Getting stronger: the relationship between a newly identified virus and Merkel cell carcinoma

Abstract

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin cancer of unknown etiology that develops on sun-exposed areas in individuals who are over 50 or immunosuppressed. DNA from a new polyomavirus, MCPyV, was recently shown to be clonally integrated into the host genome in several MCC cases. In this issue, Becker et al. show in a larger study that MCPyV DNA can be isolated from 85% of primary European MCC specimens and their metastases, while Gerneski et al. present data suggesting that the percentage of Australian MCC cases containing MCPyV may be lower than in North American cases. These reports strengthen the possibility that MCPyV may be etiologically involved in at least some cases of MCC.

Earlier this year, genetic analysis of Merkel cell carcinomas (MCC) revealed the existence of a previously unidentified human polyomavirus (5). DNA sequences from the virus, designated Merkel cell carcinoma polyomavirus (MCPyV), were detected in 8 out of 10 MCC tumor specimens but in only 4 out of 25 (16%) control skin samples. In this issue of the Journal of Investigative Dermatology, two letters confirm and significantly extend these findings to additional cases of MCC (1, 6). Becker et al. find, using PCR-based analysis, that 45/53 (85%) of MCC cases from a center in Europe contained MCPyV DNA, while Garneski et al. detected MCPyV DNA in 11/16 (69%) North American and 5/21 (24%) of Australian MCC specimens. The frequent detection of MCPyV DNA in European cases of MCC has also recently been confirmed by Kassem et al (10). The clear confirmation that most MCC tumors carry MCPyV sequences represents an important step in addressing the intriguing hypothesis that the virus is a key etiologic agent behind MCC and perhaps other forms of cancer.

Clinical and epidemiological features of MCC

Merkel cells were identified in 1875 as a distinctive, histologically translucent cell type associated with epidermal nerve endings (8). At the time of their discovery, it was reasoned that Merkel cells might participate in sensation of mechanical stimuli. Like melanocyte precursors, the precursors of Merkel cells are thought to migrate during development from the neural crest to their ultimate home in the basal layer of the epidermis and the outer root sheath of hair follicles (14). Although Merkel cells do exhibit a number of neuroendocrine features, such as cytoplasmic neuropeptide-containing granules, the idea that they are directly involved in mechanosensation remains inferential (2).

There are approximately 1500 new MCC cases per year diagnosed in the U.S. (11). MCC is thus a rare form of skin cancer compared to the more than one million cases of non-melanoma skin cancer and roughly 60,000 cases of melanoma per year in the same population. However, MCC is an aggressive, fast-growing malignancy that ultimately kills about one-third of those diagnosed with the disease. It is also important to note an alarming three-fold increase in the incidence of MCC between 1986 and 2001 (9).

MCC is more common in lighter-skinned individuals, is associated with a history of sun exposure, and occurs predominantly on sun-exposed areas, such as the face (especially around the eyes) and extremeties (12). Limited mutational screening of cellular genes in MCC identified the presence of two C to T transitions, characteristic of UVB-induced mutation, in one MCC cell line (15). In their current Letter, Garneski et al. speculate that the lower percentage of MCPyV in the Australian MCC cases they studied (24% vs. 69% in the North American cases, p=0.009) might be related to the greater degree of sun exposure in the Australian population - thus leading to a greater fraction of MCPyV-independent MCC - compared with the North American and European populations.

The MCC tumor typically presents as a firm, painless non-pigmented nodule that may be pink to violet in color. As with keratoacanthomas, MCC tumors can grow rapidly over the course of as little as a few weeks. 95% of MCC cases arise in individuals over age 50 (7). Although the relative risk of MCC is dramatically increased in immunocompromised populations, such as organ transplant recipients and patients with AIDS (4), the great majority of newly diagnosed MCC patients are not recognized clinically as immunocompromised. It is, however, conceivable that the waning immunity typically associated with advancing age could be a contributory factor in the development of MCC in older individuals. Given the established correlation between immunodeficiency and the occurrence of cancers with known viral etiology (for example, human herpes virus-8-induced Kaposi’s sarcoma and human papillomavirus (HPV)-induced squamous cell cancers in epidermodysplasia verruciformis (EV) and cervical cancer), the correlation between MCC and immunodeficiency can be seen as a possible clue indicating a viral etiology.

What are polyomaviruses?

MCPyV belongs to a diverse family of non-enveloped DNA viruses that infect various vertebrates. A number of polyomavirus types, including simian virus-40 (SV40) and J-C polyomavirus (JCV), have long been known to infect human populations. Early clues had suggested the possible existence of unidentified MCPyV-related polyomaviruses in humans (3), and a number of recent reports have identified additional human polyomaviruses from various phylogenetic branches of the family (reviewed in (16)). Although polyomaviruses can induce malignant transformation in cultured cells and in animal models, links between previously identified human polyomaviruses and cancer have been tenuous, at best. Thus, the apparently strong correlation between MCPyV and MCC has generated a significant resurgence of interest in the possible viral etiology of MCC and perhaps other cancer types.

What is the role of MCPyV in MCC?

As with most important research, the observation that most MCC tumors harbor MCPyV DNA raises important questions for subsequent studies. One important question is whether MCPyV is an etiological cause of MCC, as opposed to a mere hitchhiker in this type of tumor. The available data suggest that PCPyV will turn out to be pathogenetically involved in at lease a fraction of MCC cases. Feng et al. reported clonal integration of PCPyV DNA in the host genome in 6 of the 8 virus-positive tumors they studied, implying that the viral DNA was integrated early in the course of tumor development. Juergen et al. found that 25/25 mestastases from virus positive primary tumors contained MCPyV DNA, while 0/3 metastases from virus-negative primary tumors contained MCPyV DNA, strongly suggesting that the virus-positive metastases were derived from cells that harbored viral DNA in the primary tumors. It will be important for future studies to determine whether viral infection precedes tumor development, if viral gene expression is required for tumor maintenance, and whether MCPyV is tumorigenic in experimental systems. If the viral DNA is necessary for maintenance of the malignant phenotype, the development of targeted therapies against pathologically relevant viral genes would be a desirable goal. It would also be important to test whether MCPyV might be implicated in other tumor types and, perhaps, whether other polyomaviruses cause human tumors (16).

Is MCPyV an animal virus or a human virus?

Another key question is whether MCPyV is endemic in the human population or primarily an animal polyomavirus for which humans are an accidental host. Partial support for the latter possibility is that the polyomavirus SV40 does not appear to cause disease in its natural host, the Rhesus monkey, but can cause experimental tumors in heterologous hosts, such as hamsters. An alternate scenario is that MCPyV circulates widely in human populations and simply causes few apparent symptoms in the great majority of infected individuals. A precedent for this scenario can be found in the epidemiology of EV-type HPVs, which infect a great majority of humans asymptomatically worldwide and cause squamous cell carcinomas in only a small minority of infected individuals (13).

Answers to these questions and more will no doubt be fertile areas of research in years to come.