymes and isolation using immunoabsorption to magnetic beads (Dong et al., 1997). However, in our experience, the use of magnetic beads results in low cell yields and remaining beads are difficult to
203
dissociate from the purified cells. Techniques for isolation of human dermal microvascular endothelial cells utilize neonatal foreskin or adult human skin (Karasek, 1989; Davison et al., 1980). These techniques rely on an initial separation of the epidermal layer from the dermis. Once of highly vascularized dermis is free of epithelial cells, it is treated with enzymes such as collagenase or trypsin and the vascular fragments are extracted. Here, we have adapted this methodology to isolate endothelial cells from murine skin. The separation of epidermal and dermal layer is only achievable before the murine skin matures and the hair follicles are present. We identified P3 as the ideal time to have a successful and complete separation of the dermis and epidermis. The collagen-rich dermis has abundant microvessels, therefore it is a good source of microvascular endothelial cells. Our strategy to isolate endothelial cells from the dermis includes an enzymatic extraction of the cellular components of the dermis followed by continuous density gradient centrifugation. With these methods, we have succeeded in isolating murine dermal endothelial cells. We report for the first time the detailed isolation strategy and also the culture conditions to sustain the growth of these cells. The advantage of generating primary endothelial cultures using the skin from individual mice is of particular interest to investigators using transgenic and null murine models where the availability of animals is limited and costly. Cultured mDMEC may be used to generate models to analyze basic aspects of microvascular biology and to investigate possible applications for tissue engineering. Acknowledgments The authors wish to thank A. Murad for technical assistance on gene expression experiments. This work was supported by funding from the Skirball Foundation to the Division of Plastic Surgery of Cedars-Sinai Medical Center and by NIH grant 1R01GM66292-01A1 to M.R.S.H. References Davison, P.M., Bensch, K., Karasek, M.A., 1980. Isolation and growth of endothelial cells from the microvessels of the newborn human foreskin and cell culture. J. Invest. Dermatol. 75, 316 – 321. Davison, P.M., Bensch, K., Karasek, M.A., 1983. Isolation and long-term serial cultivation of endothelial cells from the growth the microvessels of the adult human dermis. In Vitro 19, 937 – 945. Dong, Q.G., Bernasconi, S., Lostaglio, S., De Calmanovici, R.W., MartinPadura, I., Brevario, F., Garlanda, C., Ramponi, S., Mantovani, A., Vecchi, A., 1997. A general strategy for isolation of endothelial cells from murine tissues: characterization of two-endothelial cell lines from the murine lung and subcutaneous sponge implants. Arterioscler., Thromb., Vasc. Biol. 17, 1599 – 1604. Huang, H., McIntosh, J., Hoyt, D.G., 2003. An efficient, nonenzymatic method for isolation and culture of murine aortic endothelial cells and their response to inflammatory stimuli. In Vitro Cell. Dev. Biol.: Anim. 39, 43 – 50. Imcke, E., Ruszczak, Z., Mayer-da Silva, A., Detmar, M., Orfanos, C.E., 1991. Cultivation of human dermal microvascular endothelial cells in vitro: immunocytochemical and ultrastructural characterization and effect of treatment with three synthetic retinoids. Arch. Dermatol. Res. 283, 149 – 157.
204
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
Karasek, M.A., 1989. Microvascular endothelial cell culture. J. Invest. Dermatol. 93 (2 Suppl.), 33S – 38S. King, G.L., Buzney, S.M., Kahn, C.R., Hetu, N., 1983. Differential responsiveness to insulin of endothelial cells. Biochem. Pharmacol. 71, 974 – 979. Li, J.M., Mullen, A.M., Shah, A.M., 2001. Phenotypic properties and characteristics of superoxide production by mouse coronary microvascular endothelial cells. J. Mol. Cell. Cardiol. 33, 1119 – 1131. Lincoln-li, D.W., Larsen, A.M., Phillips, P.G., Bove, K., 2003. Isolation of murine aortic endothelial cells in culture and effects of sex steroids on their growth. In Vitro Cell. Dev. Biol.: Anim. 39, 140 – 145. Magid, R., Martinson, D., Hwang, J., Jo, H., Galis, Z.S., 2003. Optimization of
isolation and functional characterization of primary murine aortic endothelial cells. Endothelium 10, 103 – 109. Marelli-Berg, F., Peek, M., Lidington, E.A., Stauss, H.J., Lechler, R.I., 2000. Isolation of endothelial cells from murine tissue. J. Immun. Methods 244, 205 – 215. Marks, R.M., Czerniecki, M., Penny, P., 1985. Human dermal microvascular endothelial cells: an improved method for tissue culture and a description of some singular properties in culture. In Vitro 21, 627 – 635. Ruszczak, Z., 1996. Human skin microvascular endothelial cells. In: Bicknell, R. (Ed.), Endothelial Cell Culture, vol. 1. Cambridge Univ. Press, New York, pp. 77 – 89.
Short Communication
A method of isolation and culture of microvascular endothelial cells from mouse skin Sung Tae Cha a,1, Dodanim Talavera a,1, Erhan Demir a,1, Anjali K. Nath b, M. Rocio Sierra-Honigmann a,* a
Department of Surgery, Division of Plastic and Reconstructive Surgery, Engineered Wound Repair Laboratory at Davis Building, Cedars Sinai Medical Center, 8700 Beverly Blvd., Los Angeles, CA 90048, USA b Department of Pathology, Yale University School of Medicine, New Haven, CT 06520, USA Received 11 March 2005; revised 5 July 2005; accepted 4 August 2005 Available online 26 September 2005
Abstract Objectives: The study of isolated microvascular endothelial cells from mice has long been impeded due to the many difficulties encountered in isolating and culturing these cells. We focused on developing a method to isolate microvascular endothelial cells from the skin fragments of newborn mice. We also aimed at establishing optimal culture conditions to sustain the growth of these cells. Methods and results: Isolation of murine dermal microvascular endothelial cells (mDMEC) from P3 newborn mice was based first on enzymatic separation of the skin epidermal layer from the dermis using dispase and then on disaggregating dermal cellular elements using collagenase. The cells obtained from the dermis were subjected to a continuous density gradient centrifugation. Cells situated between densities 1.033 and 1.047 were then cultured on collagen IV-coated culture flasks using optimized growth culture conditions. Cells were characterized by endothelial appearance and by the presence and genetic expression of endothelial markers like CD31, NOS3, VEGFR-2 and Tie-2. Uptake of acetylated lowdensity lipoprotein (Ac-LDL) was used as a functional assay. Conclusions: The methodology described herein for isolation and culture of murine microvascular endothelium offers a distinctive advantage for those using mouse models to study endothelial cell biology. D 2005 Elsevier Inc. All rights reserved. Keywords: Endothelial cell; Isolation; Mouse; Dermal; Microvascular
Introduction The dermal microvascular endothelium with its strategic location plays a central role in physiological and pathological processes in the skin including anti-thrombogenicity, blood vessel permeability and pressure control, metabolism of lipoproteins, tissue aging, antigen presentation, tumor metastasis and angiogenesis during wound healing. Studies involving murine microvascular endothelial cells have been generally delayed by the inherent difficulties in developing primary cultures of these cells. Only recently, in vitro cultures of murine macrovascular endothelial cells have been achieved and are already proving to be useful in investiga* Corresponding author. Fax: +1 310 423 2325. E-mail address: [email protected] (M.R. Sierra-Honigmann). 1 Contributed equally to this work. 0026-2862/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.mvr.2005.08.002
tions of some aspects of vascular biology (Lincoln-li et al., 2003; Huang et al., 2003; Magid et al., 2003). However, studies of cultured macrovascular endothelium may not be valid to investigate specific microvascular behavior and function (King et al., 1983). Microvascular endothelial cell cultures have been done using a variety of isolation- and culture-methods, for example human dermal microvascular endothelial cells isolated from neonatal foreskin, adult human skin or tissue derived from vascular tumors (Karasek, 1989). Methods based on continuous gradient centrifugation of cells obtained from human dermis after enzymatic and mechanical disaggregating of tissue fragments have been described over the last two decades (Marks et al., 1985; Imcke et al., 1991). However, murine models on microvascular cell isolation are still limited and there is no current description of microvascular segment isolation from the mouse skin (Dong et al., 1997; Marelli-Berg et al., 2000; Li et al., 2001).
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
199
Here, we describe a method for isolation of murine dermal microvascular endothelial cells (mDMEC) based on enzymatic disaggregating and continuous density gradient separation. We also describe optimized cultured conditions for these cells. The availability of mouse microvascular endothelial cultures would therefore offer an important asset for those attempting to answer questions related to endothelial biology through the use of murine models. Materials and methods Endothelial cell isolation and culture All animal experiments were done under an approved protocol of the Cedars-Sinai Animal Care and Use Committee. Murine dermal microvascular endothelial cells were derived from the skin of 3-day-old BALB/c mice (Jackson Laboratory, Bar Harbor, ME). Fig. 1 shows a schematic diagram of the procedure used to isolate these cells. After decapitation of the mice, the entire skin flap was dissected and placed immediately in Hank’s balanced salt solution (HBSS) without calcium and magnesium (Invitrogen Corp, Carlsbad, CA), containing 100 U/ml penicillin and 100 Ag/ml streptomycin (Invitrogen Corp, Carlsbad, CA), 2.5 Ag/ml fungiozone, 205 Ag/ml sodium deoxycholate and 10 Ag/ml gentamycin (Omega Scientific, Tarzana, CA). After washing the skin with this solution three times for 5 min each, the tissue flap was then incubated for 45 min at 37-C in 5 mg/ml dispase solution (Invitrogen Corp, Carlsbad, CA) with continuous agitation using an orbital mixer. Then, the epidermis was separated from the dermis by gentle mechanical
Fig. 2. Phase-contrast microscopy of subconfluent and confluent mDMEC monolayer. (A) mDMEC 2 h after plating. (B) Subconfluent endothelial cells 72 h after plating that have proliferated and started making contact to each other. (C) Confluent mDMEC monolayer exhibited a typical ‘‘cobblestone’’ morphology several days after plating. Scale bar = 50 Am.
Fig. 1. Schematic representation of the procedure used to isolate and culture mDMEC. The skin of 3 day-old mice was removed and the dermis was separated from epidermis after enzymatic treatment with dispase. The dermal fragments were subsequently treated with collagenase and the cell suspension was filtered and layered onto a gradient of Percoll. After a continuous density gradient centrifugation, the cells were plated with complete media on collagen IV-coated culture flasks. In several days, they reached confluence and exhibited ‘‘cobblestone’’ morphology.
dissociation with Adson forceps (Fine Science Tools Inc., Foster City, CA). The entire epidermis was removed as a single layer, leaving the intact dermis for further processing. The residual dermal fragments were then incubated on the orbital mixer overnight at 4-C in a 4% type 1A collagenase solution (Worthington, Lakewood, NJ) with 4% bovine serum albumin (BSA) in PBS. Shorter incubation times such as 1 h at 37-C were tested with similar results. At the end of the incubation, the dermal tissue fragments were passed through 100 Am nylon mesh cell strainer (BD Labware, Franklin Lakes, NJ), and the cell suspension obtained using 20 volumes of HBBS (Invitrogen Corp, Carlsbad, CA), and centrifuged for 5 min at 400 g. The cell pellet was resuspended in 200 Al of PBS and set aside on ice until the gradient was ready. Concurrently, a sterile gradient of Percoll was separately prepared by centrifuging 8 ml of a 35% Percoll solution (Amersham Biosciences, Uppsala, Sweden), in oak ridge polycarbonate centrifuge tubes (Nalgene, Rochester, NY) at 30,000 g for 15 min at 4-C. The formation of the gradient was done using maximum acceleration without brake on a fixedangle rotor (AM 10.17 angle rotor 10 10 ml) (Jouan Inc., Winchester, VA). Then, the resuspended cells were layered onto the gradient and centrifuged for 10 min at 400 g at room temperature. In parallel to any gradient tubes containing the cells, an additional gradient tube was used as density reference by substituting the sample with 200 Al of a mixture of colored density
200
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
Fig. 3. Expression of endothelial cell markers in mDMEC. Positive immunoreactivity for anti-CD31 (A), anti-NOS3 (B) and anti-vWF (C) antibodies (red) is shown, as well as uptake of DiI-Ac-LDL by mDMEC (D) (red). The nuclei were stained with DAPI for contrast (blue). Scale bar = 10 Am.
marker beads (Amersham Pharmacia Biotech, Uppsala, Sweden). Endothelial cells lay in the gradient between densities 1.033 g/ml and 1.047 g/ml, whereas nonendothelial cells have a density greater than 1.065 g/ml (Ruszczak, 1996). The cell band at the appropriate density was then collected, washed in HBSS (Invitrogen Corp, Carlsbad, CA), counted and plated at 7500 cells/cm2 on PRIMARIAi tissue culture plastic plates (BD Biosciences, Bedford, MA). The culture surface was coated with 3% type IV collagen (BD Biosciences, Bedford, MA), dissolved in 0.05 N HCl. The growth medium composition consisted of MDCB 131 medium (Invitrogen Corp, Carlsbad, CA) supplemented with 10% FBS (Omega Scientific, Tarzana, CA), 100 U/ml penicillin and 100 Ag/ml streptomycin, 10 mM HEPES, 10 mM l-glutamine (Invitrogen Corp, Carlsbad, CA), 30 Ag/ml endothelial cell growth supplement (ECGS) (Upstate, Lake Placid, NY), 15 U/ml heparin, 1 Ag/ml hydrocortisone acetate, 325 Ag/ml glutathione, 0.33 mM iso-butyl-methyl-xanthene (IBMX), 0.05 mM N6,2V-O-dibutyryladenosine 3V,5V-cyclic monophosphate (dbcAMP), 5 Ag/ml insulin, 5 Ag/ml transferrin, 0.01 mM forskolin, 5 AM 2-mercaptoethanol (Sigma, St. Louis, MO) and 5 ng/ml recombinant human vascular endothelial growth factor (rhVEGF, a gift from Genentech Inc., San Francisco, CA).
Immunofluorescence staining Endothelial cells were cultured on glass coverslips and allowed to grow to confluence. Cells were fixed with methanol/acetone at 20-C and washed three times in TBS-Tween (Sigma, St Louis, MO). Cells were rehydrated with PBS with 1% BSA (Sigma, St Louis, MO) solution for 30 min at room temperature, washed and incubated with primary antibody for 1 h at room temperature, then washed three times and incubated for 30 min at room temperature with a secondary antibody. After another washing in PBS, the cells on the coverslips were mounted onto slides using GEL/MOUNTi (Biomeda, Foster City, CA). Three antibodies were used for this study: rat anti-mouse CD31 (PECAM-1) monoclonal antibody (BD Pharmigen, San Diego, CA); anti-NOS3 rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) and anti-von Willebrand Factor polyclonal antibody (anti-vWF) (Calbiochem, San Diego, CA). Goat anti-rat IgG and donkey anti-rabbit IgG antibody conjugated to Texas red (Jackson ImmnunoResearch, West Grove, PA) were used as secondary antibodies.
DiI-Ac-LDL uptake Confluent cell monolayers were used for uptake of acetylated low-density lipoprotein labeled with 1,1V-dioctadecyl-3,3,3V,3V-tetramethylindo-carbocynamina perchlorate (DiI-Ac-LDL) (Molecular Probes, Eugene, OR) at a final concentration of 20 Ag/ml and incubated for 6 h at 37-C. For microscopic visualization, the cells were washed and fixed with Cytospray (Streck Laboratories, La Vista, NE). For nuclear counterstaining, 4V-6V-diamidino-2phenylindole dihydrochloride (DAPI) was used and cells were mounted and observed under a fluorescence microscope.
Flow cytometric analysis Flow cytometric analysis was done to verify endothelial phenotype. Confluent monolayers were treated with Dil-Ac-LDL for 6 h as mentioned above and detached using Versene 0.05% (Invitrogen Corp, Carlsbad, CA) to be analyzed by FACS Scan (Becton Dickinson, San Jose, CA). Other monolayers of endothelial cells without any treatment were detached with Versene 0.05% (Invitrogen Corp, Carlsbad, CA) and then resuspended with 1% PBS – BSA (Sigma, St. Louis, MO). These cells were then incubated with rat anti-mouse CD31 (PECAM-1) (BD Pharmigen, San Diego, CA), CD34 (Novus Biologicals, Inc, Littleton, CO) and CD144 (BD Biosciences, San Jose, CA) monoclonal antibodies for overnight at 4-C, washed three times in PBS-1% BSA (Sigma, St Louis, MO) and finally stained with fluorescein isothiocyanate (FITC) conjugated goat anti-rat antibody (Jackson ImmnunoResearch, West Grove, PA) for 1 h at room temperature. Cells were subsequently washed in PBS three times and fixed with 2% paraformaldehyde. These groups of cells were also analyzed using flow cytometry.
Gene expression using real-time RT-PCR Total RNA was isolated from mDMEC in two consecutive extractions using TrizolR (Invitrogen, Carsbald, CA) to ensure RNA purity. Before cDNA synthesis, the samples were digested with DNase I to eliminate any residual genomic DNA contamination (Ambion, Austin, TX). The cDNA synthesis was carried out using SuperScripti II (Invitrogen, Carlsbad, CA). Real-time PCR
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
201
Fig. 4. Expression of endothelial cell markers CD31 (PECAM-1), CD34, CD144 (VE-Cadherin) and DiI-Ac-LDL uptake evaluated by flow cytometric analysis. The histograms in each graph show the increase in fluorescence for each endothelial cell marker and for DiI-Ac-LDL endothelial uptake (thick line) compared with the expression of an isotype-matched irrelevant mAb and no DiI-Ac-LDL treated cells used as control (thin line). (qPCR) was performed employing the iCycler iQ real-time PCR machine (BioRad, Hercules, CA) and using HotStarTaqi DNA polymerase (Qiagen, Valencia, CA). TaqmanR probe-based chemistry was chosen for accurate realtime PCR analysis. The TaqmanR used had dual-labeled fluorogenic probe (prb) 5V-FAM/3V-BHQ (Biosearch Technologies, Novato, CA), and flanking forward (fwd) and reverse (rev) primers. The sequences used were as follows (all sequences are 5V– 3V): CD31: prb-CCAAGCTGGGATCCTGTCCG, fwdCAGGACCACGTGTTAGTGTT, rev-ACTCCTGATGGGTTCTGACT; NOS3: prb-CGTGCACAGGCGGAAGATGT, fwd-AGAGCCTGCAATTACTACCA, rev-GTGGATTTGCTGCTCTGTAG; VEGFR2: prb-TCCTGGGACTGTGGCGAAGATG, fwd- GAAGATTGTAAACCGGGATG, rev-TTGGQ TCACTCTTGGTCACAC; and Tie2: prb-CGAGAGGCGATCCCTGCAAA, fwd-TGACTTGGCAACCGATATTT, rev-AGGCACTTTGATGTTCTGCT. All the probes and primers were designed to be specific for murine sequences.
Results mDMEC isolation and culture The skin of 3-day old mice was selected as the optimal tissue sample trough a series of trial and error experiments. Comparing skin from newborn, day 3, day 10 and adult mice (data not shown), the 3-day old mice skin was found to have the optimal strength, size and manageability for the various purification steps. The initial incubation with dispase was
found to be critical because it allowed for a clean separation of dermal and epidermal layers. Dispase targets primarily the basement membrane that separates these two layers of skin. Therefore, cell isolation was possible by manipulation of the skin dermis with virtually no keratinocyte contamination. The subsequent collagenase digestion step released endothelial cells and vascular fragments from the dermis. The gradient separation procedure after overnight collagenase digestion separated microvascular cells from nonendothelial cells such as dendritic cells, melanocytes, fibroblasts and pericytes. The continuos gradient of Percoll was prepared from freshly made solution of 35% Percoll in PBS. Once the gradient was formed, the tube was kept firmly upright and the cells were carefully layered using a steady flow from the side of the tube to avoid disturbing the gradient. Assisted by a parallel tube with colorcoded density marker beads, the endothelial layer was recovered between the densities of 1.033 and 1.047. The final purified cell yielded an average of 0.5 million cells per mouse. The collagen IV-coated culture surfaces allowed adequate endothelial cell attachment within 2 h. After 24 h, cells debris and nonadherent cells were removed and fresh complete medium was added. The plated cells typically reached confluence by day 5 and then they were subcultured. At
202
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
confluence, the average cell density was approximately 4 105 cells/cm2. The culture conditions described here are similar to those previously used for the growth of human dermal
microvascular endothelial cells (Ruszczak, 1996). These culture conditions discourage the growth of nonendothelial cells and stimulate the endothelial cell proliferation. A salient
Fig. 5. Gene expression of different endothelial-cell membrane-associated proteins. Total RNA from mDMEC was used for reverse transcriptase (RT) followed by real-time PCR reaction using murine-specific TaqmanR probe-primer sets. Expression of CD31, NOS3, VEGFR-2 and Tie-2 was determined with this method. The horizontal record in each graph corresponds to a negative control where no RT was used.
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
difference between the culture conditions described here and those used by other authors is the addition of rhVEGF described herein. Morphological and phenotypic characterization of mDMEC Cultured mDMEC monolayers showed morphological characteristics that closely resembled endothelial cell cultures from other species. Figs. 2A –C show endothelial cells at low density, subconfluent and full confluent ‘‘cobblestone’’, respectively. The cells typically maintain their uniform appearance up to 3 serial passages suggesting a homogeneous cell population. The cellular identity of mDMEC was confirmed by assessment of structural and functional parameters: the isolated mDMEC cultures had positive immmunoreactivity to anti-CD31 (PECAM) (Figs. 3A and 4). In addition, mDMEC also displayed the characteristic immunoreactivity to NOS3 in the Golgivesicular compartment (Fig. 3B). The cells also reacted to antibodies directed against von Willebrand Factor (anti-vWF), which is commonly used for endothelial cell identification (Fig. 3C). A functional endothelial assay was performed on subconfluent mDMEC cultures and demonstrated uptake of DiI-AcLDL (Figs. 3D and 4). Flow cytometric analysis To further address the endothelial phenotype and determine the degree of purity of the cell preparation, flow cytometry analysis was done using anti-CD31, CD34 and CD144 antibodies. The labeled fraction of the samples showed a significant increase in the fluorescence respect to controls (Fig. 4). The histograms shown reveled greater than 90% purity of the cell preparation (CD31, 91%; CD34, 93%; CD144, 92% and DiI-AcLDL, 99.8%). Gene expression studies Gene expression of endothelial-specific markers was analyzed by real-time RT-PCR. Fig. 5 shows the expression for four different endothelial markers (CD31, NOS3, VEGFR-2 and Tie-2). The presence of mRNA for each markers further confirmed the identity of the isolated cells. Discussion Murine microvascular endothelial cells have been difficult to culture for several reasons. Primary cultures are difficult to prepare and are often contaminated by fibroblast and other stromal cells. Furthermore, the available compositions of endothelial culture media do not fully satisfy the growth requirements of murine cells (Karasek, 1989; Imcke et al., 1991; Davison et al., 1980, 1983). The most common attempts to isolate murine microvascular endothelial cells have involved incubation of tissues in digestive enzymes and isolation using immunoabsorption to magnetic beads (Dong et al., 1997). However, in our experience, the use of magnetic beads results in low cell yields and remaining beads are difficult to
203
dissociate from the purified cells. Techniques for isolation of human dermal microvascular endothelial cells utilize neonatal foreskin or adult human skin (Karasek, 1989; Davison et al., 1980). These techniques rely on an initial separation of the epidermal layer from the dermis. Once of highly vascularized dermis is free of epithelial cells, it is treated with enzymes such as collagenase or trypsin and the vascular fragments are extracted. Here, we have adapted this methodology to isolate endothelial cells from murine skin. The separation of epidermal and dermal layer is only achievable before the murine skin matures and the hair follicles are present. We identified P3 as the ideal time to have a successful and complete separation of the dermis and epidermis. The collagen-rich dermis has abundant microvessels, therefore it is a good source of microvascular endothelial cells. Our strategy to isolate endothelial cells from the dermis includes an enzymatic extraction of the cellular components of the dermis followed by continuous density gradient centrifugation. With these methods, we have succeeded in isolating murine dermal endothelial cells. We report for the first time the detailed isolation strategy and also the culture conditions to sustain the growth of these cells. The advantage of generating primary endothelial cultures using the skin from individual mice is of particular interest to investigators using transgenic and null murine models where the availability of animals is limited and costly. Cultured mDMEC may be used to generate models to analyze basic aspects of microvascular biology and to investigate possible applications for tissue engineering. Acknowledgments The authors wish to thank A. Murad for technical assistance on gene expression experiments. This work was supported by funding from the Skirball Foundation to the Division of Plastic Surgery of Cedars-Sinai Medical Center and by NIH grant 1R01GM66292-01A1 to M.R.S.H. References Davison, P.M., Bensch, K., Karasek, M.A., 1980. Isolation and growth of endothelial cells from the microvessels of the newborn human foreskin and cell culture. J. Invest. Dermatol. 75, 316 – 321. Davison, P.M., Bensch, K., Karasek, M.A., 1983. Isolation and long-term serial cultivation of endothelial cells from the growth the microvessels of the adult human dermis. In Vitro 19, 937 – 945. Dong, Q.G., Bernasconi, S., Lostaglio, S., De Calmanovici, R.W., MartinPadura, I., Brevario, F., Garlanda, C., Ramponi, S., Mantovani, A., Vecchi, A., 1997. A general strategy for isolation of endothelial cells from murine tissues: characterization of two-endothelial cell lines from the murine lung and subcutaneous sponge implants. Arterioscler., Thromb., Vasc. Biol. 17, 1599 – 1604. Huang, H., McIntosh, J., Hoyt, D.G., 2003. An efficient, nonenzymatic method for isolation and culture of murine aortic endothelial cells and their response to inflammatory stimuli. In Vitro Cell. Dev. Biol.: Anim. 39, 43 – 50. Imcke, E., Ruszczak, Z., Mayer-da Silva, A., Detmar, M., Orfanos, C.E., 1991. Cultivation of human dermal microvascular endothelial cells in vitro: immunocytochemical and ultrastructural characterization and effect of treatment with three synthetic retinoids. Arch. Dermatol. Res. 283, 149 – 157.
204
S.T. Cha et al. / Microvascular Research 70 (2005) 198 – 204
Karasek, M.A., 1989. Microvascular endothelial cell culture. J. Invest. Dermatol. 93 (2 Suppl.), 33S – 38S. King, G.L., Buzney, S.M., Kahn, C.R., Hetu, N., 1983. Differential responsiveness to insulin of endothelial cells. Biochem. Pharmacol. 71, 974 – 979. Li, J.M., Mullen, A.M., Shah, A.M., 2001. Phenotypic properties and characteristics of superoxide production by mouse coronary microvascular endothelial cells. J. Mol. Cell. Cardiol. 33, 1119 – 1131. Lincoln-li, D.W., Larsen, A.M., Phillips, P.G., Bove, K., 2003. Isolation of murine aortic endothelial cells in culture and effects of sex steroids on their growth. In Vitro Cell. Dev. Biol.: Anim. 39, 140 – 145. Magid, R., Martinson, D., Hwang, J., Jo, H., Galis, Z.S., 2003. Optimization of
isolation and functional characterization of primary murine aortic endothelial cells. Endothelium 10, 103 – 109. Marelli-Berg, F., Peek, M., Lidington, E.A., Stauss, H.J., Lechler, R.I., 2000. Isolation of endothelial cells from murine tissue. J. Immun. Methods 244, 205 – 215. Marks, R.M., Czerniecki, M., Penny, P., 1985. Human dermal microvascular endothelial cells: an improved method for tissue culture and a description of some singular properties in culture. In Vitro 21, 627 – 635. Ruszczak, Z., 1996. Human skin microvascular endothelial cells. In: Bicknell, R. (Ed.), Endothelial Cell Culture, vol. 1. Cambridge Univ. Press, New York, pp. 77 – 89.