A protocol for enrichment of CD34+ stromal cell fraction through human skin disaggregation and magnetic separation

Abstract

In normal skin, CD34 staining is limited to dendritic, endothelial and perifollicular cells, as well as discrete areas of the hair follicles, eccrine glands and nerve fascicles [ [1] Poblet E. Jimenez F. Godinez J.M. Pascual-Martin A. Izeta A. The immunohistochemical expression of CD34 in human hair follicles: a comparative study with the bulge marker CK15. Clin Exp Dermatol. 2006; 31: 807-812 Crossref PubMed Scopus (75) Google Scholar ]. CD34+ stromal cells are thought to play a role in the maturation and proliferation of adjacent mesenchymal and epithelial stem cells as well as in local immune responses [ [2] Erdag G. Qureshi H.S. Patterson J.W. Wick M.R. CD34-positive dendritic cells disappear from scars but are increased in pericicatricial tissue. J Cutan Pathol. 2008; 35: 752-756 Crossref PubMed Scopus (21) Google Scholar ]. However the function of CD34+ cells in the dermis is as yet unclear, and a reliable isolation protocol is needed to further explore their proposed physiological roles. Perifollicular CD34+ cells may interact with epithelial stem cells of the hair follicle bulge [ [3] Nickoloff B.J. The human progenitor cell antigen (CD34) is localized on endothelial cells, dermal dendritic cells, and perifollicular cells in formalin-fixed normal skin, and on proliferating endothelial cells and stromal spindle-shaped cells in Kaposi's sarcoma. Arch Dermatol. 1991; 127: 523-529 Crossref PubMed Scopus (342) Google Scholar ]. Possible relationships with desmoplastic stroma formation around some cutaneous epithelial tumors [ 4 Kirchmann T.T. Prieto V.G. Smoller B.R. CD34 staining pattern distinguishes basal cell carcinoma from trichoepithelioma. Arch Dermatol. 1994; 130: 589-592 Crossref PubMed Scopus (136) Google Scholar , 5 Tardio J.C. CD34-reactive tumors of the skin. An updated review of an ever-growing list of lesions. J Cutan Pathol. 2009; 36: 89-102 Crossref PubMed Scopus (56) Google Scholar , 6 Tellechea O. Reis J.P. Ilheu O. Baptista A.P. Dermal cylindroma. An immunohistochemical study of thirteen cases. Am J Dermatopathol. 1995; 17: 260-265 Crossref PubMed Scopus (43) Google Scholar ] as well as an implication in collagen synthesis regulation [ 2 Erdag G. Qureshi H.S. Patterson J.W. Wick M.R. CD34-positive dendritic cells disappear from scars but are increased in pericicatricial tissue. J Cutan Pathol. 2008; 35: 752-756 Crossref PubMed Scopus (21) Google Scholar , 7 Aiba S. Tabata N. Ohtani H. Tagami H. CD34+ spindle-shaped cells selectively disappear from the skin lesion of scleroderma. Arch Dermatol. 1994; 130: 593-597 Crossref PubMed Scopus (71) Google Scholar ] have also been postulated. To evaluate the abundance of CD34 cells in human skin, biopsies were disaggregated to single cell suspensions and analyzed directly by flow cytometry (Appendix A1). FACS analysis of 44 human biopsies (Table S1) showed the presence of a relevant percentage CD34+ cells (14.7 ± 8.4%). To discard the presence of endothelial and/or haematopoietic cells within the CD34+ cell fraction, we analyzed co-expression of CD34 antigen with CD31 and CD45. CD34+ cells did not express CD31 or CD45 in meaningful numbers (Fig. S1). These results demonstrated presence of a surprisingly abundant CD34+ cell fraction in human skin, apparently not attributable to migrating blood cells or small endothelial vessel constituents, although we cannot formally exclude an influence of our chosen protocol in the isolation of a particular subset of CD34+ cells. In order to assess the dermal and/or epidermal origin of CD34+ cells, we splitted skin biopsies (n = 3) into three separate but identical fragments that were then subjected to different cell isolation protocols (Fig. 1). Cell fractions obtained through different procedures as outlined in Fig. 1A were then analyzed through tissue dissociation and flow cytometry of the resulting cell suspensions (Fig. 1B and data not shown). Epidermal (0.15 ± 0.08%) and trypsin (1.98 ± 1.37%) fractions presented a markedly lower CD34+ cell percentage than dermal (5.10 ± 3.48%) and collagenase (16.00 ± 4.72%) fractions. Statistically significant differences were detected in CD34+ cell abundance among epidermal, dermal and trypsin fractions, as compared to collagenase fraction cells (p = 0.000036, p = 0.0012 and p = 0.00035, respectively). No statistically significant difference was detected among epidermal–dermal (p = 0.11) or trypsin-dermal fractions (p = 0.97), nor among epidermal–trypsin fractions (p = 1). These results suggested that most CD34+ cells isolated from whole skin seem to have a dermal origin. To better characterize CD34± skin cell fractions, we magnetically isolated both fractions as well as the Unsorted population as a positive control (Fig. 2A ). Isolation protocol was optimized to achieve purity over 85% in the CD34+ fraction (n = 22, data not shown). However purified CD34+ cell recovery was low (5.8 ± 3.3% of CD34+ fraction of initial cell population, n = 16; not shown). Cell fraction composition was further characterized both by flow cytometric and immunofluorescence analyses (Fig. 2B). In a representative example, 17% of the Unsorted population was positive for CD34 (Fig. 2B; b2). After isolation with magnetic beads, the expression of CD34 increased to 93% in the CD34+ cell fraction (Fig. 2B; b4), while the CD34− fraction still presented 8% of CD34+ cells (Fig. 2B; b6). CD34 expression in the cell membrane was confirmed by immunofluorescence and confocal microscopy. CD34+ cells were mainly detected in the Unsorted (Fig. 2B; b7) and CD34+ fractions (Fig. 2B; b9), while CD34− cells were present both in the Unsorted (Fig. 2B; b7) and in the CD34− (Fig. 2B; b11) fractions, as expected. When cell fractions were put in standard keratinocyte culture conditions, both the Unsorted and the CD34− cell populations grew consistent keratinocyte colonies (Fig. 2C; c1 and c3). In contrast, keratinocytes of the CD34+ fraction grew only in 1/14 independent experiments (Fig. 2C; c2 and c4), suggesting the presence of a high percentage of dermal cells within the CD34+ fraction that impeded keratinocyte culture development. To confirm these results, CD34± and Unsorted cell fractions were subjected to flow cytometric and immunofluorescence analyses with anti-vimentin antibodies (Fig. S2). CD34+ cell fraction was mainly vimentin positive (96–98% of double positive cells; Fig. S2, a8), suggesting that it almost entirely consisted of cells with a dermal origin. In contrast, both the Unsorted and CD34− fractions presented two distinct cell populations, one of dermal cells (50.8 and 25.7% of vimentin+ cells, respectively; Fig. S2, a7, a9) and one of cells of non-dermal origin (vimentin−). Presence of vimentin in the cell cytoplasm of all fractions was also confirmed by immunofluorescence and confocal microscopy analyses. These data confirmed that collagenase digestion followed by magnetic cell separation with anti-CD34 antibody isolated a population that almost entirely consisted of CD34+ cells of a dermal origin; although a minor keratinocyte proportion was undoubtedly present as evidenced by further purification downstream [ [8] Gutiérrez-Rivera A. Pavón-Rodríguez A. Jiménez-Acosta F. Poblet E. Braun K.M. Cormenzana P. et al. Functional characterization of highly adherent CD34 positive keratinocytes isolated from human skin. Exp Dermatol;. 2010; (in press) Google Scholar ]. In summary, clinical separation of CD34+ cells of the haemopoietic system is now a semi-automated process with systems commercially available that achieve 90–97% cell purity and 60–80% cell recovery. These clinical grade systems have already been used to isolate CD34+ microvascular endothelial cells from solid tissue [ [9] Arts C.H. de Groot P. Heijnen-Snyder G.J. Blankensteijn J.D. Eikelboom B.C. Slaper-Cortenbach I.C. Application of a clinical grade CD34-mediated method for the enrichment of microvascular endothelial cells from fat tissue. Cytotherapy. 2004; 6: 30-42 Abstract Full Text Full Text PDF PubMed Scopus (11) Google Scholar ]. Initially, we wanted to take advantage of this fact to improve isolation of human epidermal stem cells for clinical application, since CD34 is a canonical marker for murine hair follicle stem cells [ [10] Trempus C.S. Morris R.J. Bortner C.D. Cotsarelis G. Faircloth R.S. Reece J.M. et al. Enrichment for living murine keratinocytes from the hair follicle bulge with the cell surface marker CD34. J Invest Dermatol. 2003; 120: 501-511 Crossref PubMed Scopus (487) Google Scholar ]. We have since purified CD34 positive keratinocytes of human hair follicles through a distinct enrichment protocol and shown that they might be a transit-amplifying precursor for hair follicle sheath cells instead [ [8] Gutiérrez-Rivera A. Pavón-Rodríguez A. Jiménez-Acosta F. Poblet E. Braun K.M. Cormenzana P. et al. Functional characterization of highly adherent CD34 positive keratinocytes isolated from human skin. Exp Dermatol;. 2010; (in press) Google Scholar ]. Taking into account that physiological role of CD34+ stromal cells deserves further investigation, we conclude in this manuscript that disaggregation of human whole skin biopsies by collagenase protocol and subsequent isolation for CD34 antigen through magnetic beads permits substantial enrichment for CD34+ stromal cells. In our hands, magnetic isolation of CD34 positive cells permitted reasonable purity (85–90%), although rather bland recovery rates (5.8 ± 3.3%; n = 16). Our method is thus amenable for improvement. However it represents a novel tool for the phenotypic and functional characterization of live cell fractions from human skin.