Epidermal stem cells: biology, carcinogenesis, function

皮肤是一种不断更新的组织，由大量增殖潜力有限的转运扩增（TA）细胞和少量增殖潜力高且具有克隆形成能力的角质形成细胞干细胞（KSC）组成。KSC更新干细胞群并产生TA细胞，TA细胞被移位到基底上层并因终末分化而丢失。在皮肤中，干细胞群驻留在毛囊隆起中，其位于毛囊的永久部分中，并且被保护免受物理损伤和毛囊从静止（休止期）到活跃生长（生长期）循环时所经历的变化。在众所周知的两阶段鼠表皮癌发生模型中，广泛认为KSC是主要致癌物靶细胞（即，潜伏的肿瘤细胞）。支持这一观点的一个关键因素是，DMBA启动的小鼠在暴露于肿瘤促进剂（如TPA）后会发生皮肤肿瘤，无论是在启动后一周还是一年。鉴于小鼠中角质形成细胞群体每6天更新一次，启动细胞持续存在的事实表明这些细胞必须是位于受保护微环境中的缓慢循环细胞。我们目前的重点是识别和表征的细胞，引起皮肤肿瘤的小鼠，以及调查的信号通路，有助于肿瘤的发展。作为该目的的一部分，我们研究了皮肤中的造血干细胞和祖细胞标志物CD 34，并将该标志物与α-6整联蛋白和荧光激活细胞分选（FACS）结合使用，已经表明CD 34特异性地标记毛囊隆起角质形成细胞，并有助于分离代表静止的α-6整联蛋白亮细胞亚群的活毛囊隆突角化细胞（即，主要在G1/G 0）。这项工作代表了第一次使用隆起特异性细胞表面标志物，允许直接阳性富集活的角质形成细胞干细胞和祖细胞。 最近，我们比较了从TPA处理或未处理的Tg.AC小鼠中收获的CD 34+角质形成细胞，以研究肿瘤促进后的差异基因表达模式。使用来自Clontech的尼龙cDNA阵列，用从TPA处理或未处理的皮肤分离的CD 34+细胞制备的基于PCR的SMART扩增的cDNA探测，鉴定了11个基因，其表达响应于TPA处理而显著改变。特别令人感兴趣的是裂手/裂足1（Dss 1），这是与异质性肢体发育障碍。通过RT-PCR和北方分析在TPA处理的皮肤（非肿瘤承载;与未处理的皮肤中的低水平相比）以及皮肤肿瘤（包括乳头状瘤、鳞状细胞癌和梭形细胞肿瘤）中检测到Dss 1和NDPK-B的过表达。功能研究表明，增加病灶形成活性和增殖的癌前表皮细胞组成性表达Dss 1。有趣的是，Dss 1诱导的转化稳定转染的JB 6表皮细胞被废除，通过添加蛋白激酶C（PKC）特异性抑制剂，暗示可能的PKC调节Dss 1表达的作用。两者合计，这些结果表明，Dss 1是TPA诱导的基因，可能在皮肤癌发生的早期阶段发挥重要作用。此外，最近的研究表明，NDBK-B可能参与化学诱导的皮肤癌发生的早期肿瘤发展。此外，我们利用高密度微阵列技术进一步进行了这些实验，比较了从TPA处理和丙酮对照处理的Tg.AC和野生型FVB/N小鼠分离的CD 34+和CD 34-细胞开发的基因表达谱。从这一点上，与未经治疗的小鼠产生的数据集进行比较，我们已经开发了一个在表皮干细胞中受到差异影响的早期反应基因的图片。 为了了解肿瘤发展中的关键信号通路，我们正在探索Ha-ras和p19 ARF在良性和恶性肿瘤发展中的关系，使用通过杂交p19 ARF-null和Tg.AC小鼠制成的双基因小鼠品系。皮肤肿瘤发生实验中的初步发现支持p19 ARF在肿瘤发展中的作用，但在这些研究过程中，在双基因小鼠中出现了一种意想不到的肿瘤类型，胃肠道间质瘤（GIST），这是人类高度侵袭性的肿瘤，在双基因小鼠中自发发生率很高，使其成为这种肿瘤类型的潜在小鼠模型。 最后，为了进一步评估毛囊表皮干细胞群在皮肤肿瘤发展中的作用，我们利用了表皮磨损技术，其中毛囊间表皮被物理去除。由此产生的表皮再生源自角质形成细胞从下面的毛囊中迁移出来。使用高密度微阵列分析从磨损后第3、5、9、18和30天从磨损的FVB/N（野生型）和Tg.AC小鼠（其在单次磨损后发展ras转基因依赖性乳头状瘤）收集的样品制备表达谱。我们已经开发了生物信息形成学策略，旨在从全局分析的角度探索这一复杂的数据集，从而更明智地关注可能有助于肿瘤发展和表皮再生的较小基因子集。

The skin is a continually renewing tissue consisting of a large population of transit amplifying (TA) cells with a limited proliferative potential, and a smaller population of keratinocyte stem cells (KSCs) that have a high proliferative potential and are clonogenic. KSCs renew the stem cell population and give rise to TA cells, which are displaced to the suprabasal layers and are lost by terminal differentiation. In the skin, the stem cell population resides in the hair follicle bulge, which is located in the permanent portion of the hair follicle and is protected from both physical damage and the changes the hair follicle undergoes as it cycles from resting (telogen) to active growth (anagen). In the well-known two- stage murine epidermal carcinogenesis model, it is a widely held belief that KSCs are the primary carcinogen target cells (i.e., latent neoplastic cells). A key factor supporting this belief is the fact that DMBA-initiated mice will develop skin tumors upon exposure to a tumor promoter such as TPA whether it is applied a week or a year after initiation. Given that the keratinocyte population renews itself every 6 days in the mouse, the fact that initiated cells persist suggests that these must be slowly cycling cells located in a protected microenvironment. Our current focus is on identification and characterization of the cells that give rise to cutaneous neoplasms in the mouse, as well as investigation into signaling pathways that contribute to neoplastic development. As part of this objective, we have investigated the hematopoietic stem and progenitor cell marker, CD34, in the skin and using this marker in combination with alpha-6 integrin and fluorescence activated cell sorting (FACS), have shown that CD34 specifically marks hair follicle bulge keratinocytes, and facilitates isolation of live follicular bulge keratinocytes that represent a subset of alpha-6 integrin bright cells that are quiescent (i.e., predominantly in G1/G0). This work represents the first use of a bulge-specific cell surface marker allowing for direct positive enrichment of live keratinocyte stem and progenitor cells. Recently we made a comparison was made between CD34+ keratinocytes harvested from either TPA-treated or untreated Tg.AC mice to investigate differential gene expression patterns following tumor promotion. Using nylon cDNA arrays from Clontech probed with PCR-based SMART-amplified cDNA prepared from CD34+ cells isolated from TPA-treated or untreated skin, eleven genes were identified whose expression changed significantly in response to treatment with TPA. Of particular interest was Deleted in Split Hand/Split Foot 1 (Dss1), which is associated with a heterogeneous limb developmental disorder. Overexpression of Dss1 and NDPK-B was detected by RT-PCR and Northern analysis in TPA treated skin (non-tumor bearing; compared to low levels in untreated skin), as well as in cutaneous tumors, including papillomas, squamous cell carcinomas, and spindle cell tumors. Functional studies revealed an increase in foci-forming activity and proliferation of preneoplastic epidermal cells constitutively expressing Dss1. Interestingly, Dss1 induced transformation of stably transfected JB6 epidermal cells was abrogated by addition of a protein kinase C (PKC) specific inhibitor, implicating a possible PKC regulatory role in Dss1 expression. Taken, together, these results suggest that Dss1 is a TPA-inducible gene that may play an important role in the early stages of skin carcinogenesis. In addition, recent studies have implicated a potential involvement of NDBK-B in early-stage neoplastic development in chemically-induced skin carcinogenesis. In addition, we have furthered these experiments utilizing high density microarray technology, comparing gene expression profiles developed from CD34+ and CD34- cells isolated from TPA treated and acetone control treated Tg.AC and wild type FVB/N mice. From this, comparing back to datasets generated from untreated mice, we have developed a picture of early response genes differentially affected in epidermal stem cells. To understand key signaling pathways in tumor development, we are exploring the relationship between Ha-ras and p19ARF in benign and malignant tumor development using a bigenic mouse strain made by crossing p19ARF-null and Tg.AC mice. Initial findings in skin tumorigenesis experiments supported a role for p19ARF in tumor development, but in the course of these studies, an unexpected tumor type developed in the bigenic mice, Gastrointestinal Stromal Tumors (GIST), which are highly aggressive tumors in humans, and which developed spontaneously at a high incidence in bigenic mice, making this a potential mouse model for this tumor type. Finally, to assess further the role of the hair follicle epidermal stem cell population in cutaneous tumor development, we have utilized the technique of epidermal abrasion, in which the interfollicular epidermis is physically removed. The resulting epidermal regeneration is derived from keratinocytes migrating out from the underlying hair follicles. Expression profiles using high-density microarray analysis were prepared from samples collected at days 3, 5, 9, 18, and 30 post-abrasion from abraded FVB/N (wild type) and Tg.AC mice (which develop ras-transgene dependent papillomas following a single abrasion). We have developed bioinoformatics strategies designed to probe this complex dataset in terms of global analysis, allowing for more informed focus on smaller subsets of genes that may contribute to tumor development and epidermal regeneration.