Biochemical, Virological and Epidemiological Aspects of HPV Genital Infection in Cervical Cancer Progression

Transcription

1 Ph.D in Biochemistry XXIV Cycle (A.Y ) Biochemical, Virological and Epidemiological Aspects of HPV Genital Infection in Cervical Cancer Progression PhD Student Elona Bucaj Guide Professor Prof. Chiara Cini PhD Coordinator Prof. Paolo Sarti December 2011

2 To my father Dr. Thanas Bucaj for all he has been in my life...

3 PhD in Biochemistry XXIV st Cycle (A.Y ) Biochemical, Virological and Epidemiological Aspects of HPV Genital Infection in Cervical Cancer Progression PhD Student : Elona Bucaj Guide Professor: Prof. Chiara Cini PhD Coordinator: Prof. Paolo Sarti December

4 1. Introduction page Cervical Cancer Definition page Epidemiology page Etiologyand risk factors page Human Papilloma Virus page HPV and cervical cancer page Oral contraceptives page Smoking page The human immunodeficiency virus (HIV) page Sexual behavior page Parity page Cervical cancer screening and prevention page Project I Genital Human Papillomavirus infection and genotype prevalence among Albanian women: Across-sectional study 2.1 Introduction page Aim of the work page Materials and Methods page Patients page Virology page Viral typing page Results page 30 2

5 2.5 Discussion page Project II Oxidative stress contributes to HPV-driven viral carcinogenesis: Redox proteomics analysis of HPV-16 dysplastic and neoplastic tissues 3.1 Introduction page General Introduction page Oxidative Stress: ROS sources and page 42 Antioxidant systems Overview on Proteomics and page 50 Redox Proteomics methods 3.2 Aim of the work page Material and Methods page Patient enrolment page Viral analysis page HPV typing page Viral load and physical status determination page Protein extraction page Western Blot page Proteomics page Protein oxidation measurement page Redox Proteomics page Image Analysis page Trypsin digestion and protein identification by MS page 63 3

6 Protein identification by LC-MS/MS page Statistical analysis page Results page Patients page Viral assay page Viral load page Viral genome physical status page Expression levels of stress response proteins page Protein oxidation page Identification of carbonylated proteins page Discussion page Viral sample characterization page Stress markers page Redox proteomics page Conclusions page Project III The effect of HPV16 E5 variants in transfected keratinocytes: A proteomics and redoxproteomics analysis 4.1 Introduction page HPV16 E5 protein functions page Signal trafficking pathways page Immune response page 94 4

7 4.2.3 Cell proliferation page Tumour growth page Response to OS page Aim of the work page Materials and Methods page Clinical samples page HPV type characterization page HPV 16 E5 variants analysis page Variants cloning page DNA sequencing page Ligation page Lentiviral vector packaging page Cell culture page Keratinocytes infection page Results and discussion page Patients page Prevalence of HPV16 E5 variants page Cloning of HPV16E5 variants in plico plasmid page NHEK cell line infection page Work in progress page References page 110 5

8 INTRODUCTION 1.1 Cervical cancer definition Cervical cancer (CC) is malignant neoplasm of the cervix uteri. Most cervical cancers (80%) are squamous cell carcinomas, arising from the squamous epithelial cells of the cervical transformation zone (i.e.: the squamous columnar junction: the junction line between the vaginal multilayered squamous epithelium and the simple, columnar epithelium of the endocervical canal) fig.1a. A less common neoplasia arising in glandular epithelial cells is referred to as cervical adenocarcinoma. Very rarely, cancer can arise from other types of cells in the cervix. Figure 1: Colposcopically image of cervix uteri (fig. A) and cervical intraepithelial neoplasia grades (CIN) (fig. B). Schiffman M et al. JNCI J Natl Cancer Inst 2011;103: Epidemiology CC is the third most common cancer in women, and the seventh overall. With an estimated 529,000 new cases in 2008 and deaths per 6

9 year it continues to constitute a major public health concern, ranking it as the fourth most common cause of cancer mortality in women worldwide. Unlike other neoplastic diseases CC affects women between years old, at the top of their personal potentials, resulting in severe burden for individuals and societies [1]. Wide variations in incidence exist between low and high CC burden countries, whose rates range from <3 to >50 per respectively. More than 85% of the global burden occurs in developing countries, where it accounts for 13% of all female cancers. High-risk prevalence regions include Eastern and Western Africa, with a cumulative risk (0 74) of 3.8%, Southern Africa (2.9%), South-Central Asia (2.6%), Middle Africa and South America (2.5%). Risks are lowest in Western Asia, Northern America and Australia/New Zealand (0.5%) [2]. These contrasts are believed to reflect both differences in exposure to risk factors and protection offered by screening. 1.3 Etiology and risk factors Human papilloma virus (HPV) infection is the major etiologic agent of cervical carcinoma. Other co-factors like: smoking, oral contraceptives, high parity, sexual habits, diethylstilbestrol (DES), human immunodeficiency (HIV), are thought to play an important role in cervical cancer development Human Papilloma Virus Human Papillomaviruses (HPVs) belong to the Papillomaviridae family. Their implication in cervical cancer pathogenesis was originally proposed in mid-1970, upon aneddoctical observations. HPVs and their role 7

10 in squamous cervical cancer (SCC) is now soundly proved and universally accepted [3]. HPVs infect squamous epithelia (or cells with the potential for squamous maturation) inducing proliferative lesions, of which the humble skin wart is a typical example. These viruses are highly species-specific and exclusively tissue tropic, undergoing a complete infectious cycle only in fully differentiating squamous epithelium. More than 100 HPV types are described based on the L1 ORF region sequence [4]. Viruses mainly infecting the anogenital mucosal epithelia, i.e. the HPV types included into the alpha-papillomavirus genera, may be distinguished according to their oncogenic potentials as low-risk (LR) and high risk (HR) types. Low-risk types, like HPV 6, 11 and other types member of alpha-papillomavirus genera, are predominately associated with benign anogenital warts or condyloma. Fifteen mucosal HR-HPVs (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73 82) are reported to be associated with cervical intraepithelial neoplasia (CIN) and cervical cancer. HPV 16 is the commonest HPV type detected in squamous cervical cancer (SCC) (55%), followed by HPV 18 (12.8%), and together these two types cause around 70% or more of SCC irrespective of geographical locale. The next most frequently detected HPVs worldwide are HPV 33, 31, 45, which contribution does not exceed 4% each [5,6]. HPVs are small viruses composed by a non-enveloped capsid and a circular double-stranded DNA genome with sizes close to 8 kb. Their genome is composed of 3 domains: a noncoding upstream regulatory region (URR) of approximately 1 kb; an early coding region (E genes), and a late coding region (L genes) (fig.2) [7]. 8

11 Figure 2: HPV Genome The URR is an 850 bp region, located between the end of the open reading frame (ORF) of L1 and the start of the E6 ORF. The main P97 early promoter together with the origin of replication (ori) makes up the 3 portion of the URR and regulates the expression of multicistronic transcripts. The central segment of the URR contains an epithelial cell-specific enhancer providing tissue specificity and modulates viral gene expression through the binding of different transcription factors. 9

12 The early proteins (E1, E2, E4, E5, E6 and E7) are functional proteins expressed early in the viral life cycle in lower epithelial layers. They are necessary for the transcription and replication of the HPV DNA and are involved in proliferation and survival of the host cell. Throughout the viral life cycle two viral proteins, the DNA helicase E1 and the DNA binding protein E2, are required for viral genome replication [8]. E2 is a DNA binding protein that interacts with target sequences located around the viral origin and recruits the E1 helicase, which forms a dihexameric complex that via the interaction with host proteins take part into the viral genome replication. In addition to the above process the E2 protein has been demonstrated to interact with or to modulate the activity of various viral proteins, including L2, E6, E7, and with a number of cellular proteins involved in transcriptional regulation or genome tethering, such as the TATA box binding protein (TBP), Sp1, p53 and Brd4 [9]. The E4 protein is expressed as a fusion product of the co-linear protein with five amino acids from N-terminus of E1. It is expressed in the late stages of the HPV infection and thought to be involved in the alteration of cytoskeleton network resulting in the assembly and release of the infected virions. Another gene of HR-HPV genome, whose encoded protein possessing proliferation stimulating activity and thought to be a potential oncogene, is E5 gene. HPV16 E5 protein contributes to cellular transformation by increasing the mitogenic stimulus from growth factor receptors (EGF) to the nucleus [10]. In addition the E5 protein has an antiapoptotic activity through the stimulation of the ubiquitin proteasome-mediated degradation of Bax protein, involving different pathway like: COX-2, PGE2 and PKA [11]. The oncogenic activity of the HR- HPV types is mainly mediated by the joint action of two viral oncoproteins, E6 and E7. These oncoproteins can 10

13 induce cellular immortalization and transformation interacting with cellular proteins that are involved in regulating cell cycle and apoptosis. Major activities of the E6 oncoprotein include proteasome-mediated degradation of the p53 tumor suppressor and of pro-apoptotic proteins like BAK, BAX, resulting in resistance to apoptosis and in increased chromosomal instability. E7 interacts with a number of cellular proteins, targeting mainly the prb proteins family for proteasome mediated degradation by blocking the activity of these tumor supressors. When E7 binds to prb proteins, it relieves their repression of E2F (a nuclear transcription factor), resulting in the constitutive activation of E2F-responsive genes. This action of E7 causes the cell to re-enter the S phase where cellular replication factors that are necessary of viral replication are activated [3]. The late genes L1 and L2 encode for the structural proteins of the capsid, respectively the major and minor capsid protein. Their expression occurs late in infection and is restricted to the superficial zone of the epithelium for the assembly of the virus. These two structural proteins are not expressed in precancerous and malignant cells, but are very important for vaccine development [3, 12] HPV and cervical cancer HPV is a sexually transmitted infection and very common in young sexually active people. The virus life cycle requires the availability of epidermal or mucosal epithelial cells that are still able to proliferate, i.e. the basal layer cells. In these cells, however viral replication is highly restricted. Despite this restriction, the viral genes are expressed at very low copy number. This result in enhancing the proliferation of the infected cell and its 11

14 expansion in a proliferating clone. At some time after infection, there is around of viral DNA replication, which appears to be independent of the cell cycle and which amplifies the viral copy number per cell. The infected cell is thought then to leave and enter the proliferating compartment of the epithelium, where the late viral genes expression is initiated, the circular viral genome is replicated, and structural proteins L1 and L2 are formed. In the upper layers of the epidermis or mucosa, complete viral particles are assembled and released. This infectious cycle takes about 2 3 weeks in vivo, the time taken for the basal keratinocyte to move up the epithelium and differentiate (fig.3) [13]. During these infectious cycles the viral genome is intact and maintained as plasmid or an episomal, where viral gene expression is minimal, in particular, the expression of the oncogenes E6 and E7, which is under very tight control of the E2 ORF. Figure 3: Infection cycle of HPV. Margaret Stanley. Gynecologic Oncology 117 (2010) 12

15 Sometimes, however the infection tends to persist much longer and the episomal viral DNA frequently becomes integrated into the host cell genome. This step is considered important in the progression from a simple viral infection to carcinogenesis. Indeed, in those cases in which the circular DNA molecule is opened within E2 ORF, the structural integrity of the E2 ORF is disrupted and so its protein function. This allows the E6 and E7 oncogenes to escape the E2 suppressing control and to be expressed unrestrictedly [3, 14]. Their marked transforming activity is related tothe degradation and the loss of functions of two tumor suppressor genes - p53 (regulating the G1 cell cycle checkpoint preventing entry into S phase of cells containing DNA damage, and allow DNA repair), Retinoblastoma protein (prb) (which formsa complex with E2F family of transcription factors that prevent cell proliferation). P53 is a DNA binding protein that is expressed in response to DNA damaging agents and unscheduled induction of DNA replication resulting in cell cycle arrest and apoptosis. Since HPV depends on cellular DNA synthesis machinery for replication of its own genome, the proper expression of p53 represents a major impairment for viral replication. In order to suppress the p53 fuction the E6 protein of HR-HPV types first binds to a cellular ubiquitin ligase the E6-AP and then binds the p53, recruits the ubiquitination complex of enzymes, and directs it to proteosome degradation. The degradation of p53 results in bypassing the normal growth arrest signal at G 1 /S and G 2 /M check point leading to the acquisition of numerous genetic alterations that contribute to malignant transformation. E6 has many other activities among which it also mediates the degradation of another pro-apoptotic protein, Bak, a member of the Bcl-2 family thus re-enforcing its powerful antiapoptotic activity [13]. The transforming role of high risk E7 protein is associated mainly with its role in 13

16 controlling the members of retinoblastoma (Rb) tumor suppressor family. In normal cells, Rb is hypophosphorylated in early G 1 and becomes increasingly phosphorylated towards S phase. In its hypophosphorylated form, Rb binds E2F transcription factors and actively represses transcription from promoters containing E2F sites, which include a large number of genes required for DNA synthesis, such as DNA polymerase alpha and thymidine kinase. Binding Rb in a hypophosphorylated state, E7 prevents it from binding to E2F and induce ubiquitin-mediated degradation of Rb appears to be essential to efficiently overcome cell cycle arrest. Thereby E7 promotes cell cycle progression in differentiated epithelial cells, allowing for productive replication of HPV genome. E7 from HR-HPV types have demonstrated its ability to abrogate the inhibitory activities of the cyclin-dependent kinase inhibitors (CKIs) such as p21 and p27. It has been shown that inactivation of both the Rb protein and p21 by E7 is necessary to prevent cell cycle arrest [15]. The binding of p21 and p27 by E7 interferes with the ability of p53 to induce G1 growth arrest following DNA damage. E7 proteins of HR-HPVs are found to induce cell proliferation by activation of cyclin E kinases and controlling gene regulation through interaction with AP-1 family of transcription. The binding of E7 to c-jun points to further interactions with transcription factors (fig.4) [16]. All together the above interactions of E6 and E7 lead to an abnormal cell cycle control and ultimately contribute to the accumulation of genetic changes required for initiation of malignant transformation [17-19]. 14

17 Figure 4: Mechanisms of action of HR-HPV E6/E7 oncogenes.di Domenico F. et al. Biochim Biophys Acta Oct 12. [Epub ahead of print] Most HPV infections are subclinical, transient infections with the presence of viral DNA and the absence of colposcopically detectable lesions or with minimal lesions like low squamous intraepithelial lesions (LSIL) in cytology screening, and resolve in 90% of cases by the cell-mediated immunity in 6-12 month after appearance. In a small percentages (probably 10 15%) of cases the HPV infection persistent, with a constant production of infectious virus, due to an unsuccessful cell-mediated immune response. This group of women, persistently infected with a HR-HPV, are at risk for the 15

18 development of high-grade intraepithelial lesions (HSIL) and carcinoma in situ (fig.5) [20]. Figure5: Natural history of genital HPV infection.margaret Stanley. Gynecol.Onc. 117 (2010) S5 S10 The progression of HPV infection to LSIL, HSIL and invasive cancer is determined by failure of immune control mechanisms including: intracellular controls by cyclin-dependent kinase inhibitors, paracrine signaling cascade (macrophages, cytokines like TNF-α and interferon β) and decreasing immunological control (humoral and cellular immune response) [21]. Histological feature of the epithelium determines the presence of cervical intraepithelium neoplasia (CIN) lesions. The degree of CIN is determined by the portion of thickness of the epithelium showing cellular atypia, such as loss of cytoplasmic maturation, nuclear pleomorphism, 16

19 nuclear hyperchromasia, disordered growth (characteristic and atypical mitotic figures) [22]. The lesions where cellular dysplasia is confined to the lower third of the epithelium thickness are regarded as CIN I, those confined to the lower two third of the epithelium as CIN II. In CIN III lesion cellular alterations extend over more than two third of epithelium thickness and in carcinoma in situ (CIS) indicates full-thickness lesions (fig.6)[23]. The acquisition of this genomic chaos is considered the crucial stage in progression from low-grade CIN I to high-grade CIN II/III. Genetically unstable lesions are capable of progression due to the increased probability of acquiring further mutations of host genes that could lead to invasive cervical cancer (ICC) [24]. Figure 6: Histological classification of cervical intraepithelial neoplasia (CIN). The last panel shows a nest of cells growing beyond the basal membrane layer (mico invasion), an event marking the onset of invasive carcinoma. Despite the outlined ancogenic potential of viral infection, it is well recognized that most HPV infected women do not develop CC suggesting that HPV infection alone is not sufficient for neoplastic transformation, and that host and environmental factors, causing additional molecular genetic 17

20 alterations or aberrant expression of cell proliferation pathways, are required to activate malignant conversion [25] Oral contraceptives The first reports on possibly increased risk of cervical cancer (CC) among oral contraceptive (OC) users were followed by a large number of epidemiological studies reporting contradictory results. Many reports failed to establish an increased risk for CC associated with OC use, while others have reported OC use as a positive risk factor. The data have been repeatedly reviewed by the International Agency for Research on Cancer (IARC) experts, and evidences indicate that longterm use of OC is a co-factor that increases the risk of CC in women who are positive for HPV DNA [26]. The relative risks for cervical cancer associated with duration of OC use of approximately less than 5 years, 5 9 years, and 10 years or longer, respectively, were: 1.1 (95% CI ), 1.6 ( ), and 2.2 ( ). Studies presented data on the risk of cervical cancer according to time since ceasing use of OC, and the relative risk declines with increasing time since last use and after 10 years or more of cessation the risk returned to that of never users [27,28]. The use of OC is not an independent risk factor for cervical carcinogenesis. The association of OC use and CC is far more complex, because other factors like HPV infection, sexual behavior (irrespective of OC use) are the risk factors predisposing women to high risk HPV, development of high-grade CIN (HSIL), and also influence on the outcome of their cervical disease/high risk HPV infection [29]. It is thought that OC might affect the development of HPV infection. A study reported that the use of OCs for >6 years was found to be associated with an increased risk of infection with any HPV compared to never users, 18

21 independently of sexual behavior and cervical abnormalities. These data suggest that exogenous hormonal exposure in the form of OC use could influence both viral and host factors associated with the outcome of HPV infection and cervical cancer [30]. OC exposure was shown to affect cervical cells directly, increasing cell proliferation, stimulating transcription E6/E7 gene of HPVs and modulating the host immune response to the HPV 16 virus-like particle through upregulation of anti-inflammatory (IL10,6) and regulatory immune markers (INFγ, TGFβ). The possibility also exists that reproductive hormones may promote the integration of HPV DNA into the host genome, which would be consistent with the observed increased risk for cervical cancer after prolonged use of OC [31, 32] Smoking The International Agency for Research on Cancer (IARC) has classified tobacco smoking as a cause of cervical cancer. Data from 23 epidemiological studies provide evidence that current smokers, compared to never smokers, are at a significantly increased risk of developing squamous cell cervical carcinoma and that this risk increases with the number of cigarettes smoked per day and with decreasing age at starting smoking. Surprisingly, smoking was not found to increase the risk of adenocarcinoma of the cervix [33]. In another study the relationship between tobacco smoking and natural history of low-grade cervical abnormalities was examined, showing a strong dose response relationship between smoking and increased risk of persisting LSIL and HPV infections. Results suggest that tobacco smoking can contribute to cervical carcinogenesis by interfering with regression of 19

22 HPV-induced lesions and through the suppression of T-cell responses against HPV infections [34]. Accordingly smoking habit has a strong influence on local inflammation and immune response, as indicated by the finding that among HPV 16 infected women, smokers have a delayed antibody response when compared to never smokers. These results give rise to new insights about potential mechanisms by which tobacco smoke could increase the viral infection persistence and tumor onset [35]. Detection of cigarette smoke carcinogen-specific products, such as N- nitrosamines and nicotine, in the cervical mucus suggests that these mutagens can induce gene mutations and chromosome aberrations that become the hallmarks in the subsequent cancer cells. However, it has been known that carcinogenic effects of tobacco smoke should be modulated by interindividual differences in activity and efficiency of metabolic and detoxification pathways. In particular, polymorphisms in xenobiotic metabolizing gene enzymes, like Glutathione-S-transferases (GSTM1, GSTT1, GSTP1) involved in activation and biotransformation of environmental carcinogens, have shown to increase the risk for cervical cancer [36]. However, the role of tobacco smoking in multistage carcinogenesis is not fully understood because of a paucity of prospective data The human immunodeficiency virus (HIV) Women infected with the human immunodeficiency virus (HIV) are at increased risk of HPV-related cervical disease, cervical dysplasia (CIN) and invasive carcinoma of the uterine cervix. They present with invasive cervical cancer 10 to 15 years earlier than HIV-negative women, manifest 20

23 with more advanced disease, and have a poorer prognosis compared with HIV-negative counterparts [37]. Several studies exploring the interaction between HIV and HPV have demonstrated that compared to their HIV-uninfected counterparts, HIVinfected women have increased overall HPV prevalence and increased rate of multiple infections [38]. HIV infection has also shown to alter the natural history of HPV related carcinogenesis. This may be caused by HIV attenuation of the systemic immune response or local cellular immunity in the cervix, modified cytokine expression altering HPV regulation, up-regulation of HPV expression by HIV utilizing the Tat gene that leads to increased expression of HPV E1 and L1 genes. The impaired ability to clear HPV infection causes viral persistence, allowing the progression of infection through CIN increasingly severe CIN up to invasive cancer [39] Sexual behavior Sexual behaviour is considered another risk factor in cervical cancer development. Early age at first sexual intercourse (AFSI) or shorter intervals between menarche and sexual debut, have been postulated as risk factors for HPV infection acquisition. However, AFSI does not seem to be independently associated with HPV positivity, but is more likely a predictor of the number of lifetime sexual partners. The earlier a woman starts, the more sexual partners she is likely to have and the higher chance of getting an HPV infection which is a major causative agent of cervical carcinogenesis [40]. 21

24 1.3.6 Parity A recent pooled analysis of 25 epidemiological studies that analyzed the association of reproductive factors and cervical carcinoma, showed that women with more than sevenfull-term pregnancies (FTP) were at higher risk of developing cervical carcinoma than those who had one or two FTPs. Early age at first FTP (17 years versus 25 years) was also associated with an increased risk of both CIN3/carcinoma in situ and invasive carcinoma. However, no relationship was found between HPV positivity and number of FTP or early age at first FTP among controls, emphasizing the potential role of these factors in the transition of HPV infection to neoplastic lesions [41]. 1.4 Cervical cancer screening and prevention Cytology-based population screening programs have resulted in a dramatic decline in CC incidence [42]. Cytological testing, synonymous of Papanicolaou test (Pap test), involves collection of exfoliated cells from the cervix and microscopic examination of these cells after staining. The test is based on the knowledge that tumour cells lose the cohesive properties of normal cells and are therefore readily dislodged when the ecto- and endo-cervix are scraped and brushed, thus allowing the detection of the earliest and smallest neoplastic lesion by cytological examination before it is visible to the naked eye [43]. The most widely used system to report cytological evaluation is Bethesda system. It uses a two-tier grading system for reporting neoplastic changes in squamous cells: HSIL (severe dyskaryosis) and LSIL (mild dyskariosis). 22

25 Although the introduction of liquid-based cytology in mid-1990 has improved the performance of the cervical conventional cytology test, Papanicolaou (Pap) test sensitivity remains low despite its high specificity [44]. In several studies, the Pap smear has been shown to have a high falsenegative rate, and the sensitivity of cytology to detect high-grade cervical lesions is on average 53%, hencing several pre-cancerous cervical lesions not be identified nor treated on time. Another drawback of cervical cytology is its high false-positive rate, which is primarily restricted to the diagnosis of minor abnormalities, and may be attributed to the difficulty in distinguishing between inflammatory reactive atypia and LSIL [45]. As high-risk HPV is a necessary cause for cervical cancer, HPV testing could be safely used in primary cervical screening. It results to be more sensitive than cytology to detect pre-cancer and grants a high long-term protective effect of developing cervical lesions to women testing negative [46]. The hybridization method (Hybrid Capture HC2) approved by FDA (Food and Drug Administration) in combination with cytology (co-testing) are used as primary screening in the USA [47]. Preliminary studies have suggested that it may have superior sensitivity compared with cervical cytology for the detection of HSIL or cancer [48]. Detection of high-risk HPV DNA is considered to be potentially useful in four clinical applications: (1) as a triage test to select which women who have low-grade cytological abnormalities in routine screening require immediate referral for colposcopy rather than cytological surveillance; (2) follow up of women with abnormal screening results who are negative at colposcopy and biopsy; (3) follow up for women treated for high-grade CIN with local ablative or excisional treatment; (4) as a primary screening test, 23

26 either alone or in combination with cervical cytology to detect cervical cancer precursors. Because the vast majority of HPV infections represent acute HPV infections that are destined to clear without causing cancer, HPV testing has poor specificity and positive predictive value for cervical cancer screening [49]. The combination of HPV DNA testing with cervical cytology and other biomarkers may represent the potential to further improve the detection of pre-invasive disease and thereby reduce the incidence of high-grade lesions compared with cytology testing alone. Although screening and management guidelines may evolve with our improved understanding of the biology and immunology of HPV, current recommendations and methodologies for testing are not impacted by the more recent advent of the HPV vaccine. At present, there are two HPV vaccines commercially available, Cervarix (the bivalent vaccine directed against HPV types 16 and 18 and Gardasil (quadrivalent HPV vaccine directed against four HPV types: HPV 6, 11, 16 and 18). Both are sub-unit protein vaccines consisting of the L1 coat protein assembled into macromolecular structures virus like particles that mimic the wild type virus capsid morphologically and antigenically, although contain no DNA and are therefore not infectious [50,51]. These vaccines (recommended for females in the 9-26 age range) have demonstrated high efficacy in the prevention of infection by the named HPV types, related dysplastic and neoplastic disease. However, they have no therapeutic effect on existing infections or cervical lesions. One of the most important clinical benefits of HPV vaccination will be the reduction of cervical cancer, but it will take decades before the impact of vaccine on invasive cervical cancer is observed 24

27 [52]. Therefore, the use of HPV vaccines does not alter current cervical cancer screening and management guidelines. 25

28 PROJECT I Genital human papillomavirus infection andgenotype prevalence among Albanian women: across-sectional study 2.1 Introduction Genital infection with human papillomavirus (HPV) is the most common sexually transmitted disease and represents a major concern for teenagers and young adults. The genital HPVs, presently included in the genus Alpha-Papillomavirus, family Papillomaviridae [4], can be classified either as HR-HPVor as LR-HPV based on their association with the onset of cervical dysplasia and cervical cancer [53,54]. HR-HPVs are represented in human cancers with considerably different prevalences from one type to another. Namely type 16; 18; 45; 31; 33; and 58, account for over 90% of cases whereas the remaining 10% of cases are caused by the many, so called rare types [55,56]. Consequently, screening campaigns are under way, diagnostic tools have been devised and vaccine formulations are being proposed to prevent cervical cancer. However, the world HPV circulation is not uniformly represented by available data. Some continents, namely Africa and Asia, are comparatively under-represented while North America, Europe, and Australia are conversely over-represented. Within the continents large areas are poorly represented or not represented at all, while the vast majority 26

29 of data are derived from people living in a few major cities within each country and thus, conceivably refer to specific populations and social groups. Nonetheless, there is significant evidence of HR-HPV interregional variation of HPV prevalences. Within this context the Mediterranean area represents an interesting environment. This is a highly specific region, with a unique climate, crowded with different people with marked different social, economic, and cultural conditions and sharing increasingly intense migratory exchanges. Thus, conditions promoting and conditions hampering sexually transmitted diseases are simultaneously present and may affect unpredictably HPV circulation. In this project the pattern of genital HPVs circulating in adult females living in Albania is described. In this small country, due to the deep political transformation of the 1990s and to the intense migratory flux toward Western Europe [57], a severe increase of cervical carcinoma has been observed [58]. In addition data on the circulation of HPVs and genotype prevalence are not available. 2.2 Aim of the work This is an epidemiologic study done to obtain information about the local HPV features, its distribution in the population and to characterize the most frequent types. These are the only virological data in concern to this country and are relevant for the implementation of screening and prevention campaigns at a regional level, as well as for a more detailed understanding of HPV infection in Southern Europe and the Mediterranean area. 27

30 2.3 Materials and methods Patients Patients included in this study were selected from women attending the Gynaecological Out-Patients Clinic of the Oncology Department at the Nene Tereza University Hospital Centre of Tirana and from women attending the Family Practice at the Gynaecological Clinic of Durres. In order to be eligible, women had to be 18 years of age or older, not pregnant, not to have any history of neoplastic disease, and to have presented solely for routine gynecological check. The study design and enrolment criteria had been approved previously by the Nene Tereza University Hospital Centre s local Ethical Committee. During the period, January 2004 to September 2010,1367 women were selected among those providing a written informed consent and matching the above inclusion criteria. Enrolled patients were given a standardized questionnaire to obtain data on epidemiological and personal history. This included details about their sexual behavior, a history of sexually transmitted diseases, tobacco smoking, reproductive history, contraceptive methods used, personal hygiene, and details about their compliance with programs for cytological screening of cervical lesions. At entry, enrolled patients underwent a full gynecological examination, including cytology and colposcopy. Histology was carried out on a selection of patients based on gynecological indication according to standard criteria Virology Viral detection and typing were performed on affinity purified total DNA (QIAamp DNA Mini Kit, QIAGEN GmbH, Hilden, Germany) obtained 28

31 from the residual cytological material. Viral detection was undertaken by the combined use of two different PCR methods. Samples were first assayed using the MY09/MY11 primer pair as described [59,60]. These highly degenerated primers have been used widely in epidemiological and clinical studies. They generate an approximately 450 bp long amplicon appropriate for precise and sound type identification by downstream direct sequence. As a counterpart this fairly long amplified region makes them rather prone to fail amplification once suboptimal quality DNA is used as the starting material. Considering that parts of samples were collected in rural facilities and that during the transport might have suffered because of cold chain interruption, MY09/MY11 negative samples were retested by the GP5+/GP6+ primers [61]. This primer set amplifies a roughly 150 bp long fragment and is therefore adequate for analysis of samples with partially degraded DNA. Both reactions were performed in a 50 µl final volume containing 500 nm of each primer, 200 µm of each dntp (Roche Diagnostics S.p.A Roche Applied Science, Monza, Italy), 1 unit of thermostable Platinum Taq Polymerase and 1x reaction buffer (both from Invitrogen Life Technologies, s.r.l.-san Giuliano Milanese, Italy). MgCl 2 concentration was 1.5 mm for the MY09/MY11 mixture and 3.5 mm for the GP5+/GP6+ one. Amplification reactions were performed in an IQ4 Thermal Cycler (Bio-Rad Laboratories s.r.l., Segrate, Italy). Amplification protocol consisted of an initial 150 sec at 95 C for DNA denaturation/polymerase activation step followed by 35 cycles of annealing for 30 sec at 55 C (for 60 sec at 42 C in the case of GP5+/GP6+ primed amplification), extension for 45 sec at 72 C and denaturation for 30 sec at 95 C, plus a final cycle with a 10-min 72 C extension step. Results were evaluated by visual inspection of ethidium bromide stained 2.5% agarose gel under UV-B trans-illumination. 29

32 Adequacy of starting materials to PCR evaluation was assessed by amplification of a 268-bp fragment of the β-globin gene as described [62]. Positive and negative controls consisted in purified DNA extracted from Siha and from HaCaT cells, respectively Viral Typing Viral typing was detected by direct sequencing of amplified products. Briefly, amplicons were purified by silica gel affinity chromatography (QIAqick PCR purification kit QIAGEN GmbH), resuspended in a 20µl final volume with 600nM the upstream primer used for the amplicon generation (i.e., either MY11 orgp5+) and transferred to the BIO-FAB Research s.r.l. (Pomezia, Italy) where they were sequenced by the BigDye Terminator 1.1 Cycle Sequencing Kit (Sanger method). Cycle sequencing products have been purified by Sephadex G-50 (GE Healthcare Bio-Sciences MedicalDiagnostics, Milan, Italy) gel filtration. Sequences were analyzed by capillary electrophoresis in 3730 DNA Analyser (Applied Biosystems, Inc., Foster City, CA) and aligned to prototype viral sequences through theblast resource at the NCBI ( 2.4 Results Patients enrolled ranged in age from 19 to 75 years (median=36; mean=38). With regard to education level, 677 patients (49.5%) had an university degree or an equivalent qualification, 526 patients (38.5%) had a high school qualification, 125 (9.2%) had an intermediate school education, and 39 (2.6%) had no qualification at all or did not give any answer. 30

33 The results obtained by the β-globin amplification indicated that PCR quality DNA was obtained from 1315 out of 1367 cytological specimens, while no amplification was consitently observed in 52 samples (data not shown). These 52 samples were thus filed as inadequate for molecular assay and excluded from further analysis. Among the 1315 patients with adequate DNA specimen, 288 (approximately 21.9% of the cohort) were found to be infected with a genital HPV. This figure was higher (23.51%) among women aged 30 or over and lower (15.38%) among those between (Table I). Young women had a relative risk of genital infection of 2.125, with a confidence interval ranging from to Table I shows HPV16, the prototype of the species 9 of Alpha- Papillomaviruses, as the most frequent type accounting for 224 out of 288 typed cases (77.78%). HPV 18, belonging to species 7, and HPV 31, belonging to species 9, were the second and third most common HR HPV types being detected in 8 and 7 cases out of 288 (2.78%, 2.43%) respectively. HPV53, belonging to species 6, was detected in 5 out of 288 typed cases (1.73%). HPV 35 was detected in 4 out of 288 cases. HPV 59, member of species 7, was detected in 3 cases (1.04%), as well as HPV 66 (3 cases, 1.04%), HPV 56 (2 cases, 0.69%) both members of species 6, HPV 67 (1/288), HPV 74 (1/288) and HPV X (sequence analysis not matched with already identified HPV types) (11/288). The two LR types HPV 81 and HPV 84, belonging to species 3, were detected in 9 and 4 cases, respectively, with a prevalence of 3.12% and 1.39%, while the two LR members of species 10, namely HPV 6 and HPV 11, were detected in 3/288 (1.04%) and 2/288 (0.69%) cases respectively. 31

34 In eight cases post-pcr analysis did not confirm the presence of HPV-specific sequences in amplified products. These samples were therefore regarded as inconsistent/inconclusive results. Type-specific prevalences were also calculated for both young and mature women (as defined above) and are shown in Table I. General Young Patient (<30) Mature Patient (>30) population N% N% N% Samples HPV negative HPV positive Inconclusive Genotype HPV HPV HPV HPV HPV HPV HPV HPV HPV HPV HPV HPV HPV HPV HPVX Table I: Prevalence of Genital HPV Infection Among Albanian Women based on patients ages HPV types distributions regarding the patients age are depicted in Figures 7 and 8. 32

35 Figure 7: HPV types distribution regarding the patients age Figure 8: Distribution of all HPV types, except HPV16, regarding patients age 33

36 2.5 Discussion Available epidemiological data about HPV circulation are still unsatisfactory with large areas being under-represented or totally neglected [6,55,56]. Indeed, for a number of countries in the Mediterranean area only limited and preliminary data are available [63-66], for other countries no data are available at all. This study therefore, provides data on the prevalence of genital Alpha-papilloma Viruses among a cohort of 1315 women living in Albania. The crude overall prevalence of HPV infection here reported for the whole sample (21.9%), as well as those found among young and mature patients (15.31% and 23.51%, respectively) is within the expected range for patient [67,68] and close to those reported about other Balkan regions [69,70], although they are rather higher than those reported for unselected women in other Southern European countries [6]. Conversely, the prevalence of specific HPV genotypes appears very different from the one expected on the basis of global viral circulation data [6,55,56]. Indeed, although HPV 16 confirmed to be the most prevalent type, its prevalence of 77.78%, while well in tune with the one reported for Greece in a recent study [71] and similar to the one described in Northern Africa [66,71], is unusually high for unselected patients at global level, as well as in other Southern European countries. The second most prevalent type washpv18, which was detected in eight cases (i.e., a 0.6% prevalence among the general population and 2.79% among cases) followed by HPV31, present in 7 cases with a prevalence of 2.43%. Data are according to what is consistently reported in large studies 34

37 that consider these HPV types as the next most relevant contributors to the viral and cervical cancer burden worldwide [6,55,56,72]. HPV 53 was found in 5 cases with a prevalence of 0.4% among general population. Its finding is also reported in two recent studies about Greece [71,73] where it was found the second most prevalent type. HPV 53 is a member of alpha-papillomavirus species 6, whose oncogenic potential is still a matter of debate. Indeed, it is considered a HR-HPV on the basis of sequence analysis, however because of limited epidemiological data available, it is not possible to assign it any definite oncogenic potential. Accordingly, the definition of probable high-risk type was suggested [67,72]. Interestingly, two other members of species 6, namely the HR HPV 56 and the probable HR HPV 66, were also detected in2 and 3 cases, respectively. Thus, considering the whole data set, members of species 6 were detected with a frequency of 0.8% in the general population and 3.5% among the 288 infected women. HPV 18 and HPV 59 were the only HR types of species 7 (HPV 18; 39; 45; 59; and 70) represented in the population, detected in 3 cases. With respect to LR HPV types, HPV 6, a member of species 10 reported to be frequently present both in low and high-grade lesions as well as in cytologically normal specimens, was found nearly with the same prevalence of the less common HPV 11, detected in three and two cases, with prevalence of 1.04% and 0.69% of cases, respectively. Interestingly, nine cases of HPV 81 and four cases of HPV 84 were diagnosed. These two LR types, both members of species 3, with their global 4.5% prevalence of cases, represent the third most common species among Albanian women, whereas they are generally very uncommon in other regions, apart from South-Eastern 35

38 Asia. In this region, close to the Mediterranean basin in terms of climate, genetics and epidemiological features, HPV 81 was reported to be not unusual, ranking fourth among cytologically normal women [6]. Finally, represented data display a peak of HPV infection in women aged over 30, while no significative differences, in terms of type-specific prevalences, were observed in young and mature sub cohorts in comparison with the whole set of patients, with only minimal changes in the rank of minor types. These data were obtained from patients living in two major cities of Albania and characterized by a higher level of education. This might have partially biased the results, which thus need to be confirmed by larger, more detailed studies. However, this is the first report, based on existing evidence, dealing with HPV type circulation in Albania. These preliminary data show that HPV type circulation is far more complex than assumed commonly. Indeed, general features (i.e., HPV 16 is the most prevalent type almost in every region, followed by HPV18), associated with regional patterns such as the unexpectedly high prevalence of type 53 (reported also for neighboring Greece) and with specific features such as the high frequency of LR type 81 and 84 among Albanian women, underline a number of points: (i) supposedly HR HPV rare types may be rather frequent in small epidemiological pockets and might cause HR type shift in areas where HPV 16/HPV18 based vaccination campaigns had been implemented successfully (ii) vaccination campaigns, formulated on the basis of large scale type prevalences, could be grossly inappropriate leading to a failure of prevention programs or, worse, boost non-vaccine HR HPV types in the population 36

39 (iii) detailed knowledge of the HPV genotype circulating patterns in specific local areas is needed urgently to design adequate strategies for the surveillance of vaccinated as well as non-vaccinated women. 37

40 PROJECT II Oxidative stress contributes to HPV-driven viral carcinogenesis: redox proteomics analysis of HPV 16 dysplastic and neoplastic tissues 3.1 Introduction General Introduction Carcinoma of the cervix is the second cause of cancer mortality in females. Large body of knowledge supports the view that HR-HPV have the ability to transform normal cells initiating the progression toward the malignant phenotype. Among the HR types, HPV16 is the most prevalent type in premalignant and malignant cervical lesions [55,72]. The products of two early genes, E6 and E7, are mainly responsible for the carcinogenic activity of the virus. They exert their functions by binding and then promoting the degradation of two important tumor suppressors proteins, the p53 and the Retinoblastoma protein (prb), respectively. These combined actions deregulate many cellular functions such as the cell cycle control, DNA repair, terminal differentiation, apoptosis etc, providing the conditions for genetic damage accumulation and cancer progression. Persistence of the viral infection is considered a key event in virusinduced carcinogenesis. As shown by its high prevalence in cervical 38

41 carcinoma, HPV16 is likely to persist and to cause progression into cervical intraepithelial neoplasia (CIN) than other HR-HPV types [3,74]. A crucial step in viral oncogenesis is supposed to be the viral integration into the host genome. This event, once occurring with the disruption of the E2 gene, results in two fundamental events for carcinogenesis: first, the viral E6 and E7 oncogenes are permanently inserted within the host genome thus preventing any eventual healing, second the E2 disruption abrogate its functions and namely its negative regulatory control on E6 and E7 which are highly and continuously expressed. Experimental and epidemiological evidences however indicate that only a minor part of the infections actually progresses to invasive cancer, while the vast majority of them will eventually regress spontaneously. These data, together with the long delay between the onset of persistent infection and the emergence of the malignancy, indicate that HPV infection, although necessary, is not per se sufficient to induce cancer. Other factors have to be involved in the progression of infected cells to the full neoplastic phenotype. The present poor knowledge about the neoplastic progression mechanisms has dramatic consequences on the clinical side. As a matter of fact the current screening methods have a rather unsatisfactory predicting value making not possible to distinguish a potentially progressive CIN lesion from a lesion commites to spontaneous regression [75]. Thus while it is well known that the vast majority of cervical lesions regress spontaneously after a variable length of time and just a very minor part of them tend to persist and expose the patient to the risk of cancer, all of them have to be regarded as a potentially progressive lesions. Consequently a large number of patients have 39

42 to be treated as potentially progressive patients and are therefore submitted to ultimately unnecessary surgical treatments. In the search of a molecular marker able to predict the clinical outcome of dysplastic lesions many viral, host related and environmental factors have been taken into account and examined. Nevertheless HPV related carcinogenesis remains poorly understood and current screening protocol still wait improvements. Among viral factors, in addition to the obvious presence of the E6/E7 genes and the above mentioned viral integration, the multiplicity of infection, i.e.: the viral load is believed to be a determinant for lesion outcome. The host cell cycle markers such as p16 INK4a, Ki-67, p21, p27, cytokeratins, the host genetic background, as well as concurrent genital infection and life style have been deeply examined. Among environmental factors Oxidative Stress (OS), although appearing a good candidate as cancer promoting factor, has been comparatively neglected so far. OS is a condition arising from an increased production of reactive oxygen species (ROS) associated with a decreased antioxidant capability of the cell. ROS are constantly generated in aerobic cells by the incomplete reduction of molecular O 2 to H 2 O during mitochondrial oxidative phosphorylation, as well as during a number of processes such as inflammation, infections, mechanical and chemical stresses, exposure to UV and to ionising irradiation [76-78]. ROS cause oxidative damage to cell membrane lipids, proteins, and nucleic acids having the potential to induce both acute and chronic degenerative processes including ageing and cancer [79-82]. Indeed increased OS is well documented in transformed cells [81-83] and redox-sensitive networks are involved in cell proliferation and death pathways [84,85]. The cell will stand and counteract the OS by an array of 40

43 many different defense mechanisms, ranging from free radical scavengers and antioxidant enzymes to sophisticated and elaborated DNA repairing mechanisms. Despite this wide set of protecting mechanisms, oxidative damage eventually accumulates contributing to senescent decay and to a number of degenerative conditions including cancer. There are currently no clinically useful molecular markers for detecting the transition from infected cells, which proliferate simply in response to viral oncoprotein expression, and virally transformed cells, which have accumulated additional genetic and epigenetic changes during a latency period. Proteomic methodologies provide a general route to biomarker discovery analyzing cancer tissue [86-88]. After having seen in vitro how the OS evokes an orchestrated response affecting multiple cellular pathways that might contribute to the carcinogenesis mechanisms in HPV-infected cell [89] and how these cells resist to OS induced apoptosis [90], we were interested to study what happens in vivo. How are the levels of OS in different stages of the cervical cancer disease, beginning from the first step of HPV cell infection going to dysplastic lesions and finally to squamous cervical cancer? Considering that cervical intraepithelial neoplasia (CIN) is a dynamic process in which the lesions may persist, progress or regress, evaluating by redox proteomics the molecular mechanisms affected by OS may help us to have an insight in the possible role of OS in cervical cancer pathogenesis. 41

44 3.1.2 Oxidative Stress: ROS sources and Antioxidant systems Oxidative stress (OS) in a physiological setting can be defined as an excessive bioavailability of Reactive Oxygen and Nitrogen Species (ROS and RNS), which is the net result of the imbalance between cellular presence of ROS and the ability of cells to efficiently defend against them [91,92]. ROS sources and toxicity Reactive oxygen species are highly reactive oxygen radicals such as superoxide anion, hydroxyl radical, hydrogen peroxide and hypochlorite radical [93]. They can be produced both endogenously and exogenously. Endogenous OS could be the result of normal cellular metabolism in mitochondria (ROS areformed during the respiratory chain and by electron transport chains present also in the ER and nuclear membranes), metabolism involving oxidative phosphorylation (cytochrome P450, peroxisomes) and inflammatory cell activation (neutrophils and macrophages) during inflammation. Exogenous source of ROS such as: drugs, hormones, other xenobiotic chemicals, UV light, X-rays and gamma ray, can produce ROS by either direct mechanisms by ionizing the intracellular water and producing OH. and H 2 O 2 or by indirect mechanisms by activating an endogenous inflammatory response [94]. Superoxide anion is a fairly abundant ROS. Its formation takes place in different cell compartments, especially in the electron-rich aerobic environment in proximity of the inner mitochondrial membrane with the respiratory chain. Superoxide (as well as hydrogen peroxide) is also produced endogenously by flavoenzymes, e.g., xanthine oxidase, lipoxygenase and cyclooxygenase [95]. The NADPH-dependent oxidase of phagocytic cells, a 42

45 membrane-associated enzyme complex, constitutes an example of deliberate high-level O 2 - production. Superoxide can exist in the form of either O 2 - or, at low ph, hydroperoxyl (HO 2 ) [96]. The O 2 - form acts as a powerful nucleophile, capable of attacking positively charged centers, and as an oxidizing agent that can react with compounds capable of donating H (e.g, ascorbate and tocopherol). The most important reaction of superoxide radicals is its dismutation. One molecule is oxidized to oxygen, and the other is reduced to hydrogen peroxide. 2O 2 + 2H + SOD H 2 O 2 + O 2 Although H 2 O 2 molecules are considered reactive oxygen metabolites, they are not radical by definition; they can, however, cause damage to the cell at a relatively low concentration. Direct activities of H 2 O 2 include degradation of heme proteins with release of iron, inactivation of enzymes, and oxidation of DNA, lipids, SH groups, and keto acids [97]. Hydroxyl radical (OH ) is probably capable of doing more damage to biological systems than any other ROS. It can be generated by degradation of H 2 O from ionizing radiation. It is a powerful oxidizing agent that can react at a high rate with most organic and inorganic molecules in the cell, including DNA, proteins, lipids, amino acids, sugars, and metals [98]. Nitric oxide (NO ) is a reactive radical generated in biological tissues by specific nitric oxide synthases (NOSs), neuronal NOS, endothelial NOS (enos), and inducible NOS (inos), which metabolize arginine to citrulline with the formation of NO via a five-electron oxidative reaction. Under inflammatory conditions NO reacts with superoxide anion producing a much 43

46 more reactive molecule, peroxynitrite anion (ONOO ). This species may mediate several cytotoxic effects, such as the destruction of FeS centers in enzymes, persistent blockade of cytochrome c oxidase, which may lead to the release of free calcium ions from the mitochondrial matrix into the cell cytosol. Nitric oxide also reacts with lipophilic peroxyl radicals, important propagating species in the biological chain reaction of lipid peroxidation, to generate alkyl peroxynitrates (LOONO). These appear far more stable than ONOO-. It can cause DNA damage such as breaks, protein oxidation and nitration of aromatic amino acid residues in proteins (e.g, 3- nitrosotyrosine) [99]. All these ROS and RNS species are cytotoxic in that they cause peroxidation of membrane phospholipids, which results in increased membrane permeability, loss of membrane integrity, enzyme inactivation and structural damage to deoxyribonucleic acid, all of which lead to cell death. DNA damage (base pair mutations, rearrangements, deletions, insertions), deamination or nitration of DNA forming mutagenic species such as 8- nitroguanine, can result either in arrest or induction of transcription, alteration of cytoplasmic and/or nuclear signal transduction pathways, replication errors and genomic instability, damage to tumour-suppressor genes and enhanced expression of proto-oncogenes, all of which are associated with carcinogenesis [100, 101]. 44

47 Figure 9: Cytotoxic effects of ROS The effect of ROS in cell membrane is the peroxidation of polyunsaturated fatty acid residues of phospholipids, which are extremely sensitive to oxidation. The major aldehyde products of lipid peroxidation are malondialdehyde (MDA) and 4-hydroxy-2-nonenal (HNE). HNE is weakly mutagenic but appears to be the major toxic product of lipid peroxidation. In addition, HNE has powerful effects onsignal transduction pathways, which in turn have a major effect on the phenotypic characteristics of cells [102,103]. Proteins are another cell component sensitive to ROS damage, which lead to the fragmentation of polypeptide chain, oxidation of amino acid residue side chains, generation of protein carbonyl derivates and formation of protein-protein cross-linked aggregates. These damages cause loss of their structure and function affecting different cell signaling pathways, gene expression and inducing apoptosis (fig.9). 45

48 The presence of carbonyl groups in proteins has therefore been used as a marker of ROS-mediated protein oxidation, and several sensitive methods for the detection and quantization of protein carbonyl groups have been developed [ ]. Antioxidant systems Considering that ROS are normally formed during metabolism, cell is furnished of defense mechanisms to protect itself from this toxicity. Antioxidant systems counteract the effects of these radicals and thereby protect cell membranes from lipid peroxidation, and the cell to escape apoptosis. The ability of a tissue to buffer the effects of ROS is called total antioxidant capacity. There are enzymatic and non-enzymatic antioxidants. The most efficient enzymatic antioxidants involve superoxide dismutase, catalase and glutathione peroxidase [84]. Regarding non-enzymatic antioxidants the most efficient are Vitamin C, Vitamin E, carotenoids, thiol antioxidants (glutathione, thioredoxin and lipoic acid), natural flavonoids and melatonin [107]. Superoxide dismutase (SOD) catalyzes the dismutation of O 2 to O 2 and to the less-reactive species H 2 O 2. In humans there are three forms of SOD: cytosolic Cu, Zn-SOD, mitochondrial Mn-SOD, and extracellular SOD (EC-SOD). SOD reduces the highly reactive species by oxidizing the metal ion in its reactive site [108]. Catalase is located in a cell organelle called the peroxisome. The enzyme has a highly turnover rate and very efficiently promotes the conversion of hydrogen peroxide to water and molecular oxygen [109, 110]. 46

49 2H 2 O 2 CAT 2H 2 O + O 2 Glutathione peroxidase is an essential antioxidant system in cell metabolism. There are two forms of the enzyme, one of which is seleniumindependent (glutathione-s-transferase, GST,) while the other is seleniumdependent (GPx). Glutathione peroxidases reduce reactive species by oxidizing glutathione [84]. 2GSH + H 2 O 2 / ROOH GPx GSSG + 2H 2 O / ROH Vitamin C (or L-ascorbic acid or L-ascorbate) is an essential nutrient for humans and certain other animal species, obtained by the diet (fruits and vegetables). In living organisms ascorbate acts as an antioxidant by protecting the body against oxidative stress. When L-ascorbate carries out its reducing function, it is converted to its oxidized form, L-dehydroascorbate. L- dehydroascorbate, that can then be reduced back to the active reduced form by enzymes and glutathione. In vitro, ascorbate can act as antioxidant and scavenge a variety of ROS including hydroxyl, peroxyl, thyil, and oxosulphuric radicals and HClO and peroxynitrous acid [111]. Vitamin E is used to refer to a group of fat-soluble compounds that include both tocopherols and tocotrienols. Vitamin E can be found most abundantly in wheat germ oil, sunflower, and safflower oils. α-tocopherol, the most biologically active form of vitamin E, is an important lipid-soluble antioxidant. It performs its functions as antioxidant by the glutathione peroxidase pathway and protects cell membranes from oxidation by reacting with lipid radicals produced in the lipid peroxidation chain reaction. This would remove the free radical intermediates and prevent the oxidation 47

50 reaction from continuing. The oxidized α-tocopheroxyl radicals produced in this process may be recycled back to the active reduced form through reduction by other antioxidants, such as ascorbate, retinol or ubiquinol [112]. Thioredoxin (Trxs) proteins area group of small (10±12 kda) ubiquitous redox-active peptides, which have a conserved -Trp-Cys-Gly-Pro- Cys-Lys- catalytic site that undergoes reversible oxidation/reduction of the two Cys residues. The redox activity of this catalytic site is necessary for the biological activity of Trx [113]. Thioredoxin contains two adjacent SH groups in its reduced form that are converted to a disulphide unit in oxidized thioredoxin, which is undergoing redox reactions with multiple proteins. The reduction of the disulphide back to the dithiol form is catalyzed by thioredoxin reductase (TR), the source of electrons being NADPH. Some of the major functions of mammalian Trx proteins are to supply reducing equivalents to enzymes such as ribonucleotide reductase and thioredoxin peroxidase, and through thiol ± disulphide exchange, to reduce key Cys residues in certain transcription factors, resulting in their increased binding to DNA and altered gene transcription (fig.10) [114]. 48

51 Figure 10: Functions of Trx/TrxR system in the cell. Mustacich D. Biochem. J. 2000, 346 Intracellular accumulation of ROS can lead to cell death. Trx has a direct interaction with the apoptotic pathway through binding to ASK1 (a member of the MAPK family), controlling NF-kB action and the expression of other anti and proapoptotic genes like Bcl-2 family etc [115]. Glutathione reductase/glutathioneis another thiol redox system found in cells, which utilizes NADPH as its source of reducing equivalents. The glutathione system plays a key role in protecting cellular macromolecules from damage due to reactive oxygen species (ROS) and electrophilic species [116]. GSH can react with oxidized glutaredoxin, with electrophiles or oxidized macromolecules, and in a reaction with electrophiles catalysed by glutathione S-transferase (fig.11). 49

52 Figure 11: Scheme of glutathione reductase/glutathione system. Mustacich D. Biochem. J. 2000, 346, 1-8 The main protective roles of GSH against oxidative stress are: (1) It is a cofactor of several detoxifying enzymes, e.g. GST and others; (2) It scavenges hydroxyl radical and singlet oxygen directly, detoxifying hydrogen peroxide and lipid peroxides by the catalytic action of GPx; (3) It is able to regenerate the most important antioxidants, vitamins C and E back to their active forms [117] Overview on Proteomics and Redox Proteomics methods Proteomics is a method that analyses the proteome of the cells and tissues. It gives qualitative and quantitative information not only about the native proteins but also about their post translation modifications, their structures and functions [118]. Proteomics uses a 2D polyacrylamide gel electrophoresis separation of proteins, which consists in separating the proteins based on their two physicochemical properties. In the first dimension proteins are separated by their isoelectrical point and in the second dimension by their molecular weight. Thousand spots are present in a 2D Coomassie 50

53 stained gel image, each of them represents a protein. Multiple gel replicates of controls and experimental samples are analysed by PD-Quest software. It compares the spot intensities between groups, to calculate differences in expression between various samples and identifies the statisticaly differently expressed spots. They are further digested and identified by the mass spectrometry. Proteomics is revealed to be a useful method to identify different proteins and specific pathways involved in diseases pathogenesis, especially in neurodegenerating diseases, such as Parkinson and Alzheimer disease, andit can provide important information and identify pathways involved in [119]. Redox proteomicsis an analysis used to identify oxidatively modified proteins, i.e. proteins modified by reactive species of oxygen. The modification of proteins by the OS consists in the formation of carbonyl groups and protein-bound HNE. The proteome of cells or tissues is separated in 2D electrophoresis and the gel is blotted in a nitrocellulose membrane. Due to the absence of an antibody that recognises the carbonyl groups, proteins carbonyls are derivatized to hydrazones by chemicals such as 2,4- dinitrophenylhydrazine before the separation of proteins or following the 2D separation of proteins. The DNP adducts are visualized with anti DNP antibody. The 2D blots and gels are separately analysed by the PD-Quest software. The software normalizes the level of spot oxidation with the level of its expression in the gel, identifying the spots differently oxidized in disease versus control groups. The oxidatively modified proteins are digested and analysed by mass spectrometry (fig.12) [120]. 51

54 Figure 12: Identification of oxidatively modified proteins by Redox Proteomics OS has been shown to be involved in many diseases such as neourodegenerating and cancer disease. Protein modifications due to OS can alter their structure and function. Identifying these modifications can give an insights into the mechanisms of diseases and may help to identify diseaseassociated markers or targets for specific therapy. The field of redox proteomics is rapidly expanding and evolving and it is now being developed for clinical diagnosis and in establishing biomarkers for disease and drug discovery [121]. 52

55 3.2 The aim of the work The aim of this work was to identify new molecular markers correlating with neoplastic progression in HPV 16 transformed cervical cells. Hopefully, these markers could improve the positive predictive value of current screening protocols and might provide precious insight into the biology of HPV-dependent carcinogenesis. To this purpose, the viral load, the viral genome physical status, the expression of OS related proteins and the pattern of oxidative adducts on cell proteome (the redox proteomics) were assessed on cervical tissues from patients with HPV 16 infection. Overall, our results indicated that an increased oxidant environment associated with an increased antioxidant activity in dysplastic tissues affect proteins involved in cell differentiation pathways and pro-survival network. Their potential role in cancer progression is discussed. 3.3 Material and Methods Patients enrolment Patients attending either the Gynecological Department of the Regina Elena Cancer Institute of Rome (Italy) or the Oncological Hospital at the University Hospital center Mother Teresa of Tirana (Albania) and presenting with clinically/colposcopic evidence of dysplastic lesions or invasive cervical cancer, were invited to participate to the study. Women with evidence of uterine fibroleiomyoma were also invited to participate as non neoplastic controls. Eligibility criteria: To be eligible patients had to be at least 18 years old, to be not pregnant, not suffering from any other diseases apart the reason for asking gynaecological advise, to have no/to have never 53

56 suffered from other neoplastic disease. The study design and enrolment criteria had been approved previously by the Regina Elena s local Ethical Committee. During the period, January 2008 to December 2009, 73 women were selected among those providing a full written informed consent and matching the above inclusion criteria. At entry, enrolled patients underwent a full gynaecological examination. Ecto-cervical cells were collected and stored in Thin-Prep transfer medium for liquid based cytology and virological assays. Histological samples were taken and divided into two halves, one used for standard histo-pathological evaluation and processed under current criteria. The second one was immediately frozen in dry ice/ethanol bath and used later for viral assay and for protein extraction. Any decision about patients diagnosis, treatment and follow-up, had been based merely on current clinical criteria irrespective of any research need Viral analyses DNA was extracted from a small piece of tissue by the QIAamp DNA Mini Kit (QIAGEN Gmbh, Hilden, Germany) used according to the manufacturer s instructions. The purity and concentration of DNA preparations was evaluated by the A260/A280 ratio measured in a Nanodrop 2000C microcuvette spectrophotometer (Thermo Fisher Scientific Inc. Waltham, MA). Samples adequacy to PCR analysis was assessed by a parallel β -globin gene amplification with primers GH-20/PC-04 (tab. 1) as described by Saiki et al [62]. For HPV detection the samples were amplified using the MY09/MY11 primer couple [59] with minor modifications. Briefly a 50 ng of total DNA were mixed with 500 nm of MY09 and MY11 primers (table 1), 1.5 mm MgCl 2, 200 µm each of dntp, 5 µl of 10x reaction buffer and 1 54

57 U of Platinum TAQ DNA polymerase (both purchased from Invitrogen Life Technologies, s.r.l. - San Giuliano Milanese, Italy), in a 50 µl final volume. In both cases amplification consisted in an initial step at 95 C for 150 sec for Polymerase activation, followed by 35 cycles of denaturation at 95ºC for 30 sec, annealing at 56ºC for 30 sec and of extension at 72ºC for 40 sec plus a final cycle with a 10 min long extension for optimal chain termination. Amplified products were separated by 2.5% ethidium bromide stained agarose gel electrophoresis and visualized by gel direct inspection under UV-B trans-illumination. MY negative samples were further examined with the GP5+/GP6+ primers as described [61]. These primers, spanning a much shorter L1 segment have the potential to successfully detect HPV from partially degraded suboptimal samples as well as from samples carrying low viral genome amounts. The GP5+/GP6+ PCR conditions were the same as above apart from a 3 mm MgCl 2 concentration and an annealing step at 45 C for 50 sec. Several cell line were used as positive and negative control. The HPV-16 positive Siha, CaSki squamous cell carcinoma cell lines and the HPV 16 in vitro transformed human keratinocytes (HK- 168 cell line) previously described were used as positive controls [90, ]. HaCaT cells [125] were used as negative controls HPV typing Viral typing was detected by direct sequencing of amplified products. Briefly, amplicons were purified by silica gel affinity chromatography (QIAqick PCR purification kit QIAGEN GmbH), resuspended in a 20µl final volume with 600nM the upstream primer used for the amplicon generation (i.e., either MY11 or GP5+) and transferred to the BIO-FAB Research s.r.l. (Pomezia, Italy) where they were sequenced by the BigDye 55

58 Terminator 1.1 Cycle Sequencing Kit (Sanger method). Cycle sequencing products have been purified by Sephadex G-50 (GE Healthcare Bio- Sciences Medical Diagnostics, Milan, Italy) gel filtration. Sequences were analyzed by capillary electrophoresis in 3730 DNA Analyser (Applied Biosystems, Inc., Foster City, CA) and aligned to prototype viral sequence through the BLAST resource at the NCBI ( Viral load and physical status determination HPV-16 positive samples the viral load was determined by a SYBR Green quantitative PCR (qpcr) procedure. Briefly total DNA was extracted and purified from 10 7 Siha cells and adjusted to 1.0 ml final volume in Tris EDTA Buffer ph 8.0. This cell line is known to have a nearly triploid genome [126] carrying 2 integrated, almost complete copies of HPV-16 [122, 127] per cell, thusit contains the viral oncogene E6, the human β globin gene and presents a disruption of E2 ORF. Tenfold serial dilutions were derived from the above stock solution producing a titration series of 10 4 ; 10 3 ; 10 2 ; 10 1 ; 10 0 cell genome/µl and used to derive target-specific standard curves used to quantify, in real time PCR, the E6 gene and the β globin gene in our samples. The average lesions viral load was then calculated as the E6/βglobin ratio and expressed as viral Copies per Haploid Cellular Genome (CHCG). All PCR were carried out in a IQ4 BIO-RAD Cycler with Iq Sybrgreen Supermix (both from Bio-Rad Italia Srl, San Donato Milanese, Italy) Reaction were set up in 25 µl final volume containing 1x reaction mixture; 500 nm of each primer and 50 ng/µl of standard or sample DNA. Amplification conditions consisted in: TaQ polymerase thermal activation at 95 C for 150 sec followed by 35 cycles of denaturation at 95 C for 30 sec; annealing at 56 C for 30 sec; extension at 72 C for 40 seconds; sample 56

59 reading at 75 C for 10 sec. A final cycle extension at 72 C for 10 min followed by a melting curve ranging from 70 C to 95 C with 0.5 C incremental temperature/10 second step were incorporated. Primers were chosen to give amplicons of similar length with close annealing temperature in order to keep to a minimum the bias due to differential target amplification. The E6 primers used for the viral load (E6/β-globin ratio) encompassed the region from nucleotide (nt) 26 to nt 233 generating a 207 bp long amplicon. The primers for β-globin (GH-20 and PC04) spanned the start codon of the human β-globin amplifying a 260 bp tract (Table II). PRIMERS SEQUENCES Bp MY09 CGTCCMARRGGAVVACTGATC 450 MY11 GCMCAGGGVVCATAAYAATGG 450 GP5+ TTTGTTACTGTGGTAGATACTAC 150 GP6+ GAAAAATAAACTGTAAATCATATTC 150 GH20 GAAGAGCCAAGGACAGGTAC 260 PC04 CAACTTCATCCACGTTCA CC E6S26 AAGGGCGTAACCGAAATCGGT E6AS233 CATATACCTCACGTCGCAG E2S3438 CTTGGGCACCGAAAGAAACAC E2AS3789 TTGGTCACGTTGCCATTCAC E2S2734 AGGACGAGGACAAGGAAAA E2AS3846 GGATGCAGTATCAAGATTTG 1139 Table 2: Primer sequences used in PCR The copy number of the viral genes E6 and E2 and the cell number (i.e: the number of human β-globin copies) in experimental samples were 57

60 evaluated referring the amplicon threshold cycle (assayed in duplicate) to its specific standard curve (assayed in triplicate). Three different methods were used to obtain information about the physical status of viral genome. I) The presence of a PCR detectable full length HPV 16 E2 ORF (1112 bp) [128] assumed as an index for the presence of episomal viral genomes. II) The ratio between the E2 and the E6 [129], assumed as a marker of episomal state when close to the unit (i.e. even amount of E2 and E6) or of integrated state if close to zero (i.e. negligible amount of E2 respect to E6) was evaluated in real time PCR [130]. For the quantification E2 and E6 genes present in each samples, in real time PCR, two standard curves, one for each gene, were constructed using 10 fold serial dilution (10 4 copies copies) of HPV16 plasmid. The physical status of the virus was further characterised by comparing viral loads from parts of E2 ORF that are mostly deleted during integration, and those of E6 ORF. The interval of E2/E6 ratio in the episomal viral form, determined in real time PCR using the serial dilution of HPV16 plasmid, was considered as the mean value of this ratio ± 2SD. The E2 primers used to estimate the E2/E6 ratio, i.e.: the episomal/integrated status of viral genomes, spanned the 177bp long region between the nt 3383 and nt 3560 of E2 ORF, located in a region that is most often deleted on HPV16 integration [131]. The E6 primers were the same used for viral load assay. III) The Rolling Circle Amplification (RCA) [132,133] as a positive proof of episomal forms. RCA was performed with the TempliPhi 100 Amplification Kit (Amersham, UK) used according to the manufacturer s instructions. Briefly 1µl of sample DNA was combined with reaction buffer, dntp, random enzymes and Phi 29 DNA polymerase in a 11.7 µl final 58

61 volume and incubated at 30 C for 18h. The isothermal amplification was then interrupted by enzyme inactivation at 65 C for 15 min and amplification products digested with the single cut Bam-HI enzyme. The presence of HPV- 16 episomal genomes was then revealed by a band of the expected 7900 bp size on EB stained agarose gel electrophoresis Protein extraction Fresh or frozen histological samples were cut in small pieces and incubated for 20 min in ice with the lysis buffer (10 mm Hepes ph 7.9, 10 mm K 2 EDTA, 5 mm NaCl, 1% TritonX-100, 10 mm β-mercapto-ethanol, aprotinine 5 mg/l). Lysates were then disrupted with a mechanical mincer (Ultra-Turrax IKA T10) and successively with a potter device, until a fine cloudy suspension was obtained. The suspension was then hen clarified in a JA-21 Beckman Super-centrifuge at 16,000 g for 20 min at +4 C and protein concentration in the supernatant was determined by the Coomassie Plus Pierce Protein Assay (Rockford, IL, USA). Samples were then divided into aliquots and stored at -80 C until use Western blot Endoplasmic Reticulum protein 57 (ERp-57), glutathione S- transferase (GST), inducible nitric oxide synthase (i-nos) and thioredoxin reductase 2 (TRX-R2) levels were evaluated by Western blot analyses. Sample aliquots (40 µg of protein) were subjected to 12.5% SDS-PAGE and electroblotted (1 h at 100 V) to nitrocellulose membranes (Bio-Rad) using 25 mm Tris, 192 mm glycine and 20% (v/v) methanol. Equal protein loading was confirmed by staining with 0.2% v/v Ponceau S in 7% acetic acid. Blotted membranes were blocked with 3% albumin in T-TBS and challenged with appropriate primary antibodies, namely anti-erp-57 rabbit polyclonal 59

62 antibody, anti-gst mouse monoclonal antibody, anti-inos2 rabbit polyclonal antibody and TRX-R2 goat polyclonal antibody for 1 h at room temperature. Unbound antibodies were removed by washing twice with Trisbuffered saline containing 0.1% Tween 20, for 5 minutes. The membranes were then incubated with horseradish peroxidase-conjugated secondary antibody diluted 1:5000. Protein bands were visualized with ECL PlusTM (Amersham) according to the manufacturer s protocol. Blots were scanned on a GS880 densitometer (Biorad) and quantified by QuantityOne image software Proteomics Proteins (200 µg) were precipitated in TCA (15 % final). After 10 min in ice, pellets were collected by centrifugation at x g for 2 min, washed three times with 0.5 ml of ethyl acetate/ ethanol (1:1) and resuspended in 200 µl of fresh rehydration buffer (8 M urea, 2 M thiourea, 20 mm dithiotreitol, 2.0% CHAPS (w/v), 0.2 % Biolytes, bromophenol blue) and placed in agitation for 3 hours. Solubilized proteins were finally sonicated twice for 30 sec. For the first-dimension electrophoresis, 200 µl of sample solution were applied to a ReadyStrip IPG strip ph 3-10 (Bio-Rad). The strips were soaked in the sample solution for 1 h to allow uptake of the proteins. The strips were then actively rehydrated in Protean IEF Cell Apparatus (Bio-Rad) for 16 h at 50 V. The isoelectric focusing was performed at 300 V for 2 h linearly; 500 V for 2 h linearly; 1000 V for 2 h linearly, 8000 V for 8 h linearly and 8000 V for 10 h rapidly. All the processes above were carried out at room temperature. The focused IEF strips were stored at 80 C until second dimension electrophoresis was performed. For second dimension electrophoresis, thawed strips were equilibrated for 10 min in 50 mm Tris- 60

63 HCl (ph 6.8) containing 6 M urea, 1% (w/v) sodium dodecyl sulfate (SDS), 30% (v/v) glycerol, and 0.5% dithiothreitol, and then re-equilibrated for 15 min in the same buffer containing 4.5% iodacetamide in place of dithiothreitol. 12% Precast criterion gels (Bio-Rad) were used to perform second dimension electrophoresis. Precision ProteinTM Standards (Bio-Rad) were run along withthe sample at 200 V for 65 min. After electrophoresis, the gels were incubated in fixing solution (7% acetic acid, 10% methanol) for 20 min. Approximately 40 ml of Bio-Safe Coomassie Gel Stain (Bio-Rad) were used to stain the gels for 1 h, on a gently continuous rocker. The gels were placed in deionised water overnight for de-staining Protein oxidation measurement Protein oxidation was measured according to Butterfield et al. [134]. Briefly, samples (5 ml) were added with 5 ml of 12% SDS and derivatized with 10mM 2,4-dinitrophenylidrazine (DNPH) at room temperature for 20 min. Samples were neutralized with 7.5 ml of neutralization solution (2 M Tris in 30% glycerol). Derivatized samples (250 ng) were then blotted onto a nitrocellulose membrane under vacuum using a slot-blot apparatus (Bio- Rad). Membranes were blocked with 3% BSA in TBS-T for 1 h and next incubated with rabbit antibody to protein-bound DNP (diluted 1:150) for 90 min. After washing with TBS-T, membranes were incubated with anti-rabbit IgG alkaline phosphatase secondary antibody (1:5000) in TBS-T for 1 h at room temperature. The membrane was washed in TBS-T and developed using a solution of NBT (0.2 mm) and BCIP (0.4 mm) in alkaline phosphate buffer (0.1 M Tris, 0.1 M NaCl, 5 mm MgCl 2 ; ph 9.5). Dried blots were quantified using QuantityOne image analysis (Bio-Rad). 61

64 3.3.9 Redox proteomics To identify oxidized proteins, samples (200 µg proteins) were incubated at room temperature for 30 in four volumes of 10 mm 2,4 dinitrophenylhydrazine (DNPH) in 2N HCl for protein carbonyl derivatization, according to Levine et al.[105]. Gels were subjected to 2-DE as above described. Then, the proteins from the second dimension electrophoresis gels were transferred to nitrocellulose membrane using Criterion Blotter apparatus (Bio-Rad) at 100V for 1 h. The carbonylated proteins were challenged detected by a DNP-protein adduct primary rabbit antibody (Millipore Corp., MA, USA) 1:100 followed by a secondary goat anti-rabbit IgG alkalinephosphatase conjugated antibody 1:2000. (Santa Cruz, CA, USA) and revealed by 5-bromo-4-chloro-3-indolyl phosphate/ nitro blue tetrazolium (BCIP/NBT) solution (SigmaFast tablets, SigmaAldrich Srl Milan, Italy) Image Analysis The 21 gels (n=7 controls, n=7 dysplasia and n=7 carcinoma) and 21 nitrocellulose blots were scanned and saved in TIF format using a GS-800 densitometer (Bio-Rad). PD-Quest 2D Analysis software (version 7.2.0, Bio- Rad) was used for matching and analysis of visualized protein spots among differential gels and membranes. The anti-dnp immune-reactivity of individual proteins was normalized to protein content evaluated by the intensity of Coomassie blue stained spots. After completion of spot matching, the normalized intensity of each protein spot from individual gels was compared among the groups using statistical analysis. 62

65 Trypsin digestion and protein identification by mass spectrometry Selected spots were manually excised from gel and submitted to trypsin proteolysis [94]. Briefly, after 15 min de-staining with 50 mm ammonium, 10 min with 50% acetonitrile in 50 mm ammonium bicarbonate and 15 min 100% acetonitrile, spots were vacuum dried and then digested with 100 ng of trypsin (Trypsin Gold, Mass Spectrometry Grade, Promega, Madison, WI, USA), in 10 ml of a 25 mm ammonium bicarbonate digestion buffer at 37 C overnight. An aliquot of the peptide mixture was mixed with an equal volume of a-cyano-4-hydroxy-trans-cinnamic acid matrix solution (5mg/ml) in 70% acetonitrile containing 0.1% TFA (v/v) and spotted onto an appropriate MALDI target plate. MALDI-ToF MS analyses were performed in a Voyager-DE STR minstrument (Applied Biosystems, Framingham, MA, USA) equipped with a 337 nm nitrogen laser and operating in reflector mode. Mass data were obtained by accumulating several spectra from laser shots with an accelerating voltage of 20 kv. Two tryptic autolytic peptides were used for the internal calibration (m/z and ). Data were analysed by MoverZ program (v.2002, according to default parameters. Identification by peptide mass fingerprint (PMF), with the monoisotopic mass list, after exclusion of expected contaminant mass values bypeakerazorprogram( eakerazor.html), was performed using the Mascot search engine (v. 2.2) against human SwissProt database [SwissProt 2010_04 ( sequences; residues)]. Up to one missed cleavage, 50 ppm measurement tolerance, oxidation at methionine (variable modification) and carbamidomethylation at cysteine (fixed modification) were considered. 63

66 Identifications were validated when the probability-based Mowse protein score was significant according to Mascot [135] Protein identification by LC-MS/MS Selected peptide mixtures were separated by on-line reverse-phase (RP) capillary liquid chromatography and analyzed by electronspray tandem mass spectrometry (ESI-MS/MS). The samples were loaded onto a 15 cm reverse-phase fused-silica capillary column (BioBasic-18, inner diameter 180 mm, 300 Å, 5mm, Thermo Scientific), using Dionex Ultimate 3000 system (LC Packings, Dionex, Amsterdam, The Netherlands), by an autosampler. Peptides were fractionated with a 70 min gradient from 5% to 95% acetonitrile in 0.1% of formic acid at a flow rate of 2 ml/min. The HPLC system was connected to a linear ion trap-orbitrap hybrid mass spectrometer (LTQ-Orbitrap Discovery, Thermo Fisher Scientific GmbH, Bremen, Germany) equipped with a nanoelectrospray ion source (Thermo Fisher Scientific), operating in the positive ionization mode with a spray voltage of 1.9 kv. The eluted peptides were detected in a precursor MS scan mode by Orbitrap ( m/z, resolution at m/z 400), followed by sequential data-dependent MS/MS scans in which the three most abundant ions were fragmented in CID and analyzed in the linear trap (minimal signal required 2000, isolation width of 3 m/z, normalized collision energy 35%, removal of 1+ ions or ions with unassigned charge state and selection of 2+, 3+, and 4+ ions) Statistical analysis Statistical significance was assessed by a two-tailed Student s t-test, the method of statistical analysis most appropriate for proteomic analysis of 64

67 small number of protein spots [136] and to analyze differences in protein levels between dysplasia and carcinoma. P values of less than 0.05 were considered statistically significant. The significance of the change in carbonylation of specific proteins in the proteomics study was evaluated via nonparametric Mann-Whitney- Wilcoxon test. P < 0.05 were considered significant for comparison between control and experimental data. 3.4 Results Patients During the period from January 2008 to December 2009 a total of 87 patients entered the study. Among them 35 had an invasive squamous cell carcinoma (SCC), 1 an adeno-carcinoma, 12 were affected by cervical dysplastic lesions and 23 were suffering for a uterine fibroleiomyoma (for the sole purpose of this work here are considered as control patient). The remaining 16 patients turned out to be affected by other inflammatory or chronic/degenerative pelvic diseases and together with the adeno-carcinoma one were excluded from further analyses. The mean age of the patients at the time of the diagnosis was 52 years old (52±10) for the controls, 35 years old (35 ± 8) for the dysplastics and 51 years old (51 ± 20) for patients with invasive cancer. All patients were multiparae, with no abortion history and no smokers Viral assay The DNA adequacy was confirmed in all 70 examined samples by the presence of the β-globin gene in PCR amplification (fig.13). 65

68 9C 4C 13D 16D 4K 6K 12K HPV16 MVIII H 2O 260bp 320bp 242bp Figure 13: PCR amplicons of β-globin gene of controls (C), dysplastic(d) and SCC (K) samples,ran in ethidium bromide stained agarose gel. H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVIII: Molecular marker VIII.Inverted image. The presence of HPV-16 was confirmed in 25/35 patients with invasive SCC; in 6/12 patients with dysplastic lesion (HSIL) and in 7/23 control patients. Other HR-HPV, like HPV 18, 33 and 58, were found respectively in 5/35, 2/35, 1/35 patients with invasive SCC. We found also multiple infections in 2/35 patients with SCC. One of them was simultaneously infected with HPV 16, 33, 35 and the other one with HPV 33, 58, all HR-HPVs (fig.14). Figure 14: PCR GP5+/GP6+ amplicons from a selelction of controls (C), dysplastic(d) and SCC (K) samples, ran in ethidium bromide stained agarose gel. H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVIII: Molecular marker VIII.Inverted image. Only HPV16 positive samples that will be lately used for 2DE analysis, were further characterized by determining the viral load and virus physical status. 66

69 3.4.3 Viral load The viral load and the physical status of viral genome have been claimed to be relevant determinants in HPV infection outcome and in clinical evaluation of dysplastic and neoplastic lesions. Various methodologies are described for the evaluation of viral load and physical status [ ]. In this study we worked with the Sybr green method [130] quantifying the E6 gene (208 bp) (fig.15), E2 and of the human β globin gene based on a standard curve generated by a logarithmic dilution series of Siha. The viral load was calculated as E6/β globin ratio. 44C 45C 47C 49C 13D 16D 21D 24K 3K 9K 12K 17K HPV16 MVIII H 2O 208bp 501bp 242bp 190bp Figure 15: PCR amplicons of E6 gene of controls (c), dysplastic(d) and SCC (k) samples, ran in ethidium bromide stained agarose gel. H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVIII: Molecular marker VIII. Inverted image. The viral load obtained for each specimen is plotted in figure16. We found, in control tissue (i.e.: cervical tissue devoid of clinically evident dysplastic lesion) a mean viral load of 0.96 x 10-2 CHCG. All but one single value clustered close to the mean value and all of them were clearly below the 10-1 CHCG level. A very similar finding was obtained for dysplastic samples with a mean viral load of 2.20 x 10-2 CHCG and much less scattered individual values. Conversely a mean value of 1.65 x 10 2 viral CHCG was observed among invasive cancer. The difference of the viral load found in controls and dysplasia versus the one found in scc samples were statistically significantly. The SiHa and CaSki cell lines, here used as low and high ratio 67

70 positive control, in agreement with reported data [127, 128, 130], consistently yielded values around 10 0 and viral CHCG, respectively (data not shown). Figure 16:Scatter plot ofviral loads in controls, dysplastic and neoplastic tissues calculated as E6/β-globin ratio Viral genome physical status The interest for viral genome physical status in lesions is based on the view that viral integration, once occurring within the E2 sequences, involves two irreversible step in viral carcinogenesis: first, it permanently inserts the E6 and E7 oncogenes in the host genome thus abolishing any eventual clearing of the viral infection and second, it abrogates the physiological negative regulatory control of E2 on E6 and E7 transforming genes. Three different methods were used to gain information about the physical status of viral genomes such as: PCR amplification of the full length 68

71 E2 gene, whose absence is assumed as an index of viral genome integration, E2/E6 ratio quantified in real time PCR and RCA (rolling circle amplification) method. The presence of the full length E2 region was detected in all specimens but the eight neoplastic samples (Fig.17) bp 9c 34c 44c 35c 13d 16d 21d20k 22k 23k 24k 36k 37k 1k 11k MVIII HPV16 H 2O Figure 17: Full length E2 gene amplicons from a selection of controls (c), dysplastic(d) and SCC (k) samples, ran in ethidium bromide stained agarose gel. H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVIII: Molecular marker VIII.Inverted image. The E2/E6 ratio values, quantified in real time PCR, ranged between 0 and 0.02 in a minority of cases (n=8) that were all found in SCC and in two of them E2 was not detected, indicating complete virus integration (Fig. 18). In 19 cases the E2/E6 ratio ranged between 0.02 and 0.65 (values compliant with the presence of mixed forms). These included 14 SCC patients, 1 control and 4 dysplastic patients. The further 11 cases ranged between 0.65 and 2.85 (values compliant with the predominance of the episomal form) found in 6 controls, 2 dysplastic and 3 SCC patients. 69

72 34C 45C 13D 16D 21D 2K 3K 6K 20K 1K 36K HPV16 MVIII H 2O 351 bp Figure 18: PCR amplicons of ORF E2 gene of controls (C), dysplastic (D) and SCC (K) samples, ran in ethidium bromide stained agarose gel, used in RT-PCR for E2 gene quantification. H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVIII: Molecular marker VIII. Inverted image. With the RCA method conversely, episomal forms were positively identified just in one normal sample and in two neoplastic ones (fig.19). 9C 35C 44C 45C 47C 13D 16D 21D 24D 1K 11K 14K 16K 20K 23K M II HPV16 Caski H 2O 23000bp 7900 bp Figure 19: RCA amplicons of control (C), dysplastic (D) and SCC (K) samples digested with Bam-HI restriction enzyme, ran in ethidium bromide stained agarose gel H 2 O: no template control, HPV16: positive HPV16 plasmid control, Caski: positive control, MII: Molecular marker II. Inverted image Expression levels of stress response proteins Expression levels of Endoplasmic Reticulum protein 57 (ERp57), Glutathione S-Transferase (GST), inducible Nitric Oxide Synthase (inos) and Thioredoxin Reductase 2 (TrxR2) were evaluated in control, dysplastic and neoplastic tissues (Figure 20). 70

73 Figure 20:Expression levels of stress markers (ERp57, GST, TRX-R2 and inos). Protein expression levelsin CTR, DYS and SCC cervical tissues were measured by Western blot analysis using specific antibodies for ERp57 (A), GST (B), TRX-R2 (C) and inos (D). Immunoblots were scanned by densitometry and all values were normalized to β-actin levels. Densitometric values shown are given as percentage of the control group, set as 100%. Data are expressed as mean ± SEM. p < 0.05 versus control (Student's t-test). The expression of ERp57 in dysplastic lesion was slightly increased respect to control tissue (110%). In neoplastic tissues ERp57 expression was further increased to a 143 % respect to control (Fig. 20, A). GST, a detoxifying enzymes, was sharply increased in dysplastic and in neoplastic cells (up to 180% and 600 % above control respectively) (Fig. 20, B). 71

74 TrxR2, specifically localized in mitochondria, is believed to play an important role in the regulation of cellular redox status. As it can be seen in Fig. 20, C, TrxR2 expression was significantly increased in dysplastic lesions and to a lower extent in neoplastic lesions as compared with control tissue (175% and 125% above control, respectively). i-nos, the inducible isoform of Nitric Oxide Syntase, is frequently involved in malignant pathologies. As it can be seen (Fig. 20, D) it was found to be more expressed in control specimens than either in dysplastic (65% of control) or neoplastic (25% of control) samples Protein oxidation To assess the extent of total protein oxidation, protein carbonyl levels were evaluated by slot-blot analysis (figure 21). Figure21: Total protein oxidation. Top: Quantification of protein carbonyls levels in controls (CTR), dysplastic (DYS) and neoplastic (SCC) tissues. Bottom: Protein carbonyl slot- blots from CTR, DYS and SCC samples. 72

75 With respect to control samples, protein carbonyls were significantly increased in dysplastic tissues, while levels detected in neoplastic tissues were surprisingly similar to control ones Identification of carbonylated proteins For redox proteomics analysis, a set of 7 samples of SCC were selected to be compared with the 6 dysplastic tissues and the 7 control samples. Two representative 2D gels, and the corresponding blots, from control and dysplastic samples are pictured in figure 22. Figure 22: Oxidized protein detection by redox proteomics (DYS vs. CTR) Top: 2D maps of CTR (left) and DYS (right) cervical tissues. Bottom: 2D carbonyl immunoblots of CTR (left) and DYS (right) cervical tissues. The spots showing significant increased carbonyl levels are labelled. The identified proteins are listed in table III. 73

76 Five proteins were found to be more oxidized in dysplastic tissues compared with controls, namely cytoskeletal Keratin 6 (fragments/isoforms A, B & C), Cornulin, Actin, Retinal Dehydrogenase (RDH) and Glyceraldehyde-3-phosphate dehydrogenase (GADPDH). These are proteins involved in cytoskeleton scaffolding (keratin and actin), epidermal differentiation (cornulin, RDH), and intermediate metabolism (GADPDH). Redox proteomics analysis was also performed to compare dysplastic and neoplastic samples. Results indicated that five proteins, namely Serpin B3, Annexin 2, ERp57, PDI A3 and PIN-1 were more oxidized in dysplastic samples versus neoplastic ones (Fig. 23). Figure 23: Oxidized protein detection by redox proteomics (DYS vs. SCC) Top: 2D gel maps of DYS (right) and (left) cervical tissues. Bottom: 2D carbonyl immunoblots of DYS (right) and SCC (left) cervical tissues. The spots showing significant carbonyl levels are labelled. The identified proteins are listed in table III. 74

77 Table III shows the proteins successfully identified by mass spectrometry along with the peptide hits, sequence coverage, Mw and pi values and the increase/decrease of specific carbonyl levels, indexed as fold oxidation. Protein Keratin, type II cytoskeletal 6A Keratin, type II cytoskeletal 6B Keratin, type II cytoskeletal 6C Glyceraldehyde-3- phosphate dehydrogenase Swiss Protein code P02538 P48668 P04259 Dysplasia vs Controls Theorical Sequence Mw/pI Coverage % 60293/ / / Score Fold oxidation P / Cornulin Q9UBG / Retinal P / Dehydrogenase 1 Actin, cytoplasmic 1 Q96HG / Actin, cytoplasmic 2 P / SCC vs Dysplasia Glyceraldehyde-3- P / phosphate dehydrogenase Peptidyl-prolyl cistrans P / isomerase A Protein disulfideisomerase P / A3 Serpin B3 Q8IXI / Annexin A2 P / Table III: List of oxidized proteins successfully identified by mass spectrometry along with the peptide hits, sequence coverage, Mw and pi values and the increase/decrease of specific carbonyl levels, indexed as fold oxidation. 3.5 Discussion Viral sample characterization Infection with HR-HPV is considered the main ethio-pathogenic factor involved in cervical cancer development. By using PCR method, five 75

78 different HR-HPV, namely types 16,18,33,35, 58, were found in our samples. As expected HPV16 was the most frequently viral type found. Its prevalence increased from 30% in the controls to 50% in dysplastic and up to 72% in cervical cancer tissues. The second most frequent HPV type was HPV 18 found only in 5 cervical cancer tissues, followed by HPV 31, 33, 35 and 58. Considering the high prevalence of HPV transient infection, detection of the virus alone is not sufficient to identify women at risk for developing cervical cancer. Viral load and viral genome physical status have been claimed to be relevant determinants in HPV infection outcome and in clinical evaluation of dysplastic and neoplastic lesions. Thus HPV 16 infected patients were further evaluated for their viral load and viral genome physical status. We based the quantification of the viral E6, and of the human βglobin gene on a Siha cell standard curve. This is an extensively characterised cell line known to host a single, integrated, transcriptionally active, almost complete HPV16 copy per haploid human genome [ , 127]. Its use allows expressing the viral load in terms of virus/cell ratio, the actual functional unit in viral carcinogenesis, and eliminates the need to quantify DNA by spectrophotometry, a source of potential error implied in the method of expressing the PCR target abundance relative to gdna mass. Both in normal and dysplastic lesions, viral loads below 10-1 CHCG were found. The average HPV 16 copy number per cell found in controls and dysplastic lesions were not significantly different from each other. In contrast, in carcinoma samples a sharply higher mean lesional viral load was found (1,65 x 10 2 CHCG). This value was significantly different (p<0.05) in respect to the mean viral load found in controls and dysplastic lesions. In SCC group (fig 12) values scattered in the range mainly between 10 0 to 76

79 10 3 CHCG. This result is consistent with the finding that SSC cell lines have largely different copy number [127] but it may also be influenced by the tissue sampling, where peripheral areas of samples with non neoplastic (and non viral) epithelial or stromal cells are dissected during tumour excision. Higher levels of viral load are found as the disease progress from dysplasia to cervical cancer. Based on literature data, considerable viral load variations has been observed within histopathological grades of the disease [140, 142], making it hard to define cut-off values between these stages and help to identify if a simple infection will progress to CIN or the last one to cervical cancer. Considering the physical status of the viral genome within the infected cell, three different conditions can occur: the viral genome can be entirely found in episomal form (i.e.. as a circulatory closed free genetic element, structurally independent from the host genome), totally integrated in host genome or can be found in the mixed form where both episomal and integrated forms are present in a cell. Viral integration was evaluated by two E2 based methods plus a RCA based method. The integrated virus form was evaluated by amplifying the full length E2 ORF, which is the one reported to be mostly deleted during virus integration in the host genome. It resulted to be absent only in eight cervical cancer lesions, presenting the presence of pure integrated viral forms. The pure episomal virus form was evaluated with the RCA method, by the presence of 7900 bp band after digestion with BamH-I restriction enzymes. The presence of the episomal form resulted only in one control sample and two neoplastic one. The mixed and episomal forms were also evaluated by identifying the distribution of E2/E6 ratio values in our samples. Using real time PCR Sybr Green the virus resulted totally integrated (8 out of 25) only 77

80 in cervical cancer lesions. The mixed status was the one mainly present in dysplastic (4/6) and cervical cancer (14/25), and in only one control sample. The episomal form was found mainly in controls (6/7), with an E2/E6 ratio > 2, in 2/6 dysplastic lesions and 3/25 cervical cancer. Experimental results clearly indicate viral integration occurs with no preferential involvement of the E2 region [143], confirming that the presence of neither an E2 partial sequence, nor a full length E2 region is allowed to infer the presence of episomal forms. There are different methods reported in the literature to identify the physical status of the virus based on E2/E6 ratio [ ]. Even though they are highly sensitive, different E2/E6 ratioswere found. Standardised procedures are needed to evaluate and determine cut-off values that can be used to discriminate smears from normal, dysplastic and cervical cancer lesions. Available methods used for determination of the viral physical status can be divided into two broad categories: (i) those that detect virus host fusion transcripts, i.e. transcriptionally active viral integrants and (ii) those that detect integrated viral DNA regardless of its transcriptional status. Although the frequency of integrated HR-HPV DNA in SILs of both grades is high, only 15% of HSILs and 0% of LSILs contain transcriptionally active viral integrants. It is important to establish the HR-HPV integration in the timeline of neoplastic progression and its functional contribution to carcinogenesis. An indication of lesions at increased risk of progression would be provided by a test able to detectintegrated HR-HPV, ideally using a method that distinguishes between latent and fully selected transcriptionally active integrants to indicate level of risk [144]. Further refinement of themethods to assess the physical status of HR- HPV in clinical samples are urgently needed. 78

81 3.5.2 Stress markers Remarkable differences were shown in stress response markers among the three groups of lesions. The ERp57 and the GST were found to be up regulated both in dysplastic and, to an even greater extent, in neoplastic tissue. The ERp57 is a ER chaperone, resident member of the proteindisulphide isomerase family, namely thiol oxidoreductases. It is localized in ER lumen, nucleus, and cytosol. ERp57 assists in the maturation and transport of unfolded secretory proteins by catalysing the thiol-disulphide exchange, thus facilitating disulphide bond formation and rearrangement reactions. Two major roles of ERp57 are: (i) a key molecular player in the quality control of newly synthesized glycoproteins and (ii) a required component of the peptide loading complex of MHC class I molecules by allowing the correct folding of the HLA class I-β 2 microglobulin (β 2 m) complex, maintaining the peptide-mhc I complex in a steady state [145]. ERp57 expression is induced by a variety of stress conditions, including neoplastic transformation [146]. ER chaperones serve as a novel class of prosurvival components protecting the host against death induced by ER stress when expressed at high levels. ERp57 expression is highly upregulated in varieties of cancer cell lines and human cancer specimens, including breast cancer, lung cancer, liver cancer and prostate cancer. This over-expression is induced by elevated ER stress due to severe hypoxic tumor environment and is thought to be correlated with metastasis and drug resistance [147]. ERp57 protects cancer cells against ER stress-induced apoptosis through multiple mechanisms: by activating UPR signaling to ameliorate misfolded protein aggregation in the ER, binding Ca 2+ preventing its efflux from the ER to the cytosol and inhibiting the activation of pro-apoptotic components, such as 79

82 BIK and BAX, as well as suppressing the cleavage of procaspase-7 and procaspase-12 through complex formation [148]. Accordingly, its expression was increased in transformed cells [149] and it has been also reported that the degree of this induction was proportional to the transforming abilities of the oncogenes [150]. Owing to the ability to control cellular redox activities through its reductase activity, enhancement of its expression could lead to activated redox potential and to the modulation of cancer relevant regulatory actors. Additionally, the role of ERp57 in themodulation of STAT3 signaling and proliferation could be exploited, and the contribute to a cancerous disease stated [151, 152]. GST expression was found induced in dysplastic tissues and even more in neoplastic tissues compared with controls. GST plays a central role in the protection from OS by-products [153]. The mu and theta classes of GST isozymes (GSTM1 and GSTT1, respectively) have common and broad range of substrate specificities and detoxify the active metabolites of benzo-apyrene and other polycyclic aromatic hydrocarbons. GST polymorphisms, especially (null) genotypes of GST that are due to a homozygous deletion of the GSTM1 or GSTT1 genes were frequently observed in lung and bladder carcinoma patients and some data suggest that this could be associated with cervical carcinoma risk [36]. The absence of this detoxificating enzyme in the nucleus of the majority of cervical carcinomas may indicate that xenobiotic compounds are not catabolized and may therefore exert their mutagenic activity, resulting in tumor progression [154]. Other authors have reported high level of GST in most of human cancers and in cervical cancer [155]. Our data also show an over-expression of GST in cervical cancer, which may represents an adaptive response mechanism, devoted to the detoxification of oxidative/nitrosative stress-related to harmful metabolites. 80

83 The level of TrxR2, a seleno-cysteine mithochondrial thioredoxin reductase protein, respect to control tissue was mostly elevated in dysplastic lesions and slightly elevated in neoplastic tissues. Two Trx systems have been described previously in organisms from yeast to mammals. One is present in the cytosol and is composed of TR1 (also called TXNRD1 or TrxR1) and Trx1. The second is a mitochondrial system that consists of TR3 (also called TXNRD2 or TrxR2) and Trx2 [156]. Together with the cytosolic isoform TrxR1 and their cognate thioredoxins (Trx1 and Trx2) constitute the thioredoxin system, the major disulfide reductase and oxido-reductase system in the cell. The thioredoxin system maintains the reduced intracellular environment, is involved in the modulation of cell growth and cell death mechanisms, in OS response, in repair of oxidatively modified proteins and acts as an electron donor in deoxyribonucleotide and DNA synthesis in tumour tissues. Moreover, it is involved in the regulation of redox-sensitive transcription factors including: DNA binding (AP1), nuclear translocation, and protein stabilization (NF-kB, p53) [157]. This system has been linked to many major physiological and biochemical functions and has become an important target for cancer therapy [114]. TrxR are known to be upregulated in a variety of human cancers, including lung, colorectal, cervical, hepaticand pancreatic and Trx overexpression has been linked to aggressive tumor growth and poorer prognosis [158]. TRx system is involved in cancer development by interfering in different mechanisms: 1. Promoting cell proliferation, either through modulation of transcription factor or protein kinase signaling cascades; 2. Insensitivity to growth-inhibitory (antigrowth) signals, including sustained progressions through the cell cycle and propagation of tumor development, thus evasion of growth inhibitory signals; 3. Evasion of 81

84 programmed cell death (apoptosis) by the inhibition of ASK-1 [159], controlling transcription factors that belong to the ubiquitination pathway [160] and increasing HIF-1a (hypoxia-inducible factor 1) [161]; 4. Limitless replicative potential, through telomerase function control; 5. Sustained VEGF function and angiogenesis [115]; 6. Tissue invasion: Trx may reduce disulfides in extracellular matrix proteins such as laminin and may stimulate both matrix metalloproteinase activity and tumor cell invasiveness [162]. In addition to its possible implication in many aspects of cancer biology [163] its specific role in protein oxidative damage repair indicate that TrxR2 induction represents an adaptive response against a condition of increased ROS generation as it occurs during cancer growth [164]. Therefore, despite providing a potential survival advantage to cancer cells, highly upregulated TrxR1/Trx pathway activity, found in HSIL dysplastic lesions, may be required for survival in light of increased ROS stress level found, helping this lesion to further progress to cancer. Compared with the level found in control tissues, inos expression was increasingly reduced in dysplastic and neoplastic lesions. There are three isoforms of NOS: inducible (inos), endothelial (enos), and neural (nnos). Inducible NOS is expressed in macrophages, neutrophils, endothelial cells, hepatocytes, and many other cell types upon stimulation by bacterial endotoxins (lipopolysaccharide, LPS) and inflammatory cytokines, such as interleukin-1b, c-interferon, and tumor necrosis factor-a, and hypoxia [165]. Recent studies have investigated the expression and the activity of inos in human cancer. An increased level of inos expression and/or activity has been found in the tumor cells of gynecological malignancies, in the stroma of breast cancer, in the tumor cells of head and neck cancer, gastric cancer, esophagus, lung and colon cancer. It is likely that several molecular 82

85 mechanisms are involved. It has been demonstrated that NO induces tumor angiogenesis through activation of vascular endothelial growth factor (VEGF) [166] and activation of the cyclooxygenase (COX) enzyme pathway [167]. Another mechanism is that of generating reactive nitrogen species (RNS), such as NOX and peroxynitrite (ONOO-) that mediate the formation of mutagenic DNA species like 8-nitroguanine (8-NitroG) [168, 169]. inos is implicated in both apoptotic and necrotic cell death by inactivation of caspase-3 throughits S-nitrosylation[170]. Also cooperating with the Src family of kinases, activating multiple signaling pathways controlling cellular growth, proliferation, apoptosis, cancer progression, and inflammation [171]. Although this inducible form of NOS has been commonly associated with malignant diseases, its role in carcinogenesis and tumour biology is far from being clarified. Accordingly, a double-edged profile with both tumour promoter and tumour suppressor activity has been described for inos with its ultimate effects being dependent on enzyme level and the cell metabolic context [172]. Low concentration of NO promotes tumor growth and angiogenesis whereas at high concentration NO has antitumor activity by inducing cytotoxicity and apoptosis [173]. This explains the fact that, at the early stages of cervical carcinogenesis, HPV-infected cells display high inos activity, very likely as a defense mechanism against virus infection, whereas, as long as the carcinogenetic process takes place, its activity is reduced in order to promote tumor growth and enhance tumor cell aggressiveness. Studies has found a positive correlation between tumor neovascularity and macrophage counts, whereas inos expression displayed an inverse relationship with macrophage density and increasing severity of CIN lesions [174]. 83

86 Within this frame, low levels of inos in dysplastic and neoplastic samples are consistent with reduced generation of cell inhibiting mediators and metabolites, while growth rate and angiogenesis are probably sustained by other unrelated mechanisms. Taken together the above data support the view that an increased generation of oxidants is associated with the dysplastic/neoplastic phenotype. Highly active detoxifying systems (ERp57; TrxR2; GST) and reduced RNS generation (inos) might be part of a complex adaptive metabolic profile allowing cell survival in an increasingly oxidant, cancer promoting environment Redox proteomics In order to better understand the role of OS in cervical cancer, we measured the extent of total protein oxidation. We found that protein carbonyls were significantly increased in dysplastic tissues, while levels detected in neoplastic tissues were surprisingly similar to control ones. This finding could suggest that oxidative damage to proteins is restricted to the putative pre-cancerous cells, as represented by dysplasia, while tumour tissues are able to counteract OS through the activation of defence systems, according to the above reported response of stress markers. Since it has been demonstrated that protein oxidation results in diminished, complete loss of, or a toxic gain in protein function [175], the identification of specific proteins, which are irreversibly modified by carbonylation, could shed light on the molecular mechanism involved in cell transformation and tumor development. 84

87 In dysplastic tissue compared with control tissues 5 proteins were identified with increased carbonylation levels, namely keratin tipe II cytoskeletal 6, actin, cornulin, retinal dehydrogenase (RDH) and GAPDH. Conversely, five proteins were less oxidized in cancer tissues compared with dysplastic tissues (serpin B3, annexin 2, PDI A3, peptidyl-prolyl- cis, trans isomerase A and GAPDH. Cytoskeleton alteration: CKs and actin Cytokeratins (CKs) contribute to cytoskeleton organization and are well known markers of cell differentiation both in normal and neoplastic epithelia [176]. Their interaction with HPV transforming activity has been reported [177] and their quantitative detection has been proposed as an adjunct to current histological evaluation [178]. Suggestions have been made by which the progressive or regressive nature of CIN may be predicted on the basis of keratin expression patterns [179]. Recently, Arnouk et al. reported a decreased expression of CK6A and other CKs in HSIL and cervical cancer [180]. Our results showed that the isoforms A, B and C of CK6, were more oxidized in dysplastic tissue with respect to controls. This finding indicates that, in addition to the cited altered pattern of CKs expression, a further level of cytoskeleton derangement seems to occur in dysplastic lesion in consequence of the oxidative alteration of CKs. Such a deregulation is potentially reinforced by the parallel increased oxidation of Actin here reported, a protein cooperating with CKs in structural and functional cell shaping [181] that is, all the same, specifically down regulated by the expression of HPV 16 E7 oncogene [182]. 85

88 Epidermal differentiation: Cornulin and Retinal Dehydrogenase Cornulin is a calcium binding protein, member of the fused-gene family [183]. Although its physiological role is presently unknown, it is normally expressed in late epidermal differentiation and is currently used as a marker of differentiation [184]. Recently, cornulin has been reported to be severely reduced in cervical dysplastic lesions and to be almost completely suppressed in neoplastic lesions [180]. It is further showed that decreased cornulin expression accurately predicted local relapse in surgically treated head and neck cancer [185]. Similar down-regulation of cornulin in nasopharyngeal, esophageal and cervical cancers may reflect their common origin in squamous epithelium [186]. Our results show that in dysplastic lesion the residual amount of this differentiation-related protein is likely to have a further reduced/aberrant activity because of its increased level of oxidation. This result lends support the hypothesis, proposed by Arnouk et al (2009) that cornulin may represent a promising candidate for a molecular marker that correlate with the neoplastic progression. Retinal Dehydrogenase (RDH) is an enzyme involved in the synthesis of retinoic acid, a fundamental regulator of cell differentiation, embryogenesis, tissue homeostasis and renewal [187]. A huge amount of clinical and experimental researches underscore the high relevance of retinoids in many biological and clinical aspects of cancer (for a comprehensive review see [188] and data have been reported about their implications in several aspects of HPV related neoplastic diseases including suppressive control on integrated E6 and E7 oncogenes [189, 190]. Studies suggest that changes in the expression level and activity of RDH could perturb retinoid homeostasis and alter the intracellular concentration of 86

89 retinoic acid, leading to abnormal differentiation and high susceptibility to HPV in the cervical epithelium [191]. The oxidation levels of RDH suggest that this fundamental pathway may have altered/reduced activity in dysplastic lesion with possible decreased control on cell differentiation maintenance as well as on pro-survival, pro-mitotic and anti-apoptotic activity of integrated viral oncogenes. Glyceraldeyde-3-phosphate dehydrogenase (GAPDH) has long been considered a simple glycolitic enzyme and has been widely (and erroneously) used as an internal standard reference for RNA expression. Indeed GAPDH is tightly regulated at both transcriptional and posttranslational level [192, 193]. In addition to its conventional metabolic role many other functions have been identified so far implying the participation of GAPDH to many cell function. These include cell adhesion and motility, endocytosis and nuclear membrane assembly, ER and Golgi trafficking, trna export, DNA repair, telomere protection, cell death among others [194,195]. Decreased oxidation of GAPDH may once again represent a prosurvival mechanism which makes tumors/adapted cell more resistant to stress stimuli and therefore able to proliferate out of normal control [195] Taken together the above data indicate that in dysplastic lesions a selective oxidation of specific proteins takes place. This oxidation is likely to alter/decrease their functions contributing to cytoskeleton derangement, suppression of terminal differentiation and reduced control on viral oncogenes activity, conditions known to promote cell malignant progression. Comparing the redox proteomic pattern of dysplastic and neoplastic tissues the following proteins exhibited a lower oxidation in neoplastic tissues: ERp57; Annexin 2; Serpin B3 (serine protease inhibitor); Pin1 and GAPDH. 87

90 Interestingly, the increased expression of ERp57 in neoplastic tissues is also associated with its reduced oxidation, possibly resulting in a much greater increase of activity than expected based on purely quantitative data. This result underscores the relevance that elevated protein folding/unfolding activity may have for cell survival in the pro-oxidant neoplastic environment and suggests that the achievement of the neoplastic phenotype is accompanied by the activation of compensatory mechanisms able to counteract oxidative damage of selected targets and allowing the cell to fit to the hostile environment. A similar lower level of carbonylation in tumour tissues compared with dysplastic lesions were also showed for Annexin 2 (Anx2). It is a member of the Annexins family, a group of Ca 2+ - and membrane-binding proteins that may play a role in regulation of cellular growth, in signal transduction pathways and membrane trafficking events [196,197]. Recent studies suggest that Anx2 might be linked to carcinogenesis through its implication in the invasion and neovascularisation processes [198] and that the protein is regulated by the cellular redox status [199]. Down-regulated Annexin A2 expression was found in highly metastatic lung cancer cells. It was suggested that Annexin A2 may play roles in p53 induced apoptosis and it is also involved in regulation of cell proliferation. We have previously reported that Anx2 is oxidized in normal human epithelial keratinocytes (NHEK) and in HPV-transformed keratinocytes upon exposure to oxidative stress [89,200], as it is susceptible to oxidative modification. However, tumour cells once adapted to a more oxidized environment have improved antioxidant/detoxifying mechanisms and therefore are able to survive under stress. Thus, neoplastic cells are paradoxically able to protect themselves from oxidative damage. 88

91 Serpin B 3, also known as Squamous Cell Carcinoma Antigen 1, is a serine protease inhibitor involved in regulation of plasminogen activation, inhibition of inflammation and promotion of epithelial proliferation [201]. It is found to be over-expressed in cancer cells of epithelium origin and is thought to promote cancer cell survival by inhibiting TNFα induced apoptosis contrasting cytochrome c release and by reducing the pro apoptotic effect of p38 MAPK [202]. It is thought even that activation of STAT3, which occupies the promoter of serpin B 3, is required for continuous activation of B 3 genes [203]. It binds directly to the HPV16-E7 antigen and is down regulated in E7 transfected cells and in HSIL lesions [180; 181]. Our results indicate that a reduced oxidation takes place in tumour tissue, which may reconstitute serpin function. Pin1 [named after the acronym: Protein Interacting with NIMA (=Never In A Mithosis)] is indeed a Peptidyl prolyl cis-trans isomerase that isomerizes phospho-serine/threonine-proline [p(s/t)-p] motifs causing them to twist between two completely distinct conformations. This conformational change affects profoundly protein activity and is a major mechanism of cellular signaling and regulation. Pin1 is a necessary enzyme for cell division, regulates the cell cycle and, once over-expressed, can promote oncogenesis through a number of signaling pathways [204]. Pin1 regulates other pro- or anti-apoptotic factors, such as tumor suppressor p53, c-jun, and Bcl-2, all of which are phosphorylated by JNK. Thus, Pin1 contributes to the regulation of apoptosis by targeting both JNK1 and its downstream pro- and anti-apoptotic factors [205]. The fact that Pin1 is less oxidized in tumor tissue compared with dysplasia might be an indirect indication that tumor cells, supposed to live in a highly 89

92 oxidant environment, are able to protect cellular components from oxidative insult by selective activation of specific protecting mechanisms. As regards GAPDH, we found decreased oxidation in neoplastic tissues vs dysplasia, while in dysplastic lesions vs control tissues an increase of oxidation was evident. This is the only protein which was significantly modulated in all the 3 groups: this result led us to hypothesize that increased oxidation/dysfunction occurs in dysplasia which represents a condition with increased risk of developing cancer but not in neoplastic cells which protect themselves from oxidative damage and proliferate like normal, non damaged cells. Decreased oxidation of GAPDH may once again represent a pro-survival mechanism which makes tumors/adapted cell more resistant to stress stimuli and therefore able to proliferate out of normal control [195]. 3.6 Conclusions In order to provide data for the identification of molecular marker(s) that correlate with the outcome of dysplastic lesions progression, we analyzed the expression of stress response proteins and the pattern of oxidative adducts on the proteome in a set of normal, dysplastic and neoplastic cervical tissues infected with the HR-HPV 16. An increasingly marked activation of stress response proteins in dysplastic and in neoplastic tissues was shown, underscoring the relevance of efficient detoxifying mechanisms for tumor cells survival. Our results depict a paradoxal scenario where dysplastic tissue represents the most deregulated entity while tumor tissue appears to be capable of controlling high ROS levels and therefore able to proliferate. 90

93 Specific oxidation of target protein, including keratins, Actin, cornulin and retinal dehydrogenase leads to their impaired function in dysplasia. ERp57, Anx2, Serpin B3, Pin1 and GAPDH appear to be specifically protected from oxidative modification in cancer tissues. The above described proteome profile further demonstrate that cancer cells have an increased ability to counteract oxidative damage although in a condition of increased rates of ROS production. High levels of oxidative stress are responsible of oncogenic mutations, however cancer cells upregulate detoxifying systems and may accumulate ROS-mediated mutagenic events, which ultimately promote tumorigenesis. Further studies are needed to better understand the effects of protein oxidation on pathways involved in cell transformation and cancer promotion. The comprehension of this phenomenon may also support current clinical protocols for screening and prognostic evaluation of cervical lesions. 91

94 PROJECT III The effect of HPV16 E5 variants in transfected keratinocytes: A proteomic and redox proteomic analysis 4.1 Introduction It is well established that HR-HPV E6 and E7 proteins are responsible for full malignant transformation both in vitro and in vivo. Another protein, namely E5, is has been recently recognised to contribute to initial stages of carcinogenesis. While E6 and E7 are expressed throughout the course of the disease and are necessary for the maintenance of a transformed phenotype, E5 is expressed during the early stages of infection and suppressed, although not always, as the lesion progresses toward malignancy. These characteristics point to a role of E5 in establishment of PV infection and the initiation of cell transformation. Studies on E5 protein are mainly based on HPV 16 or BPV (bovine papillomavirus). E5 is an hydrophobic membrane protein, 83 amino acids long, located mainly in the endomembranes of the infected cells like at the endosomal compartment, endoplasmatic reticulum, Golgi apparatus and, to a lesser extent, at plasma membranes. It is believed to have 3 anchor-like α- helices transmembrane domains, and short regions at the N and C terminal that may extend beyond the lipid bilayer. The N terminus of 16E5 is located in the ER lumen and the C terminus in the cytoplasm, the latter being responsible for different cell target interaction [206]. 92

95 4.2 HPV16 E5 protein functions HPV16 E5 protein is reported to have several biological functions playing an important role in different cell s mechanisms as described below Signal trafficking pathways It has been reported that E5 binds to the 16kDa pore-forming c subunit, a component of vacuolar H + -ATPase, an universal proton pump of eukaryocytes, which is responsible for acidifying membrane bound organelles. E5 protein prevent V-ATPase assembling and its transferin the endosomes [207], consequently leading to endosomes alkalinisation, which is thought to be compartment specific and the E5 carboxyl terminus is required for this activity. Alkalinisation of the endosomes inhibits their fusion, the trafficking of molecules into the endosomes and the direction of the latter in cell compartments alteratingthe endocytic cellular pathway [208]. Another activity of E5 protein is its ability to enhance epidermal growth factor signalling pathway. It is thought that it can be due to endosome alkalinisation that alters growth factor receptor downregulation, receptor recycling and receptor/ligand interaction [209]. Also it is possible that E5 blocks the interaction of c-cbl with the EGFR, thereby stabilizing the activated form of EGFR, decreasing its downregulation and prolonging its downstream signal [210]. The ability of E5 to interfere with EGF processing, to augment EGFR, to alter the transport of HLA proteins, cholesterol, gangliosides, and lipid raft proteins that function in controlling signal transduction pathways, may be explained with the inhibition of endosome maturation and EGF endocytic 93

96 trafficking via a ph-independent, transport-independent mechanism, most likely by altering vesicle fusion events [211] Immune response A mechanism by which E5 protein contributes to the viral persistence during the early stages of the infection, is escaping the immune system control as described below. It is reported that E5 downregulates, in vitro, selectively HLA-A and HLA-B class 1 (not HLA-C or E) molecules on the cell surface. This downregulation is thought to be mainly due to inhibition of their transport on cell surface without inhibiting their transcription [212]. In fact E5 interacts with calnexin interfering with calnexin-assisted folding of HLA class I molecules and consequently resulting in heavy chains retain in ER [213, 214]. This way the infected cell evade from cytotoxic lymphocyte T immune response by lack of viral antigens presentation on cell surface and by decreasing γinf cell levels [215]. It has been found that E5 decreases also the MHC class II surface expression by inhibiting MHC complex maturation [216]. Another mechanism by which the E5 protein can alter the host immune response is by decreasing the cell surface CD1d level, which is a major histocompatibility complex (MHC) class I-like glycoprotein that presents self or microbial lipid antigen to natural killer T (NKT) cells. Modification of CD1d was interrupted at the level of the ER by interactions between HPV E5 and calnexin. Improper folding, and/or ubiquitination of CD1d MHC, in the presence of HPV targets CD1d to cellular proteasomal degradation [217]. 94

97 4.2.3 Cell proliferation E5 protein was reported to transform tissue-cultured murine fibroblasts and keratinocytes and to enhance the immortalization potential of E6 and E7 proteins [218,219]. The expression of HPV16 E5 allows growth of human keratinocytescell line in vitro and the development of endophytic papillomas, in transgenic mice. The expression of E5 allows in vitro growth of human keratinocytes cell line and contributes to the promotion and progression stages of carcinogenesis [220,221]. E5 also enhances cell proliferation through down-regulation of tumor suppressor p21, a cyclin-dependent protein kinase inhibitor (CKI), enhancing cell cycle progression. E5 was reported to suppress p21 Wafl/Sdil/Cipl tumor suppressor gene at the transcriptional level in immortalized humankeratinocytes [222]. The reduction of p21 promoter activity was correlated with transforming activity and enhancement of mitosis, which was manifested by increased expression of c-jun [223] Tumour growth E5 protein has found to be involved in tumour growth affecting different mechanisms such as: cell proliferation, angiogenesis, apoptosis etc. Angiogenesis is an essential step in tumour growth and metastasis. Vascular endothelial growth factor (VEGF), a major mediator of tumour angiogenesis, promotes mobilization of endothelial progenitor cells, cell proliferation, migration, survival and vascular permeability. E5 enhances angiogenesis by inducing VEGF expression through EGFR signaling and COX-2 PGE 2 pathway [224]. 95

98 4.2.5 Response to OS Apoptosis is a physiological cellular response to environmental stress caused by, for example, virus infection, genotoxic agents such as chemotherapeutic drugs and ionizing radiation. E5 protein has been shown to inhibit apoptosis in vitro at different key points in the pathway. The E5 protein was reported to prevent Fas ligandrelated (FasL) or tumor necrosis factor-related apoptosis inducing ligand (TRAIL)-mediated apoptosis of HaCaT cells and to protect human keratinocytes from ultraviolet radiation-induced apoptosisby enhancing the phosphatidylinositol 3-kinase-Akt and extracellular signal-regulated kinase1/2 mitogen-activated protein kinase (MAPK) signal pathways [ ]. E5 protein inhibits hydrogen peroxide-induced apoptosis partly by stimulating the ubiquitin proteasome-dependent degradation of pro-apoptotic Bax protein through mechanism that sequentially involves COX-2, PGE2, EP2, EP4 and PKA [228,229]. Another study show a very specific and consistent downregulation of ER stress response genes (XBP-1, COX-2, and IRE1) by E5, suggesting a its potential role in repressing the cellular ER stress response following HPV infection (fig.24) [230]. 96

99 Figure 24: HPV16E5 protein functions and cell pathways involved Recent studies has demonstrated that E5 alters the differentiation programme of keratinocytes, leading to both hyperkeratinization and defective differentiation in the upper layers, disruption of the basal layer and invasion [231]. The ORF region of HPV 16E5 gene has different restriction enzymes point. At protein level a total of eight variants have been reported deriving from the presence/absence of restriction site for SspI (position ), NspI (position ) and XcmI (position ). Different prevalence of these variants is found in association with neoplatic and dysplastic lesions, raising the possibility that variation of the E5 protein might explain these associations directly [232]. 97

100 Indeed, HPV-16 E5 plays an integral role in the early stages of neoplastic transformation, with E5 mrna and protein detected in low-grade cervical intraepithelial neoplasias (CIN). These variants are reported to be associated with different oncogenic activity such as: enhancement of mitosis, of cell cell-cycle progression and cell-growth [233]. Knowing these variants biological properties may be useful to better understand the prevalence differences found in disease stages and to identify new molecular factors and mechanisms involved in cervical cancer pathogenesis. 4.3 Aim of the work The aim of this project is first: to contribute knowledge about the circulation of natural E5 variants among Italian and Albanian patients, and second to describe their differing biological properties. 4.4 Material and methods Clinical samples For this study 187 cervical brush smears and 13 cervical cancer biopsies, were collected from 200 women, attending either the Gynecological Department of Regina Elena Cancer Institute of Rome (Italy) or Gynaecological Out-Patients Clinic of the Oncology Department at the Nene Teresa University Hospital Centre of Tirana and from women attending the Family Practice at the Gynaecological Clinic of Duress (Albania). Permission for the collection of clinical specimens, was provided by local Ethical Committees. 98

101 4.4.2 HPV type characterization Cells from the cervical brush smears were re-suspended in 1 ml aliquots in PBS for DNA extraction and stored at - 20 C. A 1 ml sample in PBS was then centrifuged at g and the cell pellet was resuspended in 200 µl proteinase K (PK) solution and digested overnight at 55 C. The DNA from cervical brush smears and cervical tissue biopsies were extracted with the QIAamp DNA Mini Kit (QIAGEN Gmbh, Hilden, Germany) used according to the manufacturer s instructions. The DNA extracted was analysed for HPV detection by amplification using the MY09/MY11 primer couple [59]. The HPV-16 positive Siha, CaSki and the HPV 16 in vitro transformed human keratinocytes HK- 168 cell lines [90] were used as positive controls while HaCaT cells [125] were used as negative controls. Amplicons were analysed by 2% agarose gels electrophoresis containing ethidium bromide. HPV positive samples were typed by the BigDye Terminator 1.1 Cycle Sequencing Kit (Sanger method), as described above. Samples adequacy to PCR analysis was assessed by a parallel β-globin gene amplification with primers GH-20/PC-04 (tab. 1) as described [62] HPV 16 E5 variants analysis Allsamples resulted infected by HPV16 were investigated for the presence of E5 gene, amplifying the 237 bp of the whole E5 gene length (tab.4). Amplification consisted in an initial step at 95 C for 3 min for Polymerase activation, followed by 35 cycles of denaturation at 95ºC for 30 sec, annealing at 52ºC for 30 sec and of extension at 72ºC for 40 sec plus a final cycle with a 10 min long extension for optimal chain termination. Amplified products were separated by 2.5% ethidium bromide stained 99

102 agarose gel electrophoresis and visualized by gel direct inspection under UV- B trans-illumination. Amplicons from PCRs were purified by QIAGEN PCR Kit and DNA pellets were resuspended in 50 µl molecular biology-grade distilled water. Four 5 µl aliquots were prepared, three of which were subjected to an hour digestion at37 C with 5 U of one RE (SspI, XcmI or NspI ; New England Biolabs andboehringer-mannheim) in appropriate buffer with subsequent enzyme inactivation at 65 C for 10 min.the digested mix was analyzed in3 % agarose gels. Variant types detected by RFLP (restriction fragment length polymorphism) analysis are assigned according to Bible et al. [232] Variants cloning Three different isolated variants were cloned. They were reamplified with primers including at 5 end an artificial BamHI site and a Kozak sequence (ACCATGG) for optimal translational efficiency. The second PCR used a 5' primer that contained a TAA 'stop' codon (at codon position three) as an internal control. Both PCRs utilized the same downstream primer with a tail including an artificial EcoR1 site. These primers amplify a 297 bp sequence (tab 4). The amplification protocol was the same as already told above, but the annealing step was perform at 46 C. The amplicons ran in 2% etidium bromide agarose stained gel, were purified with Qiagen Kit as recommended by the constructer. 100

103 PRIMERS SEQUENCES Bp E5_3866_S TGCATCCACAACATTACTGGCG 161 E5_4027_AS AACACCTAAACGCAGAGGATGC 161 NHE GGAGCTAGCTCACCATGGCAAATCTTGATA 274 BAM TACAGGATCCTTATGTAATTAAAAAGCGTGTAA 274 STOP GGAGCTAGCTCACCATGGCATAACTTGATA 274 CLAE5_S_16 GGCCGGATCCACCATGTACCCATACGACGTCCCAGACTACG CTTACTGCATCCACAACATT 297 CLAE5_AS_16 TACAAAGCTTAGAATTCTTATGTAATTAAAAAGCGTGCATG 297 CLAE5_S_STOP_16 GGCCGGATCCACCATGTACTAATACGACGTCCCAGACTACG CTTACTGCATCCACAACATT E5_3863_S TACGCTTACTGCATCCACAACATT 16E5_4100_AS TTATGTAATTAAAAAGCGTGCATG 237 Table 4: Primer sequences DNA sequencing Purified E5 inserts were cloned into the TOPO-TA plasmid according to the manufacture instructions. Plasmids were transformed into Top10 competent cells with the followed protocol: 2 µl of cloning mixture were incubated with 100 µl of Top10 cells, in ice for 30 min. The mix where submitted to a termic shock at 42 C for 30 sec and then incubated in ice. Top10 were incubated with LB medium for 1 hour at 37 C. Afterward grown overnight at 37 C on agar plates with ampicillin. The positive colonies where grown in 500 ml of LB medium overnight at 37 C and the plasmid DNa was extracted with the followed protocol: the cultured was pelletted at 2500 rpm for 30 min, washed in 100 ml TNE, and incubated with lysis buffer for 5 min at room temperature. The lysate where incubated in ice with potassium acetate and acetic acid for 10 min and centrifuged at rpm for 20 min. The plasmid DNA in the supernatant 101

104 was precipitated with 0.6 volumes of isopropanol and the pellet washed with ethanol at 70%. The pellet was re-suspended in TE and adding CsCl 1g/ml plus ethidium bromide were ultracentifuged at rpm for 18 hours. The ring was extracted, washed in butanol:water 1:1 and the DNA precipitated overnight at 20 C with sodium acetate 3M and ethanol. The extracted DNA was identified by plasmid sequencing using M13 forwardand reverse primers, and by RFLP. Identities of the cloned sequences were checked by plasmid based sequence and validated inserts used to be cloned into the 2 nd generation lentiviral experimental vector Ligation The E5 variants PCR products were ligated into the BamHI and EcoR1 sites of plico virus DNA by their cohesive ends using the followed protocol: a mix of 20 µl containing a molar ratio of plasmid DNA/ insert DNA of 1:3, 1U of T4 DNA ligase and 1x of the appropriate buffer, were incubated at 16 C for 16 hours. The mix were diluted 5 times with TE ph 7.5 and 2µl were used to be transformed into Top 10 competent cells (as explained above) and grown overnight at 37 C on agar plates with kanamicin. Plico plasmid without insert was used as a control. Positive colonies were selected and grown in LB with kanamicin overnight at 37 C. Plasmid DNA was purified as described above and used to transfect the keratinocytes Lentiviral vector packaging Human kidney cell line, named 293T, was used for lentivirus formation, due to their known high transfection efficiency. Two vectors were 102

105 used for lentiviral formation: pmd 2 G, an expression vector encoding for the G protein of VSV for its high infectivity and pax 2 plasmid that encodes for the gag-env proteins of HIV-1. These vectors ideally suited for the 2 nd generation lentivirus packaging. Cells must be approximately 1/4 to 1/3 confluent on the day of transfection. One to 3 million cells were precipitated and mixed with 3µg pmd 2 G (envelope plasmid), 6.5µg pax 2 (packaging plasmid), 20 µg plico-e5 (transfer vector) in a final volume of 250 µl with water (bidistilated water with 250 mm Hepes). The mix was precipitated with 250 µl of CaCl M and added slowly on 500 µl of HeBS2x (NaCl 0.28 M, HEPES 0.05 M, Na 2 HPO4 1.5mM), while vortexing. Let stand still for 30 minutes minimum (40 minutes max) on the bench, for the precipitate to form. Then the precipitate was added slowly, dropwise on the 293T cell monolayer. The sup was harvested 72 h after transfection and centrifuged at 2500 rpm for 10 minutes at +4 C. The supernatant containing the lentivirus was stored at - 80 C until used for cell line infection Cell culture Normal human epithelial keratinocytes (NHEK) were obtained from discarded child foreskin, from four different donors, and cultivated in serumfree KGM medium. Briefly, after scraping away excess fat and subcutaneous tissue, foreskin were floated in 0.25% trypsin solution at 4 C overnight. The epidermis was lifted off, placed in 20 ml of trypsin-edta (GIBCO-BRL) at 37 C for 1 h under continuous mild stirring. Trypsin was neutralized by soybean trypsin inhibitor, the cell suspension was pelleted for 10 min at 200 g, washed twice in phosphate buffered saline (PBS) and plated on plastic flasks at cell/cm 2 density in serum-free KGM medium. To avoid bias 103

106 of senescence modification in cell metabolism, cells between third and eighth passages were used in the present study Keratinocytes infection NHEK cell lines, at 3 th -8 th, passage were used for the infection with the previously formed lentivirus. NHEK cells were harvested and replated at 3 x 10 4 cell/cm 2 into T-75 flask 24 hours before infection. Next morning the cell were incubated with the infection mix, containing 4 µg/ml polybrene, 3 ml lentivirus E5 variant and 3 ml of K-SFM, at 37 C 5% CO 2 for 24 hours. The media was changed with fresh K-SFM and the cells were incubated at 37 C 5% CO 2 for 48 hours. The cells were further scrapped and centrifuged at 8000 rpm for 5 min at +4 C. The pellet was stored at -80 C until used for protein, DNA and RNA extraction. 4.5 Results and discussion Different properties of HPV16 E5 protein are already known, but its oncogenic role in HPV related cervical cancer isn t established yet. Viral proteins are expressed in a suprabasal layer of the epithelium and in a low amount, below the level acquired to switch the immune response, leading somehow to immune tolerance. HPV 16 E5 protein is thought to have a role in the early stages of the infection, favouring the viral persistence in the host cell by escaping from the immune system control. In fact, HPV-16 E5 evades immune surveillance, by a reduced MHC-I expression [234], because of its retention in the Golgi apparatus [235]. 104

107 During viral infection the host cell undergoes apoptosis as a defense mechanism to prevent further spread of the virus in the organism. Blocking of apoptosis by E5 can inhibit the premature death of host cells and enhancing the viral persistence with further neoplastic progression [227]. In vitro studies showed that HaCaT-16E5 transfected cells formed an epithelium with dysplastic/neoplastic characteristics: defective differentiation, distruption of the basal layer with many vacuolated cells and numerous invading pockets [231]. Also findings indicate that E5 induces binucleated cell formation, a characteristic of cervical cancer, which may promote genomic instability in subsequent rounds of cell division [236]. Variants of HPV16 E5 gene and their relation with different stages of cervical cancer disease are reported in the literature [232]. This study however, is done on British patients and did not further evaluate the biological properties of these variants. We studied the variants present in Italian and Albanian population and their prevalence related to disease status using RFLP method Patients A total of 200 samples were investigated for the presence of HPV infection. HPV16 positive samples were further investigated for studying the E5 variants. Detectable HPV 16 E5 DNA was found only in 95 samples out of 113. Patients positive to HPV16 E5 DNA included 46 women with normal cytology to act as controls, 22 women with LSIL, 14 women with HSIL and 13 women with cervical carcinoma. Mean ages and the age range for these women are 49.11±10.45 (tab. 5). 105

108 RFLP variants All subjects HPV16E5 positive n =95 CTR n=46 Diseasestages CN n=49 LSIL n=22 HSIL n=14 SCC n=13 Pattern 1 24 (25.3%) 9 (19.5%) 15 (30.6%) 6 (27.3%) 4 (28.6%) 5 (38.5%) Pattern 2 52 (54.7%) 20 (43.5%) 32 (65.3%) 14 (63.6%) 10 (71.4%) 8 (61.5%) Pattern 5 19 (20.0%) 17 (37.0%) 2 ( 4.1%) 2 ( 9.1%) 0 ( 0 % ) 0 ( 0 % ) Table 5: RFLP variants detected and related to disease status Prevalence of HPV16 E5 variants RFLP assay was used to evaluate the HPV16E5 variants among women with /without neoplasia and dysplasia as suggested by Bible et al, In our study we found only three variants present. Variant 1, where all the three restriction sites (XcmI, NspI, SspI) are present and corresponds to RFLP pattern 1 profile.variants 2, which lacks the SspI restriction site,is digested only with XcmI and NspI restriction enzyme and with corresponds to RFLP pattern 2 profile. Variant 5, that lacks both XcmI and SspI restriction sites, is digested only by NspI and represented by RFPL pattern 5 profile (fig. 25). Figure 25: HPV16 E5 RFLP variants pattern. M: molecular weight marker 106

109 Variant 2 is the one most common found in our samples. It was found in 52 cases out of 95 (54.7%) investigated, while variant 1 in 24 cases out of 95 (25.3%) and variant 5, the less represented, found in 19 of 95 (20%) of samples. Variants 1 and 2 were detected with higher prevalence in neoplastic samples, with 30.6% and 65.3% respect to controls. Variant 5 was highly represented in controls samples with 37% compared to 4.1% in neoplastic patients. The pattern of variants distribution among different stages of the disease present a high prevalence of variant 2 (more than 60% in all stages), followed by variant 1, which was found to be almost homogenously distributed in all disease stages, with prevalence around 30%. Variant 2 was twice more represented in neoplasias than variant 1. Both these variants are also reported to be the most transcriptionally active [233]. Variant 5 was nine times more prevalent in asymptomatic women than in those with dysplasia. We did not find it neither in high dysplastic lesions nor in cervical cancer samples. It was only present in LSIL lesion in very low prevalence (4.1%) (tab.5) Cloning of HPV16 E5 variants in plico plasmid Each variant was cloned in plico plasmid ligating with its restriction enzymes cohesive ends. Additionally a mutated insert with a TAA stop codon at 5 end, which doesn t translate into E5 protein, was cloned and used as an internal control. This plasmid has a WPRE site, that controls and enhances the expression of the cloned gene, resulting in its high transcription efficiency once entering in the infected cell. The plasmids harboring our inserts were purifiedand a PCR of control was done to certify the presence of HPV16 E5 gene. As reported in fig. 26, 107

110 all our variants were successfully cloned in the transfer vector. We found difficulty in cloning the fifth variant, maybe due to its low amplicons level. E5 P1 E5 P2 E5 P5 E5 Stop MVI H 2O HPV bp Figure 26: PCR amplicons of E5 gene variants inserted in plico plasmid. EP1: variant 1 E5 gene, E5P2: variant 2 E5 gene, E5P5: variant 5 e5 gene, E5 Stop: control plasmid, H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVI: Molecular marker VI. Inverted image. These transfer vectors harboring our variants were packaged in a 2 nd generation lentivirus by transfecting the 293T cell lines, as described above. Four different lentivirus were obtained, three harboring our variants and one harboring the control plasmid, which were further used to infect the primary keratinocytes NHEK cell line infection To evaluate the biological properties of these variants, native human keratinocytes cell line (NHEK) was infected in vitro. We used this cell line because this model simulates better the natural infection process. It was difficult to infect this cell line, cause they do not resist many passages and also are very sensitive to E5 protein toxicity. The presence of our insert in the infected NHEK DNA was evaluated amplifying E5 using PCR (fig.27). Based on real time PCR evaluation ~ 20 % of our cells were estimated to be infected. Improvement of lentiviral efficiency, propably by using other 108

111 packaging plasmids, may enhance infection rate. This is a critical point that we are still working on. Figure 27: E5 variant amplicons in infected NHEK cell.variant 1 E5 gene, E5P2: variant 2E5 gene, E5P5: variant 5 e5 gene, E5 Stop: control plasmid, H 2 O: no template control, HPV16: positive HPV16 plasmid control, MVI: Molecular marker VI. Inverted image. 4.6 Work in progress Biological properties of these variants are being evaluated by proteomic analysis of the infected NHEK proteome. Identifying their biological properties may be useful to better understand the prevalence differences found in disease stages and their possible role in cervical cancer pathogenesis. Another interesting approach is the evaluation of the HPV 16 E5 variants role in OS cell response. We have found in vivo high level of OS in CIN lesions causing oxidative modification of different proteins involved in cell differentiation mechanisms. A redox proteomic approach may be helpful in identifying the biological mechanisms involved in HPV16 E5 protein response to OS and the possible differences among its variants. 109