August 2019, Volume XXXIII, No 5


Skin cancer

A gateway to earlier interventions

significant number of skin cancers involve recruited cells from the bone marrow as well as from hair follicle stem cells, according to findings from 14 years of research by the authors and their collaborators around the country. The research used a bone marrow transplantation model to induce carcinogenesis with a hydrocarbon carcinogen and a plant compound as a tumor promoter in mice. The results show that some squamous skin tumors have a previously unrecognized systemic component, and that carcinogen-exposed bone marrow can develop both benign and malignant skin lesions upon tumor promotion.

The research could provide a gateway to studying new therapeutic targets to earlier interventions to stop the development of skin cancer and to treat non-melanoma skin cancers.


Non-melanoma skin cancers such as basal cell and squamous cell carcinomas occur more frequently in the human population than any other type of malignancy, and more than one million new cases are diagnosed in the United States annually. It is estimated that one-third to one-half of all human cancers originate in the skin; that skin cancers exceed all others combined; and that the lifetime risk for development of skin cancer in the U.S. population is one in five. Solar ultraviolet (UV) radiation is the major known cause of non-melanoma skin cancers and is directly relevant to the etiology of skin cancer, as demonstrated by both epidemiological evidence and the tight correlation between non-melanoma skin cancer development in humans and UV radiation-induced skin carcinogenesis in mouse experimental models.

One-third to one-half of all human cancers originate in the skin.

Skin cancers as seen in the clinic are actually the results of a long history of which only the later stages are easily observed. The progression from normal maintenance of the skin to neoplastic growth involves multiple genetic and epigenetic alterations, resulting in altered cellular phenotypes and gene expression patterns.

Epithelial cells covering the epidermis, or lining spaces in the gastrointestinal and respiratory tracts, can be identified by expression of certain cytoplasmic filament proteins, called keratins, or of cell surface marker proteins such as epithelial cell adhesion molecules (EpCAM). These cells can be tracked as they move around the bodies of genetically engineered mice that express green fluorescent protein (GFP) under the control of regulatory elements that govern the expression of specific epithelial cell markers, such as certain types of cytokeratin. This enables functional studies in which we can use GFP to visualize, isolate, and then culture these cells to investigate the origins of epithelial stem cells.

There are several reports of cells with epithelial markers in the blood and bone marrow (BM) of “normal, healthy human subjects”; however, the proliferative and differentiative capacities of these cells are as of this date unknown. These bone marrow-derived epithelial cells (BMDECs) are heterogeneous with regard to their repertoire of surface markers and cytokeratin expression. For the most part, they are considered “undesirable background” in the “liquid biopsy” that seeks to measure circulating metastatic cells in the blood and BM of cancer patients. Although their presence cannot be denied, their functional and phenotypic characteristics and other attributes are unknown.

Several attributes of blood-borne epithelial cells have been associated with epithelial cancers:

  1. We know that the incidence of circulating epithelial cells with characteristics of cancer cells increases with metastatic disease.
  2. Houghton and colleagues demonstrated that some bone marrow-derived cells (BMDCs) are recruited to and contribute to development of gastric cancer in a Helicobacter pylori-mediated mouse gastric cancer model.
  3. Pantel and associates showed that circulating epithelial cells increase in some benign proliferative diseases of the colon. Moreover, Tamai and colleagues have demonstrated that a sub-set of BMDCs is recruited to the epidermis in a mouse model of epidermolysis bullosa. Further, we have recently demonstrated that BMDCs contribute to both initiation and promotion of cutaneous neoplasms (Park, Morris, et al).

Taken together, these observations lead to the conclusion that it is essential to determine the proliferative and differentiative properties of circulating epithelial cells and bone marrow-derived epithelial stem cells.

Thus, BMDECs might serve as a reserve population of progenitor cells. Certainly, they are recruited to cutaneous papillomas and non-healing ulcers. Equally certainly, carcinogen-exposed BM can initiate benign and malignant cutaneous neoplasia upon skin tumor promotion. But, beyond carcinogenesis, do BMDECs play a role in chronic wound healing? Do they actually have properties of stem cells?

Some of these tumors may have a pronounced systemic component.

Research methodology

We began to lay the groundwork for this research upon reading the paper by Houghton et al., who reported that BMDCs contributed to gastric cancer in a mouse H. pylori model. We demonstrated that bone marrow cells (BMCs) also contributed significantly to non-melanoma skin cancer in mice, and in addition, proliferated, produced epidermal keratins, and stratified as in epidermal terminal differentiation. Furthermore, we demonstrated that cells in carcinogen-exposed BM are capable of becoming tumor-initiating cells. A critical unanswered question from this work is: what is the nature of the recruited cells? No defining functional studies addressing this hypothesis have been forthcoming.

We used gender-mismatched bone marrow transplantation (male GFP-positive BMCs to female recipients) in the context of a mouse model for skin carcinogenesis to determine the recruitment of Y-chromosome+ and GFP+ bone marrow cells (BMCs) in skin tumors. We detected a significant number of clusters of BMDCs in over 40% of benign papillomas where they occupied 25% or more of the lesional area. The BMDCs clustered in the cutaneous epithelium, where they expressed skin keratins, proliferated, stratified, and contributed to lesions along with the progeny of hair follicle stem cells, but as separate populations.

We also performed ex vivo co-culture experiments where we cultured adherent BMCs in the presence of filter-separated epidermal keratinocytes, and found that the presence of keratinocytes induced some of the BMCs to make epidermal keratins in the absence of cell contact or fusion.

Furthermore, when naïve female mice receive0d bone marrow transplants from donors previously treated with a carcinogen, several benign papillomas and a squamous cell carcinoma were observed after exposure to a skin tumor-promoting agent (which by itself does not cause cancer).


This research will have far-reaching consequences for the fields of hematology and epithelial biology. In the very near future, we will solve two important problems. First, we will detect and quantify BMDECs in untreated mice and healthy human subjects to determine their phenotypic characteristics. Second, we will determine whether the BMDECs have functional characteristics of epithelial stem/progenitors in ex vivo and in vivo models. These experiments have the potential to open up new avenues of research, and will provide a basis for the development of new disease models as well as a possible source of epithelial progenitors for regenerative medicine.

Our findings have significant implications for a better understanding of the etiology of epithelial cancers, as well as for the development of its diagnosis and treatment. The take-home message is that while some squamous skin tumors originate in the skin, some of these tumors may have a pronounced systemic component.

For additional information on this research initiative—led by the authors, Timothy Cragin Wang from Columbia University Medical Center, and others at multiple institutions—visit

Rebecca J. Morris, Phd, is an I.J. Holton Professor of Stem Cells and Cancer at The Hormel Institute/University of Minnesota, where she leads her laboratory. She has maintained an interest in keratinocyte stem cells and cancer from her graduate work at Syracuse University, and has been funded by grants from the ACS, NIAMS, DOD, and NCI.

Heuijoon Park, PhD, is a postdoctoral fellow at the Fred Hutch Cancer Research Center and lead researcher for the study referenced in this article. 


PO Box 6674, Minneapolis, MN 55406

(612) 728-8600

follow us

© Minnesota Physician Publishing · All Rights Reserved. 2019


Heuijoon Park, PhD, is a postdoctoral fellow at the Fred Hutch Cancer Research Center and lead researcher for the study referenced in this article.