Gordon Wilson Lecture: Infectious Disease Causes of Cancer: Opportunities for Prevention and Treatment

Abstract

The role of infectious agents in cancer is generally underappreciated. However, approximately 20% of human cancers are caused by infectious agents and as such they rank second only to tobacco as a potentially preventable cause in humans. Specific viruses, parasites, and bacteria have been linked to specific human cancers. The infectious etiology for these specific cancers provides opportunities for prevention and treatment.

INTRODUCTION

The mechanisms by which an infectious agent can cause a cancer vary. Some infectious agents that cause persistent infection and chronic inflammation lead to the formation of reactive oxygen and nitrogen species by macrophages at the site of the infection leading in turn to cell damage and cellular proliferation. Proliferating cells can then acquire mutations contributing to cancer initiation and progression. A more direct mechanism involves expression of oncogenic genes by the infectious agent, an example of which would be the human papillomaviruses (HPVs) and cervical cancer described below.

In addition, some infectious agents, such as the human immunodeficiency virus (HIV), cause immunosuppression impairing the immunologic recognition of the infected or transformed cell.

The history of the role of infectious agents and cancer indeed has a storied past. A Nobel Prize was awarded in 1926 to Johannes Fibiger, a Danish scientist and physician, for his discovery that a parasite caused gastric cancer in rodents. The problem was that the experiments could not be repeated and the lesions that were induced by dietary deficiency were papillomas, not cancer. Today, however, it is recognized that infectious agents account for approximately 20% of the global burden for cancer.

BACTERIA AND CANCER

The best example of a bacteria and cancer relationship is Helicobacter pylori that were rediscovered by Robin Warren and Barry Marshal, who first cultured the bacterium and determined it to be the cause of most gastric and duodenal ulcers (1,2). Two different human cancers are linked to H. pylori infection: gastric cancer and mucosa-associated lymphoid tissue lymphoma of the stomach. H. pylori encodes potentially oncogenic factors that affect cell signaling pathways (3). In addition, H. pylori induces chronic inflammation and expresses a factor, VacA, a secreted vacuolating cytotoxin protein that inhibits the ability of T-lymphocytes to neutralize infection and allows the bacterium to evade the immune system to establish a chronic infection (4). Warren and Marshall shared the 2005 Nobel Prize in Medicine for their discovery of H. pylori and linking it to ulcers and cancer.

PARASITES AND CANCER

Parasites have also been linked with human cancer. In 1900, Askanazy reported a link between Opisthorchis felineus infection with liver cancer, and Goebel published a report incriminating Bilharzia infections (schistosomiasis) with human bladder cancer (5).

Two well-established associations of parasites with human cancer are Shistosomiasis with bladder cancer and liver flukes with cholangiocarcinoma (cancer of the bile duct). The major burden for parasite associated cancers is in the developing world (4).

VIRUSES AND CANCER: HISTORICAL PERSPECTIVE

Viral oncology as a discipline dates back to the early 20th century with studies in birds. Avian leukemia was first described in 1908 and the transmissibility of an avian sarcoma in chickens was described by Peyton Rous in 1911 (6,7). The full impact of these early studies, however, was not appreciated at the time, and the discovery of Rous of a viral etiology of a sarcoma in chickens was recognized by a Nobel Prize in 1966, more than 50 years after the discovery. The impact of these early studies was not appreciated until evidence established that viruses also had a role in cancers in mammals.

Richard Shope was the first to identify mammalian viruses that could cause tumors in mammals with his work with the Shope fibroma virus and the Shope papillomavirus in rabbits. Working together with Rous they established that the papillomas caused by the cottontail rabbit papillomavirus could progress to cancer (8,9). The major advance in the field of viral oncology came in the 1950s with the discoveries of Ludwig Gross and others of the murine leukemia virus and the murine polyomavirus (10–12). The identification of tumor viruses in mice began the field of experimental viral oncology and suggested that viruses might also have a role in cancers in humans.

VIRUSES AND HUMAN CANCER

The identification of a herpesvirus in cultured tumor cells from an African Burkitt's lymphoma by Anthony Epstein in 1964 marks the beginning of viruses and human cancer (13). The Special Viral Cancer Program was created in the early 1960s at the National Cancer Institute (NCI) from this interest in viral oncology and the belief that human tumor viruses would be identified (4). Much of the early focus on human cancer viruses was on the identification of human retroviruses, based to some degree on the identification of retroviruses associated with cancers in animals.

The early work in viral oncology that was supported by the NCI in the 1960s and 1970s drove some of the most important developments in modern molecular biology. These include the discovery of reverse transcriptase for which David Baltimore and Howard Temin shared the Nobel Prize, the discovery of messenger RNA splicing from studies of oncogenic adenoviruses, the discovery of oncogenes from studies of transforming retroviruses, and the discovery of p53 from studies of the transforming proteins encoded by the DNA tumor viruses (14).

Table 1 provides a list of human viruses associated with human cancer. Also listed in Table 1 are co-factors that are important in the carcinogenic processes associated with each of these viruses. These viruses associated with these human cancers are thought to be involved at an early step in carcinogenesis. Subsequent cellular genetic events such as somatic mutations are thought to then be important at the subsequent steps involved in the multi-step process of malignant progression (4). These viruses are discussed in more detail below.

TABLE 1.

Human Cancer Viruses

Virus	Cancer	Co-factors
Epstein-Barr virus	Burkitt's lymphoma	Malaria
	Nasopharyngeal carcinoma	Nitrosamines, HLA subtype
	Immunoblastic lymphoma	Immunodeficiency
	? Hodgkin's disease
	? Other cancers
Hepatitis B virus	Hepatocellular carcinoma	Aflatoxin, alcohol, smoking
Hepatitis C virus	Hepatocellular carcinoma	?
HTLV-1	Adult T-cell leukemia	Uncertain
Beta genus HPV types (5, 8, 17, . . . . .)	Skin cancer	Genetic disorder (EV), sunlight, immunodeficiency
Alpha genus HPV types (16, 18, 31, . . . . .)	Anogenital cancers	Smoking, oral contraceptives,other factors
Human herpesvirus 8 (KSHV)	Kaposi's sarcoma Primary effusion lymphomas	Immunodeficiency
Merkel cell polyomavirus	Merkel cell cancer	Age, immunodeficiency

Abbreviations: HTLV-1, human T-cell leukemia virus type 1; HPV, human papillomavirus; EV, epidermodysplasia verruciformis; KSHV, Kaposi's sarcoma−associated herpesvirus.

EPSTEIN-BARR VIRUS

The Epstein-Barr virus (EBV) was discovered in 1964 in cultures of a lymphoma that had been described as a unique entity with characteristic clinical, pathologic, and epidemiologic features by Denis Burkitt in 1958 (15,16). As such, EBV was the first virus to be recognized as a human tumor virus. EBV is also a ubiquitous virus and then 90% of people worldwide are infected before adulthood. EBV is the cause of infectious mononucleosis (17).

Endemic Burkitt's lymphoma (BL) that occurs in the malaria belt is an EBV-positive malignancy occurring several years after the primary infection with EBV. BL is a monoclonal lymphoma of B cell origin. The proliferation of the B-lymphocytes and the reduction in virus-specific cytotoxic T cells associated with malaria is believed to result in an increased EBV viral load enhancing the risk of chromosomal c-myc translocations characteristic that are of BL (18). The failure of the T-cell immune response to control this proliferation might be an early step providing the enhanced opportunity for further mutation, oncogenic transformation, and lymphomagenesis in the actively dividing B cell population. As such, the oncogenic activity of EBV in BL may not be direct, but instead the virus may serve as a polyclonal B cell mitogen, setting the stage for chromosomal translocations that deregulate c-myc expression as well as other chromosomal mutations.

EBV is associated with other human malignancies, including nasopharyngeal cancers, B cell lymphomas in immunosuppressed individuals, some types of Hodgkin's disease, and some gastric cancers (4). The molecular basis of EBV-associated malignancies is complex and likely is different for the different cancers. EBV, however, does encode genes that contribute to the malignant phenotype. These genes (including EBNA-2 and LMP-1) activate cellular signaling pathways and contribute to cellular immortalization (17).

VIRAL CAUSES OF LIVER CANCER

Liver cancer or hepatocellular carcinoma (HCC) globally is one of the commonest malignancies, particularly in China. HCC is etiologically linked to infections by two different types of viruses, hepatitis B virus (HBV) and hepatitis C virus (HCV). In the 1970s, the distribution of HCC was recognized to mirror the distribution of chronic HBV infection. Epidemiologic studies in Taiwan showed that chronic infection with HBV leading to cirrhosis contributed to the etiology of HCC (19). It is estimated that 80% of HCC worldwide occurs in individuals chronically infected by HBV (28). A number of risk factors, including chronic hepatitis associated with HCV, account for the remaining cases of HCC.

HBV is a hepadnavirus that is a partially double stranded circular DNA virus with a replication cycle similar to that of a retrovirus. It does not appear to encode a gene that contributes directly to carcinogenic progression. Evidence suggests that HBV causes cancer through chronic inflammation, which in turn leads to persistent cellular proliferation (20). An effective preventive vaccine for HBV was approved in 1981, and a reduction in HCC among vaccinated populations proves the role of HBV in HCC (21).

HCV is an RNA virus (flavivirus) and is an important cause of morbidity and mortality worldwide. HCV was first cloned in 1989 from the infectious sera of individuals with non-A, non-B post-transfusion hepatitis (22,23). Similar to HBV, it appears to cause cancer indirectly through inflammation and inducing cellular proliferation. However, there is some evidence that HCV encodes genes that might contribute to the initiation of the cancer with activities that inhibit apoptosis or cooperate with cellular oncogenes (24). Although no vaccine has yet been developed to prevent HCV infection, several very effective drug combinations have been approved by the US Food and Drug Administration (FDA) during the past year for treating infected individuals.

HUMAN T-CELL LEUKEMIA VIRUS TYPE 1 AND ADULT T-CELL LEUKEMIA

Human T-cell leukemia virus type 1 (HTLV-1) was the first human retrovirus identified; it was isolated from human T-cell leukemia cells in 1980 by Robert Gallo at NCI (25,26), and independently by Yoshida in Japan (27). The HTLV-1 is recognized as the etiologic agent of adult T-cell leukemia (ATL), which has high prevalence in southwest Japan, the Caribbean, and other parts of the world where the virus is prevalent. HTLV-1 encodes an oncoprotein called Tax that functions by targeting and activating a number of cellular transcription pathways including NF kappa B (28).

A causal relationship between HTLV-1 and ATL was initially suggested by epidemiologic studies showing geographic clustering of ATL, a pattern that is consistent with an infectious agent. A second human retrovirus, referred to as HTLV-2 was initially isolated in 1982 from a cell line established from a patient with an unusual form of hairy cell leukemia (29). However, studies have not established an association of HTLV-2 with any human malignancy. HTLV-1 also encodes a protein called the HTLV-1 basic leucine zipper factor (HBZ) that is expressed in ATL cells and promotes proliferation of T-cells (30). The data suggest that both Tax and HBZ play direct roles in leukemogenesis (28).

KAPOSI'S SARCOMA-ASSOCIATED HERPESVIRUS

In 1994 a new herpesvirus, known as the Kaposi's sarcoma-associated herpesvirus (KSHV) or human herpesvirus 8 (HSV8), was identified by Yuan Chang and Patrick Moore by representational difference analysis of Kaposi's sarcoma (KS) skin lesion in an acquired immune deficiency syndrome (AIDS) patient (31). KS was originally described and recognized as a rare, indolent tumor of elderly Mediterranean men by Moritz Kaposi in the 19th century. A major feature of KS is abundant endothelial cell proliferation. KS had also been observed among immunosuppressed organ transplantation patients and was initially the most common neoplasm associated with the AIDS. With effective HIV therapies, the incidence of KS is decreasing. KSHV encodes a number of genes that inhibit apoptotic pathways, activate cellular proliferation, and induce angiogenesis (32,33).

HUMAN POLYOMAVIRUSES AND CANCER

There are reports in the literature dating back to the 1970s claiming the presence of SV40 DNA and DNA from the human polyomaviruses BK and JC in various human cancers. In general, these have not been well substantiated and some of the SV40 DNA positivity reported may be related to laboratory contamination and polymerase chain reaction amplification of expression vector plasmid DNA containing SV40 DNA elements (34,35).

However, in 2008, Yuan Chang and Patrick Moore did identify a new human polyomavirus by deep sequencing techniques from human Merkel cell carcinomas (MCCs) (36).

MCC is a relatively rare, but aggressive skin cancer in humans described in 1972 by Cyril Toker (37). MCC is seen most often in immunosuppressed individuals and the elderly. A high percentage of MCCs contain Merkel cell polyomavirus (MCV) DNA that is integrated in a clonal manner and that is transcriptionally active, suggesting an etiologic role in the cancers. Furthermore, the expression of the virally encoded tumor antigens is required to maintain the cancer cells in tissue culture, strongly implicating the virus as the cause of these cancers. A portion of the early region encoding transforming proteins is expressed in the MCV-positive cancers (38).

HUMAN PAPILLOMAVIRUSES AND CANCER

HPVs cause benign lesions, and a subset HPVs is associated with lesions that can progress to specific human cancers. There are approximately 200 different HPVs that have been identified and have been classified by their genome relatedness into five different phylogenetic genera (39). The best studied of the HPVs are those of the alpha genus that includes HPVs that cause squamous mucosal lesions including those of the genital tract. The beta genus HPVs are associated with cutaneous lesions and are now gaining interest because of their potential role in non-melanoma skin cancers. These are viruses that were first described in lesions and skin cancers in patients with the rare genetic disorder epidermodysplasia verruciformis (EV) (40).

My own laboratory began studying the papillomaviruses in the late 1970s. Early work from my laboratory included the initial cloning of papillomavirus genomes (41,42), the demonstration of conserved sequences among papillomavirus (43), and the development of reverse genetic strategies to study them since the papillomaviruses had not yet been successfully propagated in the laboratory. Much of the initial work in the field focused on the bovine papillomavirus since it could induce tumors in rodents and could readily transform rodent cells in tissue culture (44,45). The focus of papillomavirus research in my lab shifted to the human papillomaviruses after the identification of specific HPVs (HPV types 16 and 18) in cervical cancer by Harald zur Hausen et al in 1983 and 1984 for which he was awarded the Nobel Prize in 2008 (46,47).

The alpha genus, mucosal HPVs are classified as either high risk (hr) or low risk (lr) based on their potential association with cancer. HPV6 and -11 cause venereal warts and are therefore lr HPVs. The hr HPVs cause squamous intraepithelial lesions that can progress to cancer and include HPV16 and -18 but also approximately a dozen other types (4). Virtually all cases of high grade cervical intraepithelial neoplasia (CIN) and cervical cancer contain a high-risk HPV DNA (4). HPV-positive cervical cancers and derived cell lines generally contain integrated HPV DNA that express two viral genes, E6 and E7, which function as oncogenes. My laboratory contributed to studies identifying the mechanistic role by which these high-risk, cancer associated HPVs contribute to malignant progression. The hr E6 and E7 oncogenes function in part by inactivating two important tumor suppressor gene products, p53 and pRB, respectively (48,49). The inactivation of p53 and pRB in turn cause genomic instability enabling the acquisition of additional genomic mutations for cancer progression.

In addition to cervical cancer and other anogenital cancers, the high-risk HPVs have also now been associated with head and neck cancers, in particular oropharyngeal cancers (OPCs). Of these, HPV16 is found in approximately 90% of the HPV-positive tumors. A dramatic 3-fold increase in the incidence of these HPV-positive OPCs occurred between 1988 and 2004 (50). HPV-positive OPCs usually develop at a younger age and are less often associated with cigarette smoking than are the HPV-negative cancers. As in the anogenital HPV-associated cancers, the HPV-positive OPCs express E6 and E7 that target p53 and pRB genes that remain wild type (51), in contrast to the HPV-negative tumors in which p53 gene is usually mutated.

A major advance in the prevention of human cancer has been the development of an effective preventive vaccine for the major hr HPVs. The vaccine is a sub-unit vaccine consisting of the major capsid protein (L1) that self-assembles when expressed in eukaryotic cells into virus-like particles (VLPs), which are empty capsids with morphologic and immunologic characteristics similar to authentic virions (52). L1 VLPs induce high titers of neutralizing antibodies that are type-specific. The FDA has approved two commercial VLP-based HPV vaccines. Cervarix, made by GlaxoSmithKline (Brentford, England) is a bivalent vaccine composed of L1 VLPs of HPV16 and 18, and Gardasil (Merck, Kenilworth, NJ, USA) is a quadrivalent vaccine composed of L1 VLPs of HPV6, 11, 16, and 18. These vaccines are safe and highly effective at preventing acquisition of cervical infection and low- and high-grade CIN caused by the types targeted by the vaccine (53). Limitations of the current vaccines are the high level of type specificity that is less effective at preventing infections by other hr HPVs. Although 70% of cervical cancers are caused by HPV16 or HPV18, 30% are caused by the other high-risk HPV types. Strategies have been developed for next-generation VLP vaccines that would protect against an even higher proportion of HPV infection by incorporating VLPs from more hr HPV types. Also, experimental vaccine strategies are being developed that target epitopes in L2 that are highly conserved among many if not all HPVs (54). Although not as immunogenic as the L1 neutralization epitopes, L2 neutralization epitopes induce cross-neutralizing antibodies against many papillomaviruses (55,56).

Although national programs have been centered on vaccination of young girls, from ages 9 to 15 years, Gardasil protects men from genital warts and anal cancer precursors providing rationale for including males in vaccination programs. Furthermore, the increase in HPV-positive head and neck OPC in both men and women provides strong rationale for including men in vaccination programs.

IS THERE A ROLE FOR A THERAPEUTIC TARGET FOR THE HIGH-RISK HPVs?

One shortcoming of the current VLP vaccines is the lack of a therapeutic effect for the millions of already infected individuals. Also, in the United States, HPV vaccination is not mandatory and only about one-third of girls/women of the age to receive the vaccine have received all three doses. Finally, since the major burden for cervical cancer is in the developing world, shortcomings that affect the global impact of the current vaccine include its significant cost and the current recommendation of three separate injections. Thus, there is an unmet need for HPV-specific therapeutics to treat associated cancers and pre-cancers.

The viral E6 and E7 proteins account for the oncogenic potential of high-risk HPVs and it is these two viral genes that are invariably expressed in HPV-positive cancers and therefore provide a therapeutic strategy for HPV-associated cancers. As noted above, the major cellular targets for E6 and E7 are the tumor suppressor proteins p53 and pRB, respectively. E6 and E7 are, however, polyfunctional proteins, and have other biological properties that may be relevant to their oncogenicity (57). Genetic studies indicate that E7 binding to pRB and its related pocket proteins is not sufficient to account for its immortalization and transforming functions, indicating that there are additional cellular targets and activities that are relevant to cellular transformation. A large number of putative cellular targets for E7 have been identified; however, the physiologic relevance of many of these interactions is as yet unclear (58). High-risk HPV E7 can cause genomic instability in normal human cells (59). HPV16 E7 induces G1/S and mitotic cell cycle checkpoint defects and uncouples synthesis of centrosomes from the cell division cycle (60), leading to the formation of abnormal multipolar mitoses, chromosome mis-segregation, and aneuploidy (61). Moreover, there is an increased incidence of double-strand DNA breaks and anaphase bridges, suggesting that in addition to numerical abnormalities, high-risk E7 proteins induce structural chromosome aberrations (62). Similar to SV40 large T antigen and the 55-kD protein encoded by adenovirus E1B, the high-risk HPVs E6 proteins target p53 (49). The interaction of E6 with p53 is not direct but is mediated by the cellular protein E6AP (E6-associated protein) (63) discussed further below. E6 has p53-independent functions and several additional cellular targets have been identified. Two activities that are of particular importance to E6's oncogenic activity are 1) the binding to cellular PDZ domain containing proteins through its C-terminal PDZ domain binding motif (64,65) and, 2) the ability to activate telomerase in keratinocytes (66). Nonetheless, a major function of E6 is to counter the pro-apoptotic activity of E7 and the inhibition of E6 activities mediated through E6AP provides a potential therapeutic target for HPV-positive cancers and HPV-associated premalignant lesions. Inhibition of either E6 or E6AP in HPV-positive cells leads to p53 stabilization and apoptosis (67,68), thus the targeting of this pathway has been validated. E6AP-mediated proteolysis of p53 does not occur in HPV-negative cells (68), indicating the specificity of potential therapies to HPV-positive cells.

Despite the discovery of E6AP and the pathway by which E6 hijacks E6AP to target the ubiquitin mediate proteolysis of p53 (69,70), surprisingly little is known about how this pathway may be regulated by the cell. To that end, we are conducting a whole genome siRNA screen to identify cellular genes that, when depleted from the cell, stabilize p53 in HPV-positive cancer cells. This has revealed several dozen potential hits identifying such genes. The screen also has been modified to allow us to identify small molecules that stabilize p53 in HPV-positive cancers. Such molecules could serve as tool compounds to further study the pathway or potentially even as lead compounds for drug development.

CONCLUSIONS

The identification of an infectious etiology for specific cancers provides the opportunities to prevent the cancers by preventing or controlling the infections. Depending on the infectious agent, this recognition might involve public health measures or changes in cultural practices. It could also involve the development of vaccines to prevent the initial infections, as has been achieved now for HBV and the high-risk HPVs. It could also involve the treatment of the infections with specific therapeutics, or the development of novel therapies for those agents for which there are not yet specific or effective drugs. Furthermore, some infectious agents, such as the HPVs that encode genes whose continued expression is required to maintain the cancer, provide potential therapeutic targets for treating the cancers or the cancer precursor lesions.

DISCUSSION

Billings, Baton Rouge: Since teenagers are going to continue to have risky behavior — and some of us have had our own children vaccinated for hepatitis — would it be appropriate for teenagers in these days and times, both male and female, to be vaccinated for HPV?

Howley, Boston: Absolutely. In fact, the Merck vaccine is approved for males based on its effectiveness for condyloma acuminata. I think the data linking oropharyngeal cancers to HPVs, and the epidemic that we are seeing with oropharyngeal carcinoma really, is sort of a mandate that both young boys and girls should be vaccinated.

Harrison, Boston: I wonder not about HPV but HTLV1. Is it known what explains the geographic distribution of prevalence of that virus?

Howley, Boston: I am not aware of what is responsible for the geographic prevalence. But the geographic prevalence of the cancers that is associated with it correlates very well with the presence of the virus.

Rosenblatt, Whitehouse Station: With the kind of increasing promise of some of these immune oncology approaches, do you see that they would work — potentially in combination with your approach — on E6 and E7 and other therapeutic approaches for human papillomavirus?

Howley, Boston: I would say all viral cancers, especially those that are expressing a protein that may be one of the drivers for the cancers, are terrific candidates for immunotherapy for specifically going after the expression of that viral protein. The concern with the papillomavirus is that there is evidence in literature that particularly E7, and perhaps also E6, have mechanisms for down regulating expression of histocompatibility genes and having effects both on innate and adaptive immunity. In the course of malignant progression, there may be epigenetic changes affecting expression of host cell immune genes. One can ask, “Why isn't the immune system seeing E6 and E7 in the cells already since they are foreign proteins?”

Jameson, Philadelphia: I was curious with the efforts to sequence cancer genomes, if there have been any surprises that have come out that reflect on the role of the tumor viruses?

Howley, Boston: The approach of deep sequencing was the one that effectively Yuan Chang and Patrick Moore used to identify the Merkel Cell polyomavirus in Merkel cell cancers. Instead of sequencing the DNA, they sequenced the cDNAs or the RNAs that are expressed within those cells. I know there are a number of laboratories that are looking at all of the DNA sequence data in the various databases to try to identify other infectious agents. Although it may not yet be published, I understand that there may be several cases of HPVs found in some bladder cancers. If one takes the information in all of the various cancer genome and expression databases and queries it against known DNAs and known viruses, one may very well discover some new cancer associations with infectious agents. The important thing will be to pay particular attention to those DNA sequences that don't match known genomes and look for new infectious agents. These are the types of strategies and techniques that people are beginning to look at.

Bodenheimer, New York: In hepatitis C, and to some extent hepatitis B, we clinically link cirrhosis with or the amount of fibrosis to the risk for cancer, as commented in your talk, a concern about chronic inflammation. But I wondered if there are specifically measures of intracellular matrix changes, such as fibrosis or other changes, that are linked rather than inflammation?

Howley, Boston: I am aware of studies that are looking at fibrosis and cirrhosis as perhaps a required initially for this progression. For instance, Ray Chung at MGH in Boston has been doing transcriptome analysis of lesions that are HCV-positive and HBV-positive. Hopefully studies like his will provide further insights.

High, Philadelphia: I think I have heard that HPV-associated oropharyngeal carcinomas have a better prognosis than non-associated. Do you provide any mechanistic insight about that?

Howley, Boston: They do have a better prognosis. The non-HPV types are seen in older individuals. They are usually seen in smokers and drinkers, whereas the HPV-positive ones are often in younger individuals and there is not necessarily an association. If one compares genetically the transcriptional profiles of the HPV-positive and HPV-negative cancers, they are very different cancers. I do not know whether or not we know yet why the HPV-positive cancers respond better to therapies.

Somersille (Sibley Spouse), Mountain View: In your talk you noted that it takes a long time for progression to cancer. Currently the guidelines are that we are supposed to be doing a pap smear every 4 to 6 months, which is kind of overkill. Are you or a basic scientist involved in the guidelines from the American College of Ob/Gyn?

Howley, Boston: Certainly I am not involved in that. There may be some basic scientists involved in providing some advice for these guidelines however.

Somersille (Sibley Spouse), Mountain View: It would be really helpful to have you involved in those.

Howley, Boston: The women that are at risk for developing cervical cancer are women that are usually over 30 that have a persistent infection that can be by DNA testing. That is what they now are thinking of using DNA testing for, to identify the women who are persistently infected involved in providing some advice for these guidelines however.

Somersille (Sibley Spouse), Mountain View: But the frequency of Pap smears now is every 4 months. With the progression in years, it is resources that I think that could be better put toward what you are doing, like figuring out a cure.

Howley, Boston: Yes, thank you.

Alexander, Atlanta: You mentioned several times about stabilization of p53 and its being rescued from the ubiquitin pathway. What level of p53 would you get with your treatment? Would it be within physiologic ranges? High p53 is usually associated with dysfunction of mitochondria in a pretty significant way. How do you stabilize that to get some semblance of normality?

Howley, Boston: We are not looking for normality. We are looking for the killing of the cells, specifically through apoptosis. I did not show you western blots of the p53 levels, but they are quite high within the cells when they are stabilized by, say Velcade, or knocked down to siRNA depletion of E6AP; certainly high enough to the point that they are transcriptionally active in turning on the proapoptotic genes that lead to cell killing.

Hook, Birmingham: The relationships between viral oncogenesis and these infections are quite interesting. So often they are also impacted by cofactors and co-carcinogens. I was wondering about the mechanism by which that worked, whether it was with the virus, with the host, some combination?

Howley, Boston: I think that the general thought is that the co-carcinogens are playing a major role at the level of the host rather than at the level of the virus. The virus is effectively, at least for HPVs, inactivating these protective tumor suppressor pathways such as p53 and the retinoblastoma protein. Inactivation of these tumor suppressor pathways can lead to genomic instability, which when combined with co-carcinogens can drive maligant progression and cancer.