Modulating The CD8+ T Cell Response in Human Papillomavirus-Induced Lesions: Studies in The K14-HPV16 Transgenic Mouse Model

Dissertação de candidatura ao grau de Mestre em Oncologia – especialização em Oncologia Molecular – submetida ao Instituto de Ciências Biomédicas de Abel Salazar da Universidade do Porto. Orientador – Professor Doutor Rui Gil da Costa Categoria – Investigador Afiliação – LEPABE, Faculdade de Engenharia da Universidade do Porto; Grupo de Oncologia Molecular e Patologia Viral, CI-IPOP, Instituto Português de Oncologia do Porto Co-Orientador – Professor Doutor Rui Medeiros Categoria – Professor Convidado Associado Afiliação – Instituto de Ciências Biomédicas de Abel Salazar; Grupo de Oncologia Molecular e Patologia Viral, CI-IPOP, Instituto Português de Oncologia do Porto Co-Orientador – Professor Doutor Manuel Vilanova Categoria – Professor Associado Afiliação – Instituto de Ciências Biomédicas de Abel Salazar CARLOS EDUARDO REIS DOS SANTOS MODULATING THE CD8 + T CELL RESPONSE IN HUMAN PAPILLOMAVIRUS-INDUCED LESIONS: STUDIES IN THE K14HPV16 TRANSGENIC MOUSE MODEL II III Alguns dos resultados incluídos nesta dissertação de mestrado foram submetidos para publicação: - Santos C., Ferreirinha P., Sousa H., Ribeiro J., Bastos M.M.S.M., Faustino-Rocha A.I., Oliveira P.A., Medeiros R., Vilanova M., Gil da Costa R.M. (2015) Kinetics of CD8 + T cells in human papillomavirus-induced lesions – data from K14-HPV16 transgenic mice. Submitted to the journal Tumor Biology with the number TUBI-D-15-04094. - Santos C., Neto T., Ferrerinha P., Sousa H., Ribeiro J., Bastos M.M.S.M., Oliveira P.A., Medeiros R., Vilanova M., Gil da Costa R.M. (2015) Celecoxib promotes degranulation of CD8 + T cells in HPV-induced lesions of transgenic mice. Submitted to the journal Antiviral Research with the number AVR-D-16-00002. Alguns resultados foram também publicados em atas de um congresso internacional: - Santos C., Ferreirinha P., Sousa H., Ribeiro J., Bastos M.M.S.M., Faustino-Rocha A.I., Oliveira P.A., Medeiros R., Vilanova M., Gil da Costa R.M. (2015) Role of CD8 + T cells in HPV-induced skin cancer: data from a K14-HPV16 transgenic mouse model; Proceedings of the HPV2015 – 30 th International Papillomavirus Conference & Clinical and Public Health Workshops; 17 th -21 st September 2015, Lisbon, Portugal. (Best poster presentation award) IV V “When you want to succeed as bad as you want to breathe then you’ll be successful” – Eric Thomas VI VII Agradecimentos Neste momento, que marca o final de mais uma etapa da minha vida, não poderia deixar de agredecer às pessoas que de alguma forma contribuíram para a conclusão desta etapa. Este último ano foi bastante desafiante e exigente e todo o conhecimento e apoio que recebi foi fundamental para chegar até aqui. Em primeiro lugar, gostaria de agradecer à pessoa que me recebeu e que me acompanhou mais de perto no desenvolvimento deste trabalho, o meu orientador, Professor Doutor Rui Gil da Costa. Agradeço a sua confiança em mim para levar a cabo este trabalho, todos os conhecimentos que me transmitiu, a disponibilidade para ajudar no que fosse preciso, todos os conselhos e a constante boa disposição. Ao Professor Doutor Rui Medeiros, meu co-orientador, agradeço toda a ajuda fornecida ao longo deste ano e por tudo o que me ensinou. Mesmo estando cheio de trabalho arranjava maneira de me receber sempre que aparecia no seu gabinete do IPO para me poder ajudar no que eu precisasse e sempre bem-disposto. Ao Professor Doutor Manuel Vilanova, também meu co-orientador, por tudo aquilo que continua a ensinar-me, pelo seu rigor que tanto aprecio, pela disponibilidade que arranja sempre que eu preciso de falar consigo mesmo quando a quantidade de trabalho não o “deixa respirar”. A este conjunto de orientadores agradeço todas as discussões das quais surgiram ideias de trabalho, agradeço tudo o que me ensinaram e, acima de tudo, agradeço a compreensão e o apoio que sempre demonstraram perante a minha vida de estudanteatleta. Não só compreendem como também valorizam o que faço e é realmente bom sentir esse apoio. Ao Pedro Ferreirinha, mais do que um co-orientador no passado, um amigo e colega investigador no presente. Muito do que aprendi sobre trabalhar num laboratório devo-o a ele. Para além de tudo o que me ensinou e continua a ensinar está sempre disposto a ajudar no que for preciso. Não só é dedicado ao trabalho como também é uma pessoa óptima para ir tomar um café e ter uma conversa animada. Às restantes pessoas que trabalham no Laboratório de Imunologia Mário Arala Chaves (ICBAS), em especial à Encarnação Rebelo, Joana Alves e Alexandra Correia, agradeço o carinho e a boa disposição de todos os dias, a disponibilidade para ajudar no que precisasse e as conversas alegres e enriquecedoras. À Professora Paula Oliveira, à Ana Faustino-Rocha e ao Tiago Neto (todos da Universidade de Trás-os-Montes e Alto Douro) agradeço todo o apoio que me deram durante a realização dos trabalhos experimentais. VIII Por último, mas muito importantes, a família e os amigos formam um conjunto de pilares que servem de apoio nos momentos mais complicados e de diversão quando a ocasião o permite. Àqueles que me acompanharam mais de perto durante este ano mesmo, em alguns casos, estando longe (Tiago Silva, Beatriz Silva, Helena Pinheiro, Mariana Pereira, Inês Sousa, Hildeberto Moreira, Ana Rosa, Luciana Leite, Rita Faria, Miguel Gonçalves, Cátia Oliveira, Joana Pinto, Renata Lopes, Diogo Edi, Isabel Paiva, Catarina Pereira e, claro, aos meus colegas de equipa) agradeço pela força que me deram quando as coisas pareciam não correr bem, a motivação que me passaram quando esta me faltava, as conversas enriquecedoras sobre o trabalho, os momentos de descontração, gargalhada e diversão quando estes eram possíveis e, principalmente, por serem verdadeiros amigos e compreenderem os sacrifícios que faço em prol dos meus objectivos e ambições enquanto estudante-atleta. Aos meus pais e irmã, agradeço por tudo aquilo que não é possível agradecer o suficiente. Agradeço pela educação que me deram e por tudo o que me ensinaram, por todas as condições que sempre me deram para o meu crescimento, pelas chamadas de atenção, por exigirem o melhor de mim, por me apoiarem em todas as decisões, por me orientarem. Se hoje posso dizer que tenho orgulho no que me tornei devo-o inteiramente a eles. Obrigado. IX Index Agradecimentos .............................................................................................................. VII Index ................................................................................................................................ IX Abbreviations ................................................................................................................... XI Resumo ......................................................................................................................... XIII Abstract .......................................................................................................................... XV Chapter 1: Introduction ..................................................................................................... 1 The Basis of HPV Infection and Cell Transformation ..................................................... 4 HPV Infection vs. Host Immune Response .................................................................... 5 HPV-associated Chronic Inflammation .......................................................................... 6 HPV-induced Carcinogenesis: The Best Models to Understand It ................................. 9 K14-HPV16 Transgenic Mice as a Model for Inflammation-associated Carcinogenesis12 Modulators of the immune response and their effect over CD8 + T cells ........................15 Chapter 2: Objectives ......................................................................................................17 Chapter 3: Kinetics of CD8 + T cells in human papillomavirus-induced lesions..................19 Introduction ..................................................................................................................21 Material & Methods ......................................................................................................22 Animals .....................................................................................................................22 Mice genotyping........................................................................................................22 Study Design ............................................................................................................22 Histology ...................................................................................................................22 Preparation of Single-Cell Suspensions ....................................................................23 Immunophenotyping .................................................................................................23 Statistical Analysis ....................................................................................................23 Results .........................................................................................................................24 Transgenic Mice Show Epidermal Hyperplasia and Dysplasia ..................................24 Increased CD8 + T lymphocytes numbers and activation in HPV16 +/- mice ................24 Discussion ....................................................................................................................26 Conclusion ...................................................................................................................27 Chapter 4: Ptaquiloside inhibits tumour-infiltrating CD8 + T cells in HPV-transgenic mice .29 Introduction ..................................................................................................................31 Material & Methods ......................................................................................................32 Mice ..........................................................................................................................32 XVI 1 Chapter 1: Introduction 2 3 Human papillomavirus (HPV) is the pathogen responsible for the most common sexually transmitted infection among both men and women in the United States [1]. Most sexually active people will be infected by HPV during their life. The vast majority of HPV infections in immunocompetent individuals are cleared by the immune system, but in approximately 10% of cases [2, 3] the persistence of this infection will cause the development of disease and progression to cancer. HPV has tropism for skin and mucosal epithelia especially from the anogenital and oropharyngeal areas. High-risk HPVs (mostly types 16, 18, 31 and 33) act as crucial etiologic agents in cervical carcinoma [4] – which is the sixth most frequent malignancy in women worldwide [5] – and are the cause of some head and neck neoplasms where incidence depends on the country and other associated risk factors like tobacco [6, 7]. Since 1863 [8], when Virchow first hypothesised a link between inflammation and cancer, many studies have been carried out in order to understand the underlying mechanisms of this relation and its intervenients. The importance of this relation was further reinforced by the award of the Nobel Prize to Robin Warren and Barry Marshall for their work in discovering Helicobacter pylori as the etiologic agent of inflammationassociated gastric cancer [9]. The tumour microenvironment is characterised by the presence of a variety of leucocytes which have been shown to be attracted by chemokines produced by tumour and stromal cells [10]. Lymphocytes are an important part of inflammatory infiltrates in the tumour microenvironment [8, 11] and different lymphocytic populations can play different roles. For instance, T helper (Th) 1, Tγδ and cytotoxic T cells are associated with disease-free survival. In contrary, Th2, Th17 and T regulatory (Treg) cells are allied with a poorer outcome [12]. Neutrophils, tumourassociated macrophages (TAMs), dendritic cells (DCs) and mast cells represent other leucocyte populations also frequently found in the tumour environment that may play dual roles during carcinogenesis [8, 13]. These tumour-associated inflammatory cells, together with the tumour cells, produce and secrete a wide set of cytokines (e.g.: tumour necrosis factor (TNF)-α, interleukin (IL)-1, IL-6 and monocyte colony-stimulating factor) and chemokines (e.g.: monocyte chemotactic protein-1) that may either help tumour progression or contain it [14-17]. In the case of HPV-associated carcinogenesis, the virus infects keratinocytes and an immunological response arises as cytokines (mostly TNF-α, IL-1 and type I and II interferons – IFNs) are released by the infected keratinocytes and local immune cells [18]. However, HPVs may avoid the effector functions of pro-inflammatory cytokines by different mechanisms [19-22]. Thus, if the infection persists, the chance of neoplastic transformation rises [23]. Besides its importance in the early stages of carcinogenesis, 4 innate and adaptive immune responses are critical factors involved in malignant progression. The Basis of HPV Infection and Cell Transformation Human papillomaviruses have their genome divided in three main regions: the non-coding upper regulatory region, the early region encoding six known oncoproteins and the late region, which encodes the capsid proteins. The oncoproteins encoded in the early region, mainly E6 and E7 but also E5, are key participants in the infectious cycle and are the source of the genetic alterations which potentially lead to HPV-associated carcinogenesis as reviewed elsewhere [24]. The E5 oncogene is believed to have low impact in the maintenance of the malignant phenotype of cervical cancer cells as it appears to be deleted in various carcinomas [25]. Nevertheless, when present, it takes an important part in multiple mechanisms favouring neoplastic transformation. Some in vitro studies have shown that the HPV16 E5 oncoprotein can promote cell cycle progression and DNA synthesis by down-regulating the expression of tumour suppressor proteins p21 and p27 [26, 27]. Moreover, this oncoprotein is central to evasion of the host immunosurveillance. In E5-expressing cell lines antigen-presenting major histocompatibility complex (MHC) class I molecules appear to be trapped in the Golgi apparatus which reduced their surface expression [28, 29]. The down-regulation of MHC class I on cells expressing E5 ultimately impairs their recognition and clearance by cytotoxic T lymphocytes (CTL). The early genes E6 and E7 are known to have great oncogenic potential. They play a pivotal role in HPV-mediated carcinogenesis as they react with important tumour suppressor proteins impairing their functions. Thus, E6 binds to and promotes the degradation of the p53 tumour suppressor protein [30, 31], impairing important mechanisms of defence such as DNA repair and apoptosis. Additionally, the E6 protein facilitates carcinogenesis by reducing cellular senescence (as a result of telomerase activation [32]), activating the NF-κB pathway and evading the interferon response [24]. The E7 oncoprotein binds to tumour suppressor proteins of the retinoblastoma family of proteins (pRb) and induces their inactivation [33, 34], therefore increasing DNA synthesis and boosting cellular proliferation. Furthermore, E7 is able to promote genomic instability, which increases the likelihood of malignant progression [24]. Together, the E6 and E7 oncogenes are capable of transforming and induce the immortalization of keratinocytes [35]. Until today, the three early proteins mentioned above appear to be, of all HPV proteins, the ones with higher oncogenic potential. Given their roles in evading host immunity, it is important to understand how the human immune system responds to HPV infection. 5 HPV Infection vs. Host Immune Response Along the years, mammals have grown complex innate and adaptive immune mechanisms to control autoimmunity and to protect the host against microbial infections. Viruses can be detected by cellular receptors generally designated as pattern recognition receptors (PRRs), which trigger innate immune mechanisms that subsequently may be followed by a more specific adaptive immune response. These cellular receptors are proficient at recognising several pathogen-associated molecular patterns, like viral DNA or RNA or bacterial cell wall or intracellular components, and also interacting with dangerassociated molecular patterns like intracellular molecules released from damaged cells and cells undergoing unprogrammed cell death [36]. HPVs are double-stranded DNA viruses that are most likely to be recognized by Toll-like receptor 9 (TLR-9). This receptor is intracellularly expressed in the membrane of endolysosomes and recognizes unmethylated CpG-DNA sequences harboured by pathogens like viruses, bacteria and protozoa [36]. However, HPVs appear to somehow surpass these PRRs. The oncoproteins E6 and E7 may promote the down-regulation of TLR-9 as has been shown in vitro for HPV16 [37]. Therefore, at the start, the HPV infection seems to occur without the awareness of the host immune system [38]. Furthermore, in case of a persistent infection, the virus impairs immune cell functions as well as several molecular pathways involved in the immune response [39]. Also, other innate immune mechanisms, like the IFN response, can be abrogated by the actions of some HPV oncoproteins, as mentioned above. HPV is transmitted through skin-to-skin contact, usually through micro wounds. There exists a variety of immune cells divided between the dermis and epidermis. In the epidermis, the main cellular population are keratinocytes but there can also be found Langerhans cells and T lymphocytes [39]. Keratinocytes, the main targets of HPV, are capable of some immune functions. These cells can produce and release cytokines, induce the activation of memory T cells and may also act as non-professional antigenpresenting cells (APCs) [40]. Hence, keratinocytes are of great importance in the initiation of an immune response against HPV infections. On the other hand, the majority of the socalled “classical” immune cells (macrophages, DCs, T lymphocytes and others) can be found in the dermis [39]. Both keratinocytes and immune cells on site are capable of proinflammatory cytokine production [41] and antigen presentation in response to several pathogens. Although the innate immune response to HPV infections remains poorly understood, the immunological processes described above are thought to be the main ones responsible for the initiation of an adaptive immune response, which will determine the clearance or persistence of the infection. In fact, only about 10% of HPV infections will 6 develop clinical symptoms and possibly evolve to cancer [2, 3]. Therefore, warts and low grade cervical intraepithelial neoplasms (CIN1) are in most cases forced to regress by the cell-mediated adaptive immune response [42]. It has been shown that CD4 + and CD8 + HPV16-specific T cells migrate to the skin after intradermal challenge with HPV16 peptides [43]. This cellular migration event appears to be followed by an antibodymediated response, as observed in HPV-infected women [44]. Still, antibody titres are low and many women may even not produce them. HPV-associated Chronic Inflammation The above described mechanisms are thought to explain how a human immune system reacts since the beginning of a HPV infection until the moment of immunological clearance. However, as seen earlier, HPV oncoproteins have ways of evading the host immune response. This evasion favours the persistence of the infection, which in turn may promote tumourigenesis. In this context, the cellular content of the inflammatory infiltrate is decisive for either neoplastic transformation or elimination of infected/transformed cells (Fig. 1). While some cellular populations support HPV’s immune evasion and cell transformation mechanisms, others have a role in inhibiting malignant progression or even in potentiating tumour regression. For instance, T regulatory (Treg) cells facilitate the viral immune evasion. These cells display the CD4 + CD25 + FoxP3 + phenotype and play an immunosuppressive role important to prevent self-aggression by the immune system [45]. Because of their immunosuppressive function, Treg cells are often associated to persistent infection and tumour progression. Indeed, studies have shown a correlation between increased frequencies of FoxP3 + T cells at both systemic [46] and cervical [47] level and HPV persistent infection. Furthermore, a study has shown the chemokine CXCL12 (which is not expressed in normal skin) to act as a chemoattractant for Treg cells [48]. This work shows evidence that HPV stimulates an increased production of CXCL12 which draws a rising number of FoxP3 + cells. Besides Treg cells, another subset of CD4 + T cells can have pro-tumour functions. Type-2 helper T (Th2) cells can also contribute to impair the protective immune response, predominantly of the Th1 type, in the tumour microenvironment as they produce IL-4, -5 and -13, as well as IL-10. As seen earlier, this type of immune response facilitates malignant progression [12]. Furthermore, macrophages are thought to be a major component of tumour infiltrates [8] and are highly frequent in HPV-associated tumours [49]. In the presence of Th2-type cytokines like IL-4 and IL-13, tumour-associated macrophages (TAMs) are alternatively activated and acquire the M2-phenotype [50]. Unlike the M1 macrophages, these M2 cells express 7 various features and functions that are commonly associated with immune regulation and tumour progression. For instance, M2 TAMs, together with mast cells, are an important source of growth factors (e.g.: vascular endothelial growth factor – VEGF), cytokines (e.g.: TNF-α, transforming growth factor-β), proteases (e.g.: matrix metalloproteinases) and other molecules which in turn promote angiogenesis and tissue remodelling [51, 52]. These events facilitate the intake of nutrients and provide other stimuli which help tumour growth, invasion and metastasis. On the other hand, TAMs can be classically activated upon microbial stimuli or induced by pro-inflammatory cytokines, such as IFN-γ, to play important anti-tumour functions [50]. M1-type TAMs can, therefore, act as inducer and effector cells in Th1-type immune responses against tumours. They usually display an IL12 high , IL-23 high phenotype and produce high levels of pro-inflammatory cytokines (IL-1β, IL-6 and TNF-α) and reactive nitrogen species [53]. Besides M1 TAMs, other cellular populations play important anti-tumour roles among the largely heterogeneous tumour inflammatory infiltrates. T lymphocytes are also important immune effectors. Type-1 responses, mediated by Th1 cells represent the prototypic response adequate to eradicate cancer cells. The differentiation of this type of CD4 + T cells is driven by IL-12 and they are characterised by the production of pro-inflammatory cytokines like IFN-γ. Th1 cancer cells can directly induce the killing of tumour cells through TNF-related apoptosisinducing ligand (TRAIL) and/or Fas Ligand (FasL) pathways [54] and can induce the cytotoxic activity of CD8 + T lymphocytes and other effector cells through the production of specific cytokines [55]. In turn, CD8 + T cells, which may differentiate into cytotoxic T lymphocytes (CTL), are commonly protective against intracellular pathogens, like viruses, and tumour cells. These lymphocytes were observed in higher frequency in HPV-positive carcinomas when compared with HPV-negative lesions [56]. Further, this difference correlates with the patients’ prognosis as a higher CD8 + T cell frequency is associated with augmented overall survival. Moreover, “natural killer” (NK) cells and NKT cells act similarly to CTL but are part of the innate immune response. Unlike CTL, NK cells do not require the presence of MHC I molecules on target cells to exert their cytotoxic function. Thus, if HPV successfully down-regulates MHC I expression, thereby avoiding the action of CTL, infected cells may still be targeted by NK cells. Like CTL, NK and NKT cells express IFN-γ, perforin and granzyme and are thus capable effectors in the anti-viral immune response [39] as well as being able to also participate in the killing of cancer cells [57]. Additionally, several APCs (like DCs and Langerhans cells) are of great value and are frequent in the tumour microenvironment. They are essential for the activation of effector cells through antigen presentation and cytokine production [13]. 8 Fig. 1 : Schematic representation of the immune cell infiltrate in HPVinduced carcinogenesis. From top to bottom, HPV infection induces neoplastic transformation with associated tissue remodelling events which lead to host immune response. This response could eit her contribute to disease progression or regression which is reflected in the cellular populations recruited in either case. CD8 + , Th1-type CD4 + T, NK cells and M1 macrophages are commonly associated with tumour regression while, Treg, Th2-type CD4 + T cells and M2 macrophages are usually related to malignant progression. 9 The cross-talk of immune cells with HPV-infected cells is still not fully understood. However, it is generally agreed that the inflammation associated with the viral infection facilitates the development of HPV-associated cancers. In general terms, regulatory T cells, M2 TAMs, mast cells and Th2 cells most likely represent the immune cells with the greatest contribution to persistent infections and malignant progression. Additionally, HPV-transformed keratinocytes were shown to produce the immunosuppressive cytokine IL-10 in HPV-associated cervical cancer [58], which may be triggered by the virus oncoproteins as an immune evasion mechanism. On the other hand, M1 macrophages, NK and NKT cells, CTL and Th1 cells are, with the support of APCs, the major participants in anti-viral responses and tumour regression. Still, it is important to emphasize that this whole interaction in the tumour milieu is greatly complex due to the heterogeneity of cellular populations therein, which likely justifies the difficulty in characterising the relationship between inflammatory cells and HPV infection. HPV-induced Carcinogenesis: The Best Models to Understand It Along the years, scientists have been improving our understanding of chronic HPV infection and its association with carcinogenesis. In vitro and in vivo animal models are key elements towards the evolution of scientific knowledge (Table 1) and the in vitro ones are already widely used in HPV-induced carcinogenesis research. HPV-related in vitro studies include, in most cases, the use of single-layer cultures of common cervical carcinoma cell lines (HeLa, SiHa, CaSki and others). In the last two decades, a more sophisticated type of in vitro assay was developed and used as an alternative model. Organotypic (or raft) cultures allow the proliferation and differentiation of epithelial cells at an air-liquid interface on a dermal-equivalent support [59-61]. Thus, these raft cultures are useful to study the events occurring in human stratified epithelia in the course of HPV infection and HPV-induced carcinogenesis. Alongside with the cell culture technique, in vivo animal models are extensively exploited for scientific purposes. On one hand, bovine cattle, dogs and rabbits can be used to study bovine, canine oral and cottontail rabbit papillomaviruses, respectively [62, 63]. The study of these animal papillomaviruses is important as they are etiologic agents of diseases in farm and companion animals, provide in vivo models for HPV research and help to discover new subjects to study on HPV. However, these large animals are not common objects of study because they are associated with a great deal of financial costs and ethical issues. In turn, the mouse is the most commonly used test subject in healthassociated studies. Mice display a great similarity to humans in terms of anatomy, physiology, including the immune one, and genetics which, in addition to its cost- 16 [123]. Bracken toxins are hypothesized to drive the malignant progression of papillomavirus-induced upper digestive lesions in cattle and human populations, as remarked by Chang et al. [124]. Ptaquiloside is a chemically unstable compound that has been shown to induce splenic white pulp atrophy, neutropenia, reduced NK cells activity and a B-cell lymphoproliferative malignancy [125-128]. However, until now, there are no reports of an association between ptaquiloside and CD8 + T lymphocytes. In turn, celecoxib (CXB) is an immunomodulatory drug which specifically inhibits cyclooxygenase2 (COX-2). This enzyme plays a key role in the development of an inflammatory response [129] and is over-expressed in several types of cancer [130-134]. By selectively inhibiting COX-2, celecoxib diminishes the production of prostaglandins without significantly impairing the COX-1 isoform [135]. Actually, a few studies have already enlightened the effect of this selective COX-2 inhibitor over cytotoxic T cells [136-140]. However, whether celecoxib enhances or decreases CD8 + T lymphocytes infiltration and activity varies between studies. Still, none of these studies analysed the effect of celecoxib over CTL in HPV-associated lesions. 17 Chapter 2: Objectives The purpose of this work was to assess the infiltration and activity of cytotoxic CD8 + T lymphocytes in HPV-induced skin lesions, using the K14-HPV16 mouse model. Thus, the work was focused in some more specific aims: - To perform a kinetic study of the CTL infiltration and activation by comparing lesions from different-aged mice; - To analyse the immunomodulatory effect of ptaquiloside over CD8 + T cell infiltration and activation; - To analyse the immunomodulatory effect of celecoxib over CD8 + T cell infiltration and activation. 18 19 Chapter 3: Kinetics of CD8 + T cells in human papillomavirus-induced lesions 1 1 The contents of this chapter were adapted from: - Santos C., Ferreirinha P., Sousa H., Ribeiro J., Bastos M.M.S.M., Faustino-Rocha A.I., Oliveira P.A., Medeiros R., Vilanova M., Gil da Costa R.M. (2015) Kinetics of CD8 + T cells in human papillomavirus-induced lesions – data from K14-HPV16 transgenic mice. Submitted to the journal Tumor Biology with the number TUBI-D-15-04094. 20 21 Introduction Human papillomavirus (HPV) is the main etiologic agent of cervical cancer [4], but can also originate anal, skin and head and neck malignancies [6, 141, 142]. Only a persistent infection by an oncogenic HPV type (mainly HPV 16 and 18) can induce malignant transformation. In approximately 90% of HPV infected individuals the host immune system is able to fight the virus and hinder its dissemination. However, in the remaining 10%, HPV can effectively evade immune defenses and lead to clinical disease [2, 3]. K14-HPV16 transgenic mice, created nearly two decades ago [73], are an useful in vivo animal model for the study of HPV-induced carcinogenesis. This model shares a number of morphologic and molecular similarities to HPV-related human disease [76], thus functioning as an excellent replica of the multi-stage process of carcinogenesis. Targeting of HPV16 oncogenes to keratinocytes by the keratin-14 (K14) promoter/enhancer is the key characteristic of this model [73]. HPV-induced carcinogenesis is associated with progressively intense chronic inflammation. Therefore, a great diversity of immune cells and a multiplicity of soluble mediators can be found within the tumour microenvironment [18]. The inflammatory infiltrates provide proand anti-tumour stimuli, which will favour either the development or the regression of the lesion [55, 143]. CD8 + T cells are restricted to major histocompatibility complex class I (MHC I) molecules which can present peptides generated from intracellular viruses and/or tumour cells [117]. Upon recognizing specific peptides presented on MHC I molecules on the surface of professional antigen presenting cells, which also display co-stimulatory molecules, these T cells can differentiate into cytotoxic T lymphocytes (CTL), helped by cytokine stimuli [116]. CTL are an important part of tumour-specific immunity. They may eliminate target cells either directly, through the release of lytic granules containing several enzymes (such as perforin and granzyme) or by the engagement of death receptors (e.g.: FAS-FASL), and indirectly, following an interferon-γ-dependent mechanism which leads to cell cycle inhibition, apoptosis and stimulation of macrophage anti-tumour activity [118, 119]. Thus, using the K14-HPV16 transgenic mouse model, the aim was to examine the kinetics of CD8 + T cell infiltration in HPV-induced lesions during multi-step carcinogenesis at different time points. The goal was also pointed to understand whether these cells were activated by determining the presence of the lysosome-associated membrane protein 1 (LAMP1), also known as CD107a, at the cell surface. 22 Material & Methods Animals Generation of K14-HPV16 mice on a FVB/n background has been previously reported [73]. K14-HPV16 transgenic mice were generously donated by Drs. Jeffrey Arbeit and Douglas Hanahan (University of California) through the USA National Cancer Institute Mouse Repository. The animal experiments were approved by the Universidade de Trásos-Montes e Alto Douro ethics committee (10/2013) and the Portuguese Veterinary Directorate (0421/000/000/2014). Animals were maintained and bred according to Portuguese (Decreto-Lei 113/2013, August 7 th ) and European (EU Directive 2010/63/EU) legislation, under controlled conditions of temperature (23 ± 2 ºC), light-dark cycle (12h light/12h dark) and relative humidity (50 ± 10 %), using hardwood bedding. Food and water were provided ad libitum. Mice genotyping Animals were genotyped at weaning, using tail tip samples as described previously [113, 114]. Briefly, nucleic acids were extracted and DNA quality and purity were assessed. HPV16-E6 and -E2 genes were amplified to confirm the presence of HPV DNA and mouse β-globin was used as control. Amplicons lengths were confirmed by agarose gel electrophoresis. Only hemizygous females were used for the transgenic mouse groups in the experiment. Study Design Ten wild-type (WT) (HPV16 -/- , Group 1) and twelve transgenic (HPV16 +/- , Group 2) female mice were euthanized at 24-26 weeks of age. Latter, six HPV16 -/- (Group 3) and six HPV16 +/- (Group 4) female mice were sacrificed when 28-30 weeks-old. The mice were humanely sacrificed by intraperitoneal pentobarbital overdose, followed by exsanguination through cardiac puncture. Chest skin samples (approximately 4 cm 2 ) were collected for cell isolation and flow cytometry analysis. Matched samples were collected for histological examination. Histology Skin samples were fixated in 10% neutral buffered formalin for 48h. Samples were dehydrated through graded alcohols and xylene and paraffin-embedded in an automatic STP 120 processor (Micron, Boise, ID). 2 µm-thick sections were stained with haematoxylin-eosin (H&E) for histological examination on light microscopy. Skin samples 23 were classified as normal skin, epidermal hyperplasia or epidermal dysplasia. Minor dysplastic foci on a hyperplastic background were also recorded. Preparation of Single-Cell Suspensions Chest skin samples were cleaned from excessive blood vessels and fat tissue and cut into small pieces. These fragments were incubated for 2 hours with 125 U/ml type I collagenase (Gibco, Life Technologies, Paisley, UK) in RPMI-1640 medium complemented with 1% glutamine, 1% penicillin-streptomycin-amphotericin B, 1% HEPES buffer (all from Sigma, St. Louis, MO) and 10% foetal bovine serum (BioWest, Nuaillé, France) at 37 ºC and 150 rpm in a 3031 orbital incubator (GFL, Burgwedel, Germany). The resulting cell suspension was then filtered and centrifuged at 300  for 10 min at 4 ºC. Cells were then ressuspended in phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide followed by extracellular staining. Immunophenotyping Following cell isolation, the surface phenotype of the collected cells was assessed by flow cytometry using specific monoclonal antibodies (mAb). To prevent non-specific antibody binding, cells were incubated with anti-mouse CD16/CD32 mAb for FcγR blocking. This was followed by incubation with anti-CD8 mAb phycoerythrin-cychrome 5conjugate (clone 53-6.7, BD Biosciences, San Diego, CA) and anti-CD107a mAb phycoerythrin-conjugate (clone eBio1d4b, eBioscience, San Diego, CA). Following extracellular staining, the cells were washed, fixed in 2% formaldehyde and washed with phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide. Antibody-labelled cells were analysed in an EPICS XL flow cytometer using the EXPO32ADC software (Beckman Coulter, Miami, FL). The collected data files were analysed using the FlowJo software v10.0.7 (FLOWJO, LLC, Ashland, OR). Statistical Analysis Statistical analyses were executed in the GraphPad software (version 6.0, GraphPad Software, Inc. La Jolla, CA). Statistical analysis between group pairs was performed using the Mann-Whitney test. 24 Results Transgenic Mice Show Epidermal Hyperplasia and Dysplasia Analysis of skin samples showed that the totality of WT mice (groups 1 and 3) presented normal skin histology (Fig. 2a) whilst all transgenic mice presented skin lesions. In all 12 mice composing group 2 it was possible to observe simple to papillary, diffuse, variably severe epidermal hyperplasia and papillomatosis with orthokeratotic hyperkeratosis (Fig. 2b). Inflammation was mild, with a few macrophages, lymphocytes and mast cells present in the superficial dermis. Also in this group, 2 animals (16.7%) presented small epidermal dysplastic foci. In group 4, 3 animals (50.0%) presented diffuse epidermal dysplasia (Fig. 2c); other 3 (50.0%) showed multifocal epidermal dysplasia in a hyperplastic background. Sub-epidermal angiogenesis and dermal inflammatory infiltrates were prominent, showing numerous mixed mononuclear leukocytes and neutrophils. Fig. 2: Histopathological changes induced by HPV16 oncogenes in FVB/n mice, H&E. a – WT animal; Normal skin histology,  . b – 24-26 week-old HPV16 +/- animal; Epidermal hyperplasia extending to the follicular infundibulum and isthmus,  . c – 28-30 week-old HPV16 +/- animal; Epidermal dysplasia,  . Note enhanced sub-epidermal inflammatory cell infiltration. Increased CD8 + T lymphocytes numbers and activation in HPV16 +/- mice In order to assess the presence of CD8 + T cells in HPV-associated lesions, lymphoid cells were isolated from chest skin tissue and analysed by flow cytometry (Fig. 3a). As shown in Fig. 3b, chest skin samples from HPV16 +/- mice (groups 2 and 4) presented a significantly higher percentage of CD8 + T cells when compared with those of WT animals (groups 1 and 3). Although the percentage of skin CD8 + T cells was slightly higher in group 4 than in group 2, it did not reach statistical significant difference. In order to determine if the CD8 + T cells found in the skin of HPV16 +/- mice presented evidence of cytotoxic activity, the surface expression of CD107a was evaluated. In CD8 + T lymphocytes CD107a reaches the cell surface when lytic granules suffer exocytosis, thus exposing its membrane proteins [144]. Therefore, this lysosomeassociated membrane protein is a commonly used surrogate marker of CTL degranulation [145]. As shown in Fig. 3c, the percentage of CD107a + CD8 + T cells in chest skin samples 25 of mice from groups 2 and 4 were found significantly higher when compared with respective controls of groups 1 and 3. This indicates that in the HPV16 +/- mice skin infiltrating CTL released cytotoxic granules. Moreover, in group 4 mice, 100% of which show multifocal or diffuse dysplasia, a markedly higher percentage of CD8 + T cells express CD107a (P  0.001) than in group 2 mice, of which only 16.7% show focal dysplastic lesions. This result shows a positive correlation in the proportions of activated CD8 + T cells and lesion severity. Fig. 3: Percentage of CD8 + and CD8 + CD107a + T cells within total lymphoid-gated cells obtained from chest skin samples from WT and K14-HPV16 transgenic mice. a – Representative analysis of the gating strategy employed. Numbers within graphs correspond to the percentage of the gated population. b – Percentages of CD8 + T cells and c – percentages of CD8 + CD107a + T cells in gated CD8 + T cells were determined by flow cytometry after skin tissue collection and digestion with collagenase. Chest skin samples were collected from 24-26 weeks-old HPV -/- and HPV +/- mice (Groups 1 and 2, respectively) and 28-30 weeks-old HPV -/- and HPV +/- mice (Groups 3 and 4, respectively). Group 1, n = 10; Group 2, n = 12; Group 3, n = 6; Group 4, n = 6. Each dot represents an individual animal. Bars represent the mean value in each group. ** P  0.01; *** P  0.001. 32 low MHC I expression, CD8 + T cells play a decisive role in fighting HPV-induced oropharyngeal cancers [155]. In the present work, we hypothesize that ptaquiloside, the bracken toxin, exerts its immunosuppressive effect by counteracting the action of CD8 + T cells against papillomavirus-induced lesions. This would facilitate viral persistence and the progression of early HPVand BPV-induced lesions, posing major risks for human and animal health. In order to test that hypothesis, we employed HPV16-transgenic mice and analysed the effect of ptaquiloside on the population of skin infiltrating CD8 + T cells. Material & Methods Mice Construction of K14/HPV16 mice on a FVB/n background has been previously reported [73]. These animals develop characteristic multi-stage cutaneous and uterine cervical carcinogenesis, and were generously donated by Drs. Jeffrey Arbeit and Douglas Hanahan (University of California) through the USA National Cancer Institute Mouse Repository. The animal experiments were approved by the Universidade de Trás-osMontes e Alto Douro ethics committee (10/2013) and the Portuguese Veterinary Directorate (0421/000/000/2014). Animals were maintained and bred according to Portuguese (Decreto-Lei 113, August 7 th ) and European (EU Directive 2010/63/EU) legislation, under controlled conditions of temperature (23 ± 2 ºC), light-dark cycle (12h light/12h dark) and relative humidity (50 ± 10 %), using hardwood bedding. Food and water were provided ad libitum. Mice genotyping Animals were genotyped at weaning, using tail tip samples as described previously [113, 114]. Briefly, nucleic acids were extracted and DNA quality and purity were assessed. HPV16-E6 and -E2 genes were amplified to confirm the integration of HPV DNA into the mouse genome and a fragment of mouse β-globin was also amplified to confirm the quality of the extracted DNA. Lengths of the fragments were confirmed by agarose gel electrophoresis. Only hemizygous female mice were used for the transgenic mice groups in the experiment. Ptaquiloside isolation Ptaquiloside was isolated from bracken as previously described [156] with minor modifications. Briefly, 1000 g (dried weight) bracken crosiers were harvested at Arcos de Valdevez, Portugal, 41º 49´ 12´´ N, 8º 24´ 11´´ W) and a sample was deposited at the 33 Universidade de Trás-os-Montes e Alto Douro herbarium (reference no. 18248). Bracken was blended in distilled water (10 L), stirred at room temperature for 1 hour and the extract was adsorbed on 3 L of XAD-2 resin (Supelco, Sigma, St. Louis, MO). The resin was eluted with methanol (10 L) and the methanol extract was concentrated, dissolved in water (400 ml) and extracted with butanol (5500 ml). The butanol extract was chromatographed on silica gel (Merck, Kenilworth, NJ). Fractions containing ptaquiloside were separated twice on octadecyl-sylane silica gel (Fujy-Silysia, Kasugai Aichi, Japan) using methanol-water mixtures to obtain pure ptaquiloside. The compound was distinguished from other closely-related bracken illudane toxins on the basis of its characteristic 1 H and 13 C nuclear magnetic resonance (NMR) signals [157] using an Avance III 400 MHz spectrometer (Brucker, Billerica, MA). Aliquots were prepared for each experimental week (7.5 mg), freeze-dried and kept at -20ºC until use. Study Design Thirty transgenic (HPV16 +/- ) and 15 wild-type (WT, HPV16 -/- ) female 18-20 weeksold mice, showing diffuse cutaneous crusting and papillomatosis were employed. The animals were separated into three experimental groups: group 1 (n = 15, HPV16 -/- mice), group 2 (n = 15, HPV +/- untreated mice) and group 3 (n = 15, HPV16 +/- mice treated orally with 0.5 mg ptaquiloside per week, for 10 consecutive weeks). The experimental animals were monitored daily for signs of stress or disease. All surviving mice were euthanized at 28-30 weeks of age by an intraperitoneal pentobarbital overdose, followed by cardiac puncture and exsanguination. Chest skin samples (approximately 4 cm 2 ) were collected for flow cytometry analysis. Matched samples were collected for histological examination. Ptaquiloside administration and toxicity Each week, a 7.5 mg ptaquiloside aliquot was dissolved in 300 µl ethanol (25 mg/mL) and the individual 0.5 mg dose (20 µl ethanol) was added to a standard wheat cookie (Vieira, Vila Nova de Famalicão, Portugal) fragment weighting ca. 100 mg. Each dosed fragment was allowed to dry at room temperature for 10 minutes and individually administered to a group 3 animal, in an individual empty cage; the ingestion was visually monitored. In order to confirm that ptaquiloside was indeed active at the administered dosage, we looked for a lymphoid malignancy [128]. Thus, histological analysis of kidney samples was used to confirm neoplastic lymphoid cell infiltration in this organ. 34 Skin histology Skin samples were fixated in 10% neutral buffered formalin. Samples were dehydrated through graded alcohols and xylene and, paraffin embedded in an automatic STP 120 processor (Micron, Boise, ID). 2 µm-thick sections were stained with haematoxylin-eosin (H&E) for histological examination on light microscopy. Skin samples were classified as normal skin, epidermal hyperplasia, multifocal epidermal dysplasia in a hyperplastic background and diffuse epidermal dysplasia. Isolation of a Single-Cell Suspension Chest skin samples were cleaned from excessive blood vessels and fat tissue and cut into small pieces. This fragments were incubated for 2 hours with 125 U/ml of type I collagenase (Gibco, Life Technologies, Paisley, UK) in RPMI-1640 medium complemented with 1% glutamine, 1% penicillin-streptomycin-amphotericin B, 1% HEPES buffer (all from Sigma) and 10% foetal bovine serum (BioWest, Nuaillé, France) at 37 ºC and 150 rpm in a 3031 shaking incubator (GFL, Burgwedel, Germany). The resulting cell suspension was then filtered and centrifuged at 300  for 10 min at 4 ºC. Cells were then ressuspended in phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide followed by extracellular staining. Flow Cytometry Analysis Following cell isolation, the cellular immune phenotype was assessed by flow cytometry using monoclonal antibodies (mAb). Non-specific antibody binding was prevented by incubating cells with anti-mouse CD16/CD32 mAb for FcγR blocking. This was followed by incubation with anti-CD8 phycoerythrin-cychrome 5-conjugate (clone 536.7, BD Biosciences, San Diego, CA), anti-CD107a (LAMP1) phycoerythrin-conjugate (clone eBio1d4b) and anti-CD44 phycoerythrin-cychrome 7-conjugate (clone IM7) mAb (both from eBioscience, San Diego, CA). Following extracellular staining, the cells were washed, fixed in 2% formaldehyde and washed with phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide. Antibody-labelled cells were analysed in an EPICS XL flow cytometer using the EXPO32ADC software (Beckman Coulter, Miami, FL). The collected data files were analysed using the FlowJo software v10.0.7 (FLOWJO, LLC, Ashland, OR). Statistical Analysis Flow cytometry statistical analyses were executed in the GraphPad software (version 6.0, GraphPad Software, Inc. La Jolla, CA). In column and dot graphs, each point is representative of an individual mouse and bars represent the mean value for the 35 respective group. Analysis between group pairs was performed using the Mann-Whitney test. Kaplan-Meier survival analysis coupled with a log rank test was performed using the PASW Statistics software (version 18, IBM® SPSS®, Quarry Bay, Hong Kong). Results General findings Ptaquiloside was isolated at a 0.01% yield from bracken, as previously reported [128]. The compound's structure (Fig. 4a) was confirmed using NMR analysis and no other illudane glycoside was detected (Fig. 4b, c). Transgenic animals showed characteristic diffuse cutaneous hyperkeratosis and erythema, together with variably intense pruritus, while WT mice showed normal skin. All mice in groups 1 and 2 survived the 10 weeks experimental period. Ten mice (66.7%) from group 3 succumbed before the end of the study (Fig. 4d). Histological analysis of kidney samples showed normal histology in WT and untreated HPV +/- animals (Fig. 4e) and, moderate to severe, multifocal, perivascular infiltration of lymphoblastic cells, showing highly pleomorphic nuclei and up to 3 mitotic figures per high-power field in 100% of ptaquiloside-treated animals (group 3) (Fig. 4f). 36 Fig. 4: Ptaquiloside and its leukaemogenic effect in K14-HPV16 mice. a – ptaquiloside's structural formula. b and c – partial 2D NMR-HSQC spectra (obtained in CD3OD) for ptaquiloside. Note the correlations between carbons and their corresponding hydrogens in the glucose residue (b). Note in particular the correlations between carbons 12 and 13 at 5.87 ppm and 10.86 ppm respectively, and their corresponding hydrogens in the cyclopropylidene ring (c). d – Kaplan-Meier survival analysis. HPV +/- mice treated with ptaquiloside show significantly reduced survival (P  0.001) compared with untreated HPV -/- or HPV +/- animals. e – Untreated HPV +/- animal (group 2), showing normal kidney histology; H&E 200 . f – Ptaquiloside-exposed HPV +/- animal (group 3) showing typical ptaquilosideassociated perivascular infiltration by leukaemic lymphocytes. Note marked nuclear pleomorphism and mitotic figures; H&E 200 . 37 Cutaneous lesions The histological analysis showed that all group 1 mice (100.0%) presented normal skin histology (Fig. 5a) whilst all transgenic mice (groups 2 and 3) presented skin lesions. In group 2, 3 animals (50.0%) presented diffuse epidermal dysplasia (Fig. 5b); other 3 (50.0%) showed epidermal hyperplasia with multifocal dysplasia. Sub-epidermal angiogenesis and dermal inflammatory infiltrates were prominent in dysplastic lesions, showing numerous mixed mononuclear leukocytes and neutrophils. In group 3, the totality of mice (100.0%) showed diffuse epidermal dysplasia (Fig. 5c). Fig. 5: Histopathological changes induced by HPV16 oncogenes in FVB/n mice, H&E. a – group 1 animal. Normal skin histology, 400 . b – group 2 animal. Epidermal hyperplasia extending to the follicular infundibulum, 400 . c – group 3 animal. Epidermal dysplasia, 400 . Note loss of keratinocytic polarity and differentiation. CD8 + T lymphocytes are present in ptaquiloside-treated and untreated animals In order to determine the proportions of CD8 + T cells present in chest skin with or without HPV-associated lesions, samples were collected and cells isolated, followed by flow cytometry analysis (Fig. 6a). As shown in Fig. 6b, group 2 and 3 (HPV +/- and ptaquiloside-treated HPV +/- , respectively) mice showed significantly higher percentages of CD8 + T cells compared with HPV -/- mice (P  0.05). The percentage of CD8 + T cells in chest skin was not significantly different between ptaquiloside-treated and untreated HPV16 +/- mice (groups 3 and 2, respectively). 38 Fig. 6: Percentage of CD8 + T cells within total lymphoid-gated cells obtained from chest skin samples from WT and K14-HPV16 transgenic mice. a – Representative analysis of the gating strategy utilized. The number within the graph corresponds to the percentage of the gated population. b – Percentages of CD8 + T cells were determined by flow cytometry after skin tissue collection and enzymatic digestion. Chest skin samples were collected from 28-30 weeks-old HPV -/- (Group 1), HPV +/- mice (Group 2) and ptaquiloside-treated (PTQ) HPV +/- mice (Group 3). Group 1, n = 5; Group 2, n = 5; Group 3, n = 5. Each dot represents an individual animal. Bars represent the mean value in each group. * P  0.05. Ptaquiloside decreases the number of CD8 + CD107a + T cells in HPV16 +/- mice Next, CD8 + T cells found in skin samples were analysed with the objective to determine if they were actively degranulating, which indicates ongoing cytotoxic activity. For that purpose, flow cytometry analysis was performed to assess the expression of surface CD107a in these cells (Fig. 7a). A significantly higher percentage of CD8 + CD107a + T lymphocytes was observed in HPV +/- mice as compared with HPV -/- animals (P  0.01) (Fig. 7b). Additionally, HPV +/- mice exposed to ptaquiloside also presented a higher percentage of CD8 + CD107a + T lymphocytes when compared to WT animals, but this was significantly reduced when compared to untreated HPV +/- mice (P  0.05). 39 Fig. 7: Percentage of CD8 + CD107a + T cells within total CD8 + T cells obtained from chest skin samples from WT and K14-HPV16 transgenic mice. a – Representative analysis of the gating strategy utilized. Numbers within graphs correspond to the percentage of the gated population. b – Percentages of CD8 + CD107a + T cells in gated CD8 + T cells were determined by flow cytometry after skin tissue collection and enzymatic digestion. Chest skin samples were collected from 28-30 weeks-old HPV -/- (Group 1), HPV +/- mice (Group 2) and ptaquiloside-treated (PTQ) HPV +/- mice (Group 3). Group 1, n = 5; Group 2, n = 5; Group 3, n = 5. Each dot represents an individual animal. Bars represent the mean value in each group. *P 0.05; ** P  0.01. Reduced number of CD8 + CD44 + T cells in ptaquiloside-treated transgenic animals Having determined that CD8 + T cells presented a cell surface phenotype associated with degranulation, expression of CD44, a marker indicating a memory cell phenotype [158], was also assessed. As shown in Fig. 8 (a and b), a significantly higher percentage of CD8 + CD44 + T lymphocytes was found in HPV +/- mice as compared with HPV -/- animals (P  0.01). Ptaquiloside-treated HPV +/- mice presented a significantly reduced percentage of CD8 + CD44 + T lymphocytes compared with untreated HPV +/- mice (P  0.05). In fact, group 3 did not show statistically different values to group 1 (P  0.05). 40 Fig. 8: Percentage of CD8 + CD44 + T cells within total CD8 + T cells obtained from chest skin samples from WT and K14-HPV16 transgenic mice. a – Representative analysis of the gating strategy utilized. Numbers within graphs correspond to the percentage of the gated population. b – Percentages of CD8 + CD44 + T cells in gated CD8 + T cells were determined by flow cytometry after skin tissue collection and enzymatic digestion. Chest skin samples were collected from 28-30 weeks-old HPV -/- (Group 1), HPV +/- mice (Group 2) and ptaquiloside-treated (PTQ) HPV +/- mice (Group 3). Group 1, n = 5; Group 2, n = 5; Group 3, n = 5. Each dot represents an individual animal. Bars represent the mean value in each group. * P  0.05;** P  0.01. Discussion CD8 + T lymphocytes are important effectors of the cell-mediated immune response, especially against virus-infected or tumour cells [117]. Upon activation these cells may degranulate, releasing a set of lytic enzymes capable of destroying target cells [159]. CD107a is a lysosome-associated membrane protein which can be used as surrogate marker of degranulation [145]. In CD8 + T lymphocytes, CD107a reaches the cell surface when lytic granules are exocytosed, thus exposing its membrane proteins [144]. As CD8 + T cells need to be activated in order to release cytotoxic granules, the presence of CD107a at the cell surface is useful to determine their activation status [145]. Additionally, CD8 + T cells may also acquire a memory phenotype, allowing a quicker recall response, as memory T cells can be more easily activated than naive T cells [159]. CD44 is a cell-surface glycoprotein and is currently considered the best marker to identify memory CD8 + T cells [158]. Furthermore, CD44 is thought to act as a regulator of the motility of CTL in the tumour microenvironment [160]. 41 Although many papillomavirus (including some oncogenic HPV and BPV types) down-regulate MHC I expression as a strategy to evade immune surveillance mediated by CTL [154], these cells still play a major role against papillomavirus-induced lesions, namely in HPV-induced oropharyngeal cancers [155]. Evading the host immune response is critical for maintaining a long-term infection, allowing tumour progression. In this context, environmental (e.g. dietary) immunosuppressant products such as bracken and its toxin, ptaquiloside, may play an important role. In this work, as expected, WT mice had no HPV-associated lesions, whilst agematched K14-HPV16 transgenic mice showed hyperplastic lesions with dysplastic foci or diffuse dysplasia. It is possible that the diffuse dysplastic lesions observed in ptaquilosidetreated animals represent a more aggressive stage in multi-step carcinogenesis compared with the often focal dysplasia observed in untreated animals. However, the small number of animals that survived the experimental period and the absence of obviously invasive lesions do not allow us to firmly conclude that ptaquiloside enhanced tumour progression. The increasing aggressiveness of dysplastic versus hyperplastic lesions was accompanied by an increased bulk of infiltrating immune cells. These results indicate that HPV-induced hyperplastic and dysplastic skin lesions in K14-HPV16 transgenic mice show increased infiltration of CD8 + T cells when compared to normal skin from age-matched WT mice. Ptaquiloside reduced the CD8 + T cell infiltration, although only slightly (P > 0.05). Nevertheless, ptaquiloside did induce a significant reduction in CD107a + and CD44 + CD8 + T cells. Altogether, these data show that cytotoxic T cells migrate to the site of HPVinduced lesions, are activated to degranulate and acquire a memory-phenotype and, that ptaquiloside is capable of impairing the function of these cells, explicitly by inhibiting cellular activation and memory phenotype acquisition. The molecular mechanism behind this impairment should be an alluring topic for future research. Grazing animals and human populations worldwide are easily exposed to bracken and its illudane toxins, including ptaquiloside and other related compounds [123]. Bracken is an abundant weed, especially in poorer pastures, and ptaquiloside accumulates in bovine and ovine milk and meat [161-163], besides contaminating underground waters [164]. Moreover, some human populations in different countries consume bracken shoots (known as croziers, broto de samambaia in Portuguese or warabi in Japanese) as part of their daily diet [123]. In China, the production of dry Pteridium aquilinum var. latiusculum for food uses (locally known as juecai) was estimated to involve approximately 1000 companies in a business worth 300 million USD per annum [165]. This is considered the most widely-consumed fern in China, with an estimated annual production of 1200 tons in the Zhouzhou county, province of Hunan. Bracken and ptaquiloside are well-known 48 molecularly [76], making K14-HPV16 mice a particularly useful model for HPV research. Material & Methods Animals Generation of K14/HPV16 mice on a FVB/n background has been previously reported [73]. K14-HPV16 transgenic mice were kindly donated by Drs. Jeffrey Arbeit and Douglas Hanahan (University of California) through the USA National Cancer Institute Mouse Repository. The animal experiments were approved by the Universidade de Trásos-Montes e Alto Douro Ethics Committee (10/2013) and the Portuguese General Veterinary Directorate (approval no. 0421/000/000/2014). Animals were maintained and bred according to Portuguese (Decreto-Lei 113, August 7 th ) and European (EU Directive 2010/63/EU) legislation, under controlled conditions of temperature (23 ± 2 ºC), light-dark cycle (12h light/12h dark) and relative humidity (50 ± 10 %), using corncob bedding. Food and water were provided ad libitum. Mice genotyping Animals were genotyped at weaning, using tail tip samples as described previously [113, 114]. Briefly, nucleic acids were extracted and DNA quality and purity were assessed. HPV16-E6 and -E2 genes were amplified to confirm the presence of HPV DNA and a fragment of mouse β-globin was also amplified as endogenous control. Lengths of the amplicons were confirmed by agarose gel electrophoresis. Only hemizygous females were used for the transgenic mouse groups in the experiment. Experimental Design Fifty-four 18 to 20 weeks-old female mice were divided into four experimental groups, according to their genotype and taking into consideration that HPV transgenes and celecoxib (CXB) administration could induce some mortality: group 1 (HPV16 -/- untreated animals, n = 12), group 2 (HPV16 +/- untreated animals, n = 12), group 3 (HPV16 -/- CXB-treated animals, n = 15) and group 4 (HPV16 +/- CXB-treated animals, n = 15). All surviving mice were humanely euthanized at 24-26 weeks of age by intraperitoneal pentobarbital overdose, followed by exsanguination by cardiac puncture. Chest skin samples (approximately 4 cm 2 ) were collected for flow cytometry analysis and matched skin samples were collected for histological analysis. 49 Celecoxib administration Celecoxib (Pfizer, New York, NY) was dissolved in drinking water at a concentration of 0.5 mg/ml, estimating an average daily water intake of 5.0ml per mouse, and a dose of 46.7 mg/kg/day and 2.5 mg/animal/day in an average mouse weighting 30 g. This is a well-tolerated moderate dose, as shown in previous assays [177]. However, K14-HPV16 animals dramatically increased their water intake when CXB was added (possibly due to the highly palatable lactose present in the vehicle) reaching up to 15 ml per animal. The CXB concentration was thus reduced from the start of the third week onwards down to 0.2 mg/ml, resulting in a decrease in consumption and an effective dose of 93 mg/kg/day and 2.8 mg/animal/day. The average dose during the overall experimental period was thus 124 mg/kg/day and 3.72 mg/animal/day. Histological Analysis Skin samples were fixated in 10% neutral buffered formalin for 48 hours. Samples were dehydrated through graded alcohols and xylene and paraffin embedded in an automatic STP 120 processor (Micron, Boise, ID). 2 µm-thick sections were stained with haematoxylin-eosin (H&E) for histological evaluation on light microscopy. Skin samples were classified as normal skin, epidermal hyperplasia and epidermal dysplasia. Preparation of Single-Cell Suspensions Chest skin samples were cut into small pieces after excessive blood vessels and fat tissue removal. Skin fragments were incubated for 2 hours with 125 U/ml type I collagenase (Gibco, Life Technologies, Paisley, UK) in RPMI-1640 medium complemented with 1% glutamine, 1% penicillin-streptomycin-amphotericin B, 1% HEPES buffer (all from Sigma, St. Louis, MO) and 10% foetal bovine serum (BioWest, Nuaillé, France) at 37 ºC and 150 rpm in a 3031 orbital incubator (GFL, Burgwedel, Germany). Subsequently, the resulting cell suspension was filtered and centrifuged at 300  for 10 min at 4 ºC. Cells were ressuspended in phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide followed by flow cytometry analysis. Immunophenotyping Following cell isolation, the surface phenotype of the collected cells was assessed by flow cytometry using specific monoclonal antibodies (mAb). Cells were incubated with anti-mouse CD16/CD32 mAb for FcγR blocking, to prevent non-specific antibody binding. Next, cells were incubated with anti-CD8 mAb phycoerythrin-cychrome 5-conjugate (clone 53-6.7, BD Biosciences, San Diego, CA) and anti-CD107a (LAMP1) mAb phycoerythrin- 50 conjugate (clone eBio1d4b, eBioscience, San Diego, CA). Following extracellular staining, the cells were washed, fixed in 2% formaldehyde and washed with phosphate buffered saline containing 1% bovine serum albumin and 20 mM sodium azide. Antibody-labelled cells were analysed in an EPICS XL flow cytometer using the EXPO32ADC software (Beckman Coulter, Miami, FL). The assembled data files were analysed using the FlowJo software v10.0.7 (FLOWJO, LLC, Ashland, OR). Statistical Analysis Statistical analyses were performed using the GraphPad software (version 6.0, GraphPad Software, Inc. La Jolla, CA). In column and dot graphs each point represents individual mice, while bars represent the mean for the respective group. Statistical analysis between group pairs was done using the Mann-Whitney test. Results General findings All K14-HPV16 mice showed typical cutaneous changes, including diffuse hyperkeratosis and erythema. Celecoxib-treated HPV +/- mice (group 4) showed significant mortality: 10 out of 15 mice (66.7%) succumbed before the end of the study. All mice in groups 1, 2 and 3 survived until the end of the study. Histological analysis Histological analysis of skin samples (Table 2) showed that all WT mice, groups 1 and 3, presented normal skin histology (Fig. 9a). In groups 2 and 4 (untreated and CXBtreated HPV +/- mice, respectively), the totality of mice showed simple to papillary, diffuse, variably severe epidermal hyperplasia extending to the follicular infundibula and papillomatosis with orthokeratotic hyperkeratosis (Fig. 9b). Additionally, 2 animals (16.7%) from group 2 presented multifocal epidermal dysplastic foci with parakeratotic hyperkeratosis within the hyperplastic background (Fig. 9c). There were signs of mild inflammation, with a small amount of macrophages, mast cells and lymphocytes in the superficial dermis. Dysplastic foci were associated with increased numbers of leukocytes infiltrating the dermo-epidermic junction and with intense angiogenesis. 51 Table 2: Histological classification of HPV16-induced skin lesions from 24-26 weeks-old female mice. Celecoxib – CXB. Group Cutaneous lesions Incidence (%) Normal skin Epidermal hyperplasia Epidermal dysplasia 1 (HPV - / - , n = 12) 12/12 (100%) 0/0 (0%) 0/0 (0%) 2 (HPV +/ - , n = 12) 0/0 (0%) 12/12 (100%) 2/12 (16.7%) 3 (HPV - / - + CXB, n = 15) 15/15 (100%) 0/0 (0%) 0/0 (0%) 4 (HPV +/ - + CXB, n = 5) 0/0 (0%) 5/5 (100%) 0/0 (0%) Fig. 9: Histopathological changes induced by HPV16 oncogenes in FVB/n mice, H&E. a – WT animal; Normal skin histology, 100 . b – CXB-treated HPV16 +/- animal; Epidermal hyperplasia extending to the follicular infundibulum and isthmus, 100 . c – Untreated HPV16 +/- mouse; Epidermal dysplasia, 200 . Note marked parakeratotic hyperkeratosis, loss of cell polarity, enhanced anisokaryosis and mitotic activity. The dermal-epidermal junction is obscured by severe inflammatory cell infiltration. CTL infiltration and activation in celecoxib-treated mice With the purpose to study CD8 + T cell infiltration in HPV-induced lesions, lymphoid cells from mice chest skin tissue were isolated and a flow cytometry analysis of the recovered cells was performed (Fig. 10a). Skin samples from untreated HPV +/- mice showed a significantly higher percentage of CD8 + T cells compared with untreated WT animals (P  0.01) (Fig. 10b). Also, although statistical significance was not achieved, CXB-treated WT mice showed decreased CD8 + T cell infiltration when compared to CXBtreated HPV +/- animals and the same was observed between samples from WT mice treated with CXB (group 3) and untreated WT animals (group 1) (Fig. 10b). Moreover, HPV +/- mice treated with CXB (group 4) showed less CD8 + T lymphocytes compared with untreated transgenic animals (group 2) (P  0.01) (Fig. 10b). Next, the objective was to study whether the CD8 + T lymphocytes found in mouse skin samples showed signs of cytotoxic activity (Fig. 10a). Thus, expression of CD107a (LAMP1) at the cell surface was assessed by flow cytometry. CD107a reaches the surface of CTL when lytic granules are released, exposing their membrane proteins during 52 exocytosis [144]. Thus, this lysosome-associated membrane protein is frequently used as an immunological marker of CTL degranulation [145]. Fig. 10c shows a significantly higher percentage of CD8 + CD107a + cells in samples from HPV +/- mice treated with CXB (group 4) when compared with untreated HPV +/- (group 2) (P  0.01). Fig. 10: Percentage of CD8 + and CD8 + CD107a + T cells within total lymphoid-gated cells obtained from chest skin samples from WT and K14-HPV16 transgenic mice. a – Representative analysis of the gating strategy employed. Numbers within graphs correspond to the percentage of the gated population. b – Percentages of CD8 + T cells and c – percentages of CD8 + CD107a + T cells in gated CD8 + T cells were determined by flow cytometry after skin tissue collection and digestion with collagenase. Chest skin samples were collected from WT and HPV +/- mice (Groups 1 and 2, respectively) and CXB-treated (CXB) WT and HPV +/- mice (Groups 3 and 4, respectively) at 24-26 weeks of age. Group 1, n = 5; Group 2, n = 5; Group 3, n = 5; Group 4, n = 5. Each dot represents an individual animal. Bars represent the mean value in each group. ** P  0.01. 53 Discussion The development of HPV-associated malignancies depends on a persistent HPV infection. Ultimately, the ability of the immune system to eliminate the virus is the key element to decide whether a HPV infection is cleared or evolves to cancer [2]. HPV presents well-known mechanisms to evade host immunity and delay its elimination, thus facilitating viral persistence [154]. When the virus is detected, an innate immune response occurs, and leads to the development of an adaptive immune response. A fundamental part of this adaptive response is cell-mediated immunity, characterized by the activity of a vast number of CD8 + and CD4 + T lymphocytes [39]. In fact, infiltration by these cells in HPV-associated lesions has already been shown to lead to regression [148, 178]. CTL infiltration drives the response induced by an experimental therapeutic vaccine in patients with cervical intraepithelial lesions [171] and correlates with a better prognosis in patients with HPV-positive head and neck cancer [155, 172, 173]. Chronic inflammation is a key feature associated with carcinogenesis, namely in the case of HPV infection [8]. COX-2 is an enzyme with an imperative role in the metabolism of arachidonic acid, which leads to the production of prostaglandins, which in turn promote inflammation [129]. In fact, COX-2 is over-expressed in several malignancies, including cervical cancer [130-134]. As inflammation is known to contribute for cancer progression [8], COX-2 inhibition is expected to result in tumour growth inhibition whilst reducing inflammation. Some non-steroidal anti-inflammatory drugs, like aspirin and ibuprofen, have already shown promising results as anti-tumour therapy in both patients and pre-clinical animal models [179-182]. However, these drugs have very low specificity. Selective COX-2 inhibitors such as CXB and rofecoxib have already been used to prevent the development of colorectal adenomas [183, 184]. Furthermore, this COX-2 inhibitor has been showing promising results in a tumour model of human colon cancer when combined with chemotherapeutic drugs [185]. In this study, the effect of the selective COX-2 inhibitor CXB over the trafficking and activation of CD8 + T lymphocytes was examined in the K14-HPV16 mouse model, a proper model to study HPV-induced carcinogenesis due to its great similarities with the human clinical disease [76]. High mortality was observed among CXB-treated animals, presumably due to CXB-related toxicity. The results presented herein show that CXB reduces the number of tumourinfiltrating CD8 + T cells when compared with untreated mice. Still, despite the decrease in cell numbers, CXB-treated mice have a higher percentage of activated and degranulating CTL compared with untreated animals. These findings suggest that CXB reduces the number of tumour-infiltrating CD8 + T cells while enhancing their effector functions. These 54 data are in agreement with a previous study using glioma-bearing mice where COX2 -/- mice had increasing percentages of tumour-infiltrating CD8 + CD107a + lymphocytes [175]. This may be explained by another study reporting that COX-2 activity leads to CD8 + T cell senescence and this trajectory may be opposed by COX-2 inhibition, as shown by increased levels of CD28 and interleukin-2 in CD8 + T cells [174]. Furthermore, hyperplastic epidermal lesions were evident in all surviving transgenic mice, whilst all WT animals presented normal skin histology, as expected. Multifocal dysplastic epidermal lesions were restricted to untreated HPV +/- animals. These findings suggest that CXB blocked tumour progression at the hyperplastic stage, but the small number of dysplastic lesions observed does not allow for any definitive conclusions. In fact, COX-2 inhibition boosted the efficacy of a DNA vaccine expressing the HPV E7 oncogene, by enhancing tumour-infiltrating CD8 + T cells and slowing tumour growth [186]. However, this study was performed in mice bearing allografted TC1 lung cells immortalized by the HPV16 E6 and E7 oncogenes and transformed by the c-Ha-ras oncogene. Comparisons between this model and K14-HPV16 mice are limited, because allografts do not reproduce HPV-associated multi-step carcinogenesis, being directly implanted in the subcutis with a fully malignant phenotype. Also, keratinocytes and not lung cells are the targets for papillomavirus infection. It remains unclear whether CTL are activated at regional lymph nodes or at the lesion location and this would be an interesting point to address in the future. The results presented herein suggest that CXB induces augmented CTL degranulation in HPVinduced lesions, possibly contributing to prevent malignant progression in this animal model. Conclusion The present data confirm the potential of CXB to enhance degranulation by tumour-infiltrating CD8 + T lymphocytes. Moreover, the effect of CXB seems to be more complex than previously thought, as it also reduced the overall number of tumourinfiltrating CD8 + T cells. Future studies addressing the impact of COX-2 inhibitors on the prognosis of patients bearing HPV-induced lesions should take into account both the number of infiltrating CD8 + T cells and their activation status. 55 Chapter 6: General Discussion and Conclusions 56 57 HPV-associated malignancies are a genuine health problem worldwide, with emphasis on the increase in the number of HPV-associated oropharyngeal carcinomas [152]. Prophylactic vaccination remains the most effective method to control HPV infections. Currently, there are two effective commercialized vaccines: Cervarix™ (GlaxoSmithKline Biologicals, Rixensart, Belgium), a bivalent HPV16/18 vaccine; and Gardasil™ (Merck Vaccines, West Point, PA), a quadrivalent HPV6/11/16/18 vaccine. Both are HPV L1 virus-like particles (VLP) vaccines and they act by boosting the production of neutralising antibodies directed against the L1 capsid protein [187]. Despite their effectiveness, these vaccines present a few limitations, as the virus-type restriction. The first is restricted to two HPV types (bivalent) and the second is restricted to four types (quadrivalent). Looking to increase type-specific protection, Merck managed to create the nonavalent HPV L1 VLP vaccine by adding the VLPs from five oncogenic HPV types (31, 33, 45, 52 and 58) in addition to the four types used in the quadrivalent vaccine. This resulted in a 20% increase in protection, in addition to the 70% obtained from the quadrivalent vaccine, in a total of approximately 90% protection against cervival cancer [188]. However, despite these advances in disease prevention, an effective therapeutic vaccine against HPV-induced malignancies remains absent. In order to develop a therapeutic strategy, it is necessary to gather a large amount of information on the features of disease progression as well as of the events occurring in the lesions microenvironment. It is accepted that the carcinogenesis process induced by a HPV infection is marked by the existence of a prominent immune response [11, 18]. Thus, a great deal of cellular populations and chemical factors are present and play a specific role in carcinogenesis, either promoting or impairing it. At the end, it should be the balance between all proand anti-tumour stimuli that will decide whether the viral infection is cleared or is allowed to progress towards malignancy. Therefore, a therapeutic vaccine could possible enhance the action of anti-tumour effectors and/or inhibit the function of known pro-tumour mediators. Cytotoxic CD8 + T lymphocytes, for instance, are the main cells associated with the elimination of virus-infected and transformed cells. These cells are part of the adaptive immune response and depend on antigen presentation by MHC I molecules and other costimulatory signals to become activated and exert their functions [116, 117]. However, during the course of HPV infection, expression of MHC I molecules is usually downregulated. This is one of the most common immune evasion mechanisms employed by the HPV E5 oncoprotein [154]. Infiltration of CD8 + T cells in HPV-positive lesions has already been associated with better prognosis when compared with HPV-negative ones [155, 172, 173, 189]. Thus, boosting the activity of CD8 + T lymphocytes seems a possible therapeutic strategy to eliminate HPV-associated lesions. Still, the HPV immune evasion 64 31. Scheffner, M., B.A. Werness, J.M. Huibregtse, A.J. Levine, and P.M. Howley, 1990, The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell, 63(6): p. 1129-36. 32. Klingelhutz, A.J., S.A. Foster, and J.K. McDougall, 1996, Telomerase activation by the E6 gene product of human papillomavirus type 16. Nature, 380(6569): p. 79-82. 33. Dyson, N., P.M. Howley, K. Munger, and E. Harlow, 1989, The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science, 243(4893): p. 934-7. 34. Dyson, N., P. Guida, K. Munger, and E. Harlow, 1992, Homologous sequences in adenovirus E1A and human papillomavirus E7 proteins mediate interaction with the same set of cellular proteins. J Virol, 66(12): p. 6893-902. 35. Munger, K., W.C. Phelps, V. Bubb, P.M. Howley, and R. Schlegel, 1989, The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J Virol, 63(10): p. 4417-21. 36. Takeuchi, O. and S. Akira, 2010, Pattern recognition receptors and inflammation. Cell, 140(6): p. 805-20. 37. Hasan, U.A., E. Bates, F. Takeshita, A. Biliato, R. Accardi, V. Bouvard, M. Mansour, I. Vincent, L. Gissmann, T. Iftner, M. Sideri, F. Stubenrauch, and M. Tommasino, 2007, TLR9 expression and function is abolished by the cervical cancerassociated human papillomavirus type 16. J Immunol, 178(5): p. 3186-97. 38. Frazer, I.H., 2009, Interaction of human papillomaviruses with the host immune system: a well evolved relationship. Virology, 384(2): p. 410-4. 39. Hibma, M.H., 2012, The immune response to papillomavirus during infection persistence and regression. Open Virol J, 6: p. 241-8. 40. Amador-Molina, A., J.F. Hernandez-Valencia, E. Lamoyi, A. ContrerasParedes, and M. Lizano, 2013, Role of innate immunity against human papillomavirus (HPV) infections and effect of adjuvants in promoting specific immune response. Viruses, 5(11): p. 2624-42. 41. Miller, L.S. and R.L. Modlin, 2007, Human keratinocyte Toll-like receptors promote distinct immune responses. J Invest Dermatol, 127(2): p. 262-3. 42. Stanley, M.A., 2012, Epithelial cell responses to infection with human papillomavirus. Clin Microbiol Rev, 25(2): p. 215-22. 43. van den Hende, M., M.I. van Poelgeest, J.M. van der Hulst, J. de Jong, J.W. Drijfhout, G.J. Fleuren, A.R. Valentijn, A.R. Wafelman, G.M. Slappendel, C.J. Melief, R. Offringa, S.H. van der Burg, and G.G. Kenter, 2008, Skin reactions to human 65 papillomavirus (HPV) 16 specific antigens intradermally injected in healthy subjects and patients with cervical neoplasia. Int J Cancer, 123(1): p. 146-52. 44. Carter, J.J., L.A. Koutsky, G.C. Wipf, N.D. Christensen, S.K. Lee, J. Kuypers, N. Kiviat, and D.A. Galloway, 1996, The natural history of human papillomavirus type 16 capsid antibodies among a cohort of university women. J Infect Dis, 174(5): p. 927-36. 45. Halvorsen, E.C., S.M. Mahmoud, and K.L. Bennewith, 2014, Emerging roles of regulatory T cells in tumour progression and metastasis. Cancer Metastasis Rev. 46. Molling, J.W., T.D. de Gruijl, J. Glim, M. Moreno, L. Rozendaal, C.J. Meijer, A.J. van den Eertwegh, R.J. Scheper, M.E. von Blomberg, and H.J. Bontkes, 2007, CD4(+)CD25hi regulatory T-cell frequency correlates with persistence of human papillomavirus type 16 and T helper cell responses in patients with cervical intraepithelial neoplasia. Int J Cancer, 121(8): p. 1749-55. 47. Scott, M.E., Y. Ma, L. Kuzmich, and A.B. Moscicki, 2009, Diminished IFNgamma and IL-10 and elevated Foxp3 mRNA expression in the cervix are associated with CIN 2 or 3. Int J Cancer, 124(6): p. 1379-83. 48. Jaafar, F., E. Righi, V. Lindstrom, C. Linton, M. Nohadani, S. Van Noorden, T. Lloyd, J. Poznansky, G. Stamp, R. Dina, D.V. Coleman, and M.C. Poznansky, 2009, Correlation of CXCL12 expression and FoxP3+ cell infiltration with human papillomavirus infection and clinicopathological progression of cervical cancer. Am J Pathol, 175(4): p. 1525-35. 49. Lepique, A.P., K.R. Daghastanli, I.M. Cuccovia, and L.L. Villa, 2009, HPV16 tumor associated macrophages suppress antitumor T cell responses. Clin Cancer Res, 15(13): p. 4391-400. 50. Mantovani, A. and M. Locati, 2013, Tumor-associated macrophages as a paradigm of macrophage plasticity, diversity, and polarization: lessons and open questions. Arterioscler Thromb Vasc Biol, 33(7): p. 1478-83. 51. Ribatti, D. and E. Crivellato, 2012, Mast cells, angiogenesis, and tumour growth. Biochim Biophys Acta, 1822(1): p. 2-8. 52. Condeelis, J. and J.W. Pollard, 2006, Macrophages: obligate partners for tumor cell migration, invasion, and metastasis. Cell, 124(2): p. 263-6. 53. Locati, M., A. Mantovani, and A. Sica, 2013, Macrophage activation and polarization as an adaptive component of innate immunity. Adv Immunol, 120: p. 163-84. 54. Kayagaki, N., N. Yamaguchi, M. Nakayama, A. Kawasaki, H. Akiba, K. Okumura, and H. Yagita, 1999, Involvement of TNF-related apoptosis-inducing ligand in human CD4+ T cell-mediated cytotoxicity. J Immunol, 162(5): p. 2639-47. 66 55. Ruffell, B., D.G. DeNardo, N.I. Affara, and L.M. Coussens, 2010, Lymphocytes in cancer development: polarization towards pro-tumor immunity. Cytokine Growth Factor Rev, 21(1): p. 3-10. 56. Jung, A.C., S. Guihard, S. Krugell, S. Ledrappier, A. Brochot, V. Dalstein, S. Job, A. de Reynies, G. Noel, B. Wasylyk, C. Clavel, and J. Abecassis, 2013, CD8-alpha T-cell infiltration in human papillomavirus-related oropharyngeal carcinoma correlates with improved patient prognosis. Int J Cancer, 132(2): p. E26-36. 57. Robertson, F.C., J.A. Berzofsky, and M. Terabe, 2014, NKT cell networks in the regulation of tumor immunity. Front Immunol, 5: p. 543. 58. Alcocer-Gonzalez, J.M., J. Berumen, R. Tamez-Guerra, V. BermudezMorales, O. Peralta-Zaragoza, R. Hernandez-Pando, J. Moreno, P. Gariglio, and V. Madrid-Marina, 2006, In vivo expression of immunosuppressive cytokines in human papillomavirus-transformed cervical cancer cells. Viral Immunol, 19(3): p. 481-91. 59. Merrick, D.T., R.A. Blanton, A.M. Gown, and J.K. McDougall, 1992, Altered expression of proliferation and differentiation markers in human papillomavirus 16 and 18 immortalized epithelial cells grown in organotypic culture. Am J Pathol, 140(1): p. 167-77. 60. Blanton, R.A., N. Perez-Reyes, D.T. Merrick, and J.K. McDougall, 1991, Epithelial cells immortalized by human papillomaviruses have premalignant characteristics in organotypic culture. Am J Pathol, 138(3): p. 673-85. 61. Asselineau, D., B.A. Bernard, C. Bailly, M. Darmon, and M. Prunieras, 1986, Human epidermis reconstructed by culture: is it "normal"? J Invest Dermatol, 86(2): p. 181-6. 62. Campo, M.S., 2002, Animal models of papillomavirus pathogenesis. Virus Res, 89(2): p. 249-61. 63. Gil da Costa, R.M. and R. Medeiros, 2014, Bovine papillomavirus: opening new trends for comparative pathology. Arch Virol, 159(2): p. 191-8. 64. Nguyen, D. and T. Xu, 2008, The expanding role of mouse genetics for understanding human biology and disease. Dis Model Mech, 1(1): p. 56-66. 65. Griep, A.E., R. Herber, S. Jeon, J.K. Lohse, R.R. Dubielzig, and P.F. Lambert, 1993, Tumorigenicity by human papillomavirus type 16 E6 and E7 in transgenic mice correlates with alterations in epithelial cell growth and differentiation. J Virol, 67(3): p. 1373-84. 66. Arbeit, J.M., K. Munger, P.M. Howley, and D. Hanahan, 1993, Neuroepithelial carcinomas in mice transgenic with human papillomavirus type 16 E6/E7 ORFs. Am J Pathol, 142(4): p. 1187-97. 67 67. Song, S., H.C. Pitot, and P.F. Lambert, 1999, The human papillomavirus type 16 E6 gene alone is sufficient to induce carcinomas in transgenic animals. J Virol, 73(7): p. 5887-93. 68. Herber, R., A. Liem, H. Pitot, and P.F. Lambert, 1996, Squamous epithelial hyperplasia and carcinoma in mice transgenic for the human papillomavirus type 16 E7 oncogene. J Virol, 70(3): p. 1873-81. 69. Genther Williams, S.M., G.L. Disbrow, R. Schlegel, D. Lee, D.W. Threadgill, and P.F. Lambert, 2005, Requirement of epidermal growth factor receptor for hyperplasia induced by E5, a high-risk human papillomavirus oncogene. Cancer Res, 65(15): p. 653442. 70. Maufort, J.P., A. Shai, H.C. Pitot, and P.F. Lambert, 2010, A role for HPV16 E5 in cervical carcinogenesis. Cancer Res, 70(7): p. 2924-31. 71. Jabbar, S., K. Strati, M.K. Shin, H.C. Pitot, and P.F. Lambert, 2010, Human papillomavirus type 16 E6 and E7 oncoproteins act synergistically to cause head and neck cancer in mice. Virology, 407(1): p. 60-7. 72. Stelzer, M.K., H.C. Pitot, A. Liem, J. Schweizer, C. Mahoney, and P.F. Lambert, 2010, A mouse model for human anal cancer. Cancer Prev Res (Phila), 3(12): p. 1534-41. 73. Arbeit, J.M., K. Munger, P.M. Howley, and D. Hanahan, 1994, Progressive squamous epithelial neoplasia in K14-human papillomavirus type 16 transgenic mice. J Virol, 68(7): p. 4358-68. 74. Coussens, L.M., D. Hanahan, and J.M. Arbeit, 1996, Genetic predisposition and parameters of malignant progression in K14-HPV16 transgenic mice. Am J Pathol, 149(6): p. 1899-917. 75. Arbeit, J.M., P.M. Howley, and D. Hanahan, 1996, Chronic estrogen-induced cervical and vaginal squamous carcinogenesis in human papillomavirus type 16 transgenic mice. Proc Natl Acad Sci U S A, 93(7): p. 2930-5. 76. Smith-McCune, K., Y.H. Zhu, D. Hanahan, and J. Arbeit, 1997, Cross-species comparison of angiogenesis during the premalignant stages of squamous carcinogenesis in the human cervix and K14-HPV16 transgenic mice. Cancer Res, 57(7): p. 1294-300. 77. Kimple, R.J., P.M. Harari, A.D. Torres, R.Z. Yang, B.J. Soriano, M. Yu, E.A. Armstrong, G.C. Blitzer, M.A. Smith, L.D. Lorenz, D. Lee, D.T. Yang, T.M. McCulloch, G.K. Hartig, and P.F. Lambert, 2013, Development and characterization of HPV-positive and HPV-negative head and neck squamous cell carcinoma tumorgrafts. Clin Cancer Res, 19(4): p. 855-64. 68 78. Ahn, J., J.A. Bishop, B. Akpeng, S.I. Pai, and S.R. Best, 2015, Xenograft model for therapeutic drug testing in recurrent respiratory papillomatosis. Ann Otol Rhinol Laryngol, 124(2): p. 110-5. 79. Scherer, W.F., J.T. Syverton, and G.O. Gey, 1953, Studies on the propagation in vitro of poliomyelitis viruses. IV. Viral multiplication in a stable strain of human malignant epithelial cells (strain HeLa) derived from an epidermoid carcinoma of the cervix. J Exp Med, 97(5): p. 695-710. 80. Capes-Davis, A., G. Theodosopoulos, I. Atkin, H.G. Drexler, A. Kohara, R.A. MacLeod, J.R. Masters, Y. Nakamura, Y.A. Reid, R.R. Reddel, and R.I. Freshney, 2010, Check your cultures! A list of cross-contaminated or misidentified cell lines. Int J Cancer, 127(1): p. 1-8. 81. Pattillo, R.A., R.O. Hussa, M.T. Story, A.C. Ruckert, M.R. Shalaby, and R.F. Mattingly, 1977, Tumor antigen and human chorionic gonadotropin in CaSki cells: a new epidermoid cervical cancer cell line. Science, 196(4297): p. 1456-8. 82. Friedl, F., I. Kimura, T. Osato, and Y. Ito, 1970, Studies on a new human cell line (SiHa) derived from carcinoma of uterus. I. Its establishment and morphology. Proc Soc Exp Biol Med, 135(2): p. 543-5. 83. Boshart, M., L. Gissmann, H. Ikenberg, A. Kleinheinz, W. Scheurlen, and H. zur Hausen, 1984, A new type of papillomavirus DNA, its presence in genital cancer biopsies and in cell lines derived from cervical cancer. EMBO J, 3(5): p. 1151-7. 84. Yee, C., I. Krishnan-Hewlett, C.C. Baker, R. Schlegel, and P.M. Howley, 1985, Presence and expression of human papillomavirus sequences in human cervical carcinoma cell lines. Am J Pathol, 119(3): p. 361-6. 85. Nakamura, T., L. Williams-Simons, and H. Westphal, 1997, A human papillomavirus type 18 E6/E7 transgene sensitizes mouse lens cells to human wild-type p53-mediated apoptosis. Oncogene, 14(25): p. 2991-8. 86. Leykauf, K., K. Kabsch, N. Gassler, L. Gissmann, A. Alonso, and J. Schenkel, 2008, Expression of the HPV11 E2 gene in transgenic mice does not result in alterations of the phenotypic pattern. Transgenic Res, 17(1): p. 1-8. 87. Schenkel, J., H. Gaissert, E.E. Protopapa, H. Weiher, L. Gissmann, and A. Alonso, 1999, The human papillomavirus type 11 upstream regulatory region triggers hair-follicle-specific gene expression in transgenic mice. J Invest Dermatol, 112(6): p. 8938. 88. Cid, A., P. Auewarakul, A. Garcia-Carranca, R. Ovseiovich, H. Gaissert, and L. Gissmann, 1993, Cell-type-specific activity of the human papillomavirus type 18 upstream regulatory region in transgenic mice and its modulation by tetradecanoyl phorbol acetate and glucocorticoids. J Virol, 67(11): p. 6742-52. 69 89. Pfefferle, R., G.P. Marcuzzi, B. Akgul, H.U. Kasper, F. Schulze, I. Haase, C. Wickenhauser, and H. Pfister, 2008, The human papillomavirus type 8 E2 protein induces skin tumors in transgenic mice. J Invest Dermatol, 128(9): p. 2310-5. 90. Ghim, S., A.B. Jenson, J.A. Bubier, K.A. Silva, R.S. Smith, and J.P. Sundberg, 2008, Cataracts in transgenic mice caused by a human papillomavirus type 18 E7 oncogene driven by KRT1-14. Exp Mol Pathol, 85(2): p. 77-82. 91. Schaper, I.D., G.P. Marcuzzi, S.J. Weissenborn, H.U. Kasper, V. Dries, N. Smyth, P. Fuchs, and H. Pfister, 2005, Development of skin tumors in mice transgenic for early genes of human papillomavirus type 8. Cancer Res, 65(4): p. 1394-400. 92. Daniel, D., N. Meyer-Morse, E.K. Bergsland, K. Dehne, L.M. Coussens, and D. Hanahan, 2003, Immune enhancement of skin carcinogenesis by CD4+ T cells. J Exp Med, 197(8): p. 1017-28. 93. de Visser, K.E., L.V. Korets, and L.M. Coussens, 2005, De novo carcinogenesis promoted by chronic inflammation is B lymphocyte dependent. Cancer Cell, 7(5): p. 411-23. 94. Coussens, L.M., W.W. Raymond, G. Bergers, M. Laig-Webster, O. Behrendtsen, Z. Werb, G.H. Caughey, and D. Hanahan, 1999, Inflammatory mast cells up-regulate angiogenesis during squamous epithelial carcinogenesis. Genes Dev, 13(11): p. 1382-97. 95. Coussens, L.M., C.L. Tinkle, D. Hanahan, and Z. Werb, 2000, MMP-9 supplied by bone marrow-derived cells contributes to skin carcinogenesis. Cell, 103(3): p. 481-90. 96. van Kempen, L.C., J.S. Rhee, K. Dehne, J. Lee, D.R. Edwards, and L.M. Coussens, 2002, Epithelial carcinogenesis: dynamic interplay between neoplastic cells and their microenvironment. Differentiation, 70(9-10): p. 610-23. 97. Rhee, J.S., R. Diaz, L. Korets, J.G. Hodgson, and L.M. Coussens, 2004, TIMP-1 alters susceptibility to carcinogenesis. Cancer Res, 64(3): p. 952-61. 98. de Visser, K.E., L.V. Korets, and L.M. Coussens, 2004, Early neoplastic progression is complement independent. Neoplasia, 6(6): p. 768-76. 99. Andreu, P., M. Johansson, N.I. Affara, F. Pucci, T. Tan, S. Junankar, L. Korets, J. Lam, D. Tawfik, D.G. DeNardo, L. Naldini, K.E. de Visser, M. De Palma, and L.M. Coussens, 2010, FcRgamma activation regulates inflammation-associated squamous carcinogenesis. Cancer Cell, 17(2): p. 121-34. 100. Li, J.J., Y. Cao, M.R. Young, and N.H. Colburn, 2000, Induced expression of dominant-negative c-jun downregulates NFkappaB and AP-1 target genes and suppresses tumor phenotype in human keratinocytes. Mol Carcinog, 29(3): p. 159-69. 70 101. Matthews, C.P., A.M. Birkholz, A.R. Baker, C.M. Perella, G.R. Beck, Jr., M.R. Young, and N.H. Colburn, 2007, Dominant-negative activator protein 1 (TAM67) targets cyclooxygenase-2 and osteopontin under conditions in which it specifically inhibits tumorigenesis. Cancer Res, 67(6): p. 2430-8. 102. Borsi, E., C. Terragna, A. Brioli, P. Tacchetti, M. Martello, and M. Cavo, 2014, Therapeutic targeting of hypoxia and hypoxia-inducible factor 1alpha in multiple myeloma. Transl Res. 103. Tang, X., Q. Zhang, J. Nishitani, J. Brown, S. Shi, and A.D. Le, 2007, Overexpression of human papillomavirus type 16 oncoproteins enhances hypoxiainducible factor 1 alpha protein accumulation and vascular endothelial growth factor expression in human cervical carcinoma cells. Clin Cancer Res, 13(9): p. 2568-76. 104. Li, G., L. He, E. Zhang, J. Shi, Q. Zhang, A.D. Le, K. Zhou, and X. Tang, 2011, Overexpression of human papillomavirus (HPV) type 16 oncoproteins promotes angiogenesis via enhancing HIF-1alpha and VEGF expression in non-small cell lung cancer cells. Cancer Lett, 311(2): p. 160-70. 105. Elson, D.A., H.E. Ryan, J.W. Snow, R. Johnson, and J.M. Arbeit, 2000, Coordinate up-regulation of hypoxia inducible factor (HIF)-1alpha and HIF-1 target genes during multi-stage epidermal carcinogenesis and wound healing. Cancer Res, 60(21): p. 6189-95. 106. Lu, Z.H., J.D. Wright, B. Belt, R.D. Cardiff, and J.M. Arbeit, 2007, Hypoxiainducible factor-1 facilitates cervical cancer progression in human papillomavirus type 16 transgenic mice. Am J Pathol, 171(2): p. 667-81. 107. Guo, W. and F.G. Giancotti, 2004, Integrin signalling during tumour progression. Nat Rev Mol Cell Biol, 5(10): p. 816-26. 108. Tran, T., B. Barlow, L. O'Rear, B. Jarvis, Z. Li, K. Dickeson, W. Dupont, and M. Zutter, 2011, Loss of the alpha2beta1 integrin alters human papilloma virus-induced squamous carcinoma progression in vivo and in vitro. PLoS One, 6(10): p. e26858. 109. Mohamed, M.M. and B.F. Sloane, 2006, Cysteine cathepsins: multifunctional enzymes in cancer. Nat Rev Cancer, 6(10): p. 764-75. 110. Dennemarker, J., T. Lohmuller, J. Mayerle, M. Tacke, M.M. Lerch, L.M. Coussens, C. Peters, and T. Reinheckel, 2010, Deficiency for the cysteine protease cathepsin L promotes tumor progression in mouse epidermis. Oncogene, 29(11): p. 161121. 111. Yu, W., J. Liu, M.A. Shi, J. Wang, M. Xiang, S. Kitamoto, B. Wang, G.K. Sukhova, G.F. Murphy, G. Orasanu, A. Grubb, and G.P. Shi, 2010, Cystatin C deficiency promotes epidermal dysplasia in K14-HPV16 transgenic mice. PLoS One, 5(11): p. e13973. 71 112. Shahrabi-Farahani, S., L. Wang, B.M. Zwaans, J.M. Santana, A. Shimizu, S. Takashima, M. Kreuter, L. Coultas, P.A. D'Amore, J.M. Arbeit, L.A. Akslen, and D.R. Bielenberg, 2014, Neuropilin 1 expression correlates with differentiation status of epidermal cells and cutaneous squamous cell carcinomas. Lab Invest, 94(7): p. 752-65. 113. Paiva, I., R.M. Gil da Costa, J. Ribeiro, H. Sousa, M. Bastos, A.F. Rocha, P.A. Oliveira, and R. Medeiros, 2015, A Role for MicroRNA-155 Expression in Microenvironment Associated to HPV-Induced Carcinogenesis in K14-HPV16 Transgenic Mice. PLoS One, 10(1): p. e0116868. 114. Paiva, I., R.M. Gil da Costa, J. Ribeiro, H. Sousa, M.M. Bastos, A. FaustinoRocha, C. Lopes, P.A. Oliveira, and R. Medeiros, 2015, MicroRNA-21 expression and susceptibility to HPV-induced carcinogenesis - role of microenvironment in K14-HPV16 mice model. Life Sci, 128: p. 8-14. 115. Senba, M. and N. Mori, 2012, Mechanisms of virus immune evasion lead to development from chronic inflammation to cancer formation associated with human papillomavirus infection. Oncol Rev, 6(2): p. e17. 116. Mescher, M.F., J.M. Curtsinger, P. Agarwal, K.A. Casey, M. Gerner, C.D. Hammerbeck, F. Popescu, and Z. Xiao, 2006, Signals required for programming effector and memory development by CD8+ T cells. Immunol Rev, 211: p. 81-92. 117. van Montfoort, N., E. van der Aa, and A.M. Woltman, 2014, Understanding MHC class I presentation of viral antigens by human dendritic cells as a basis for rational design of therapeutic vaccines. Front Immunol, 5: p. 182. 118. Dunn, G.P., L.J. Old, and R.D. Schreiber, 2004, The three Es of cancer immunoediting. Annu Rev Immunol, 22: p. 329-60. 119. Trapani, J.A. and M.J. Smyth, 2002, Functional significance of the perforin/granzyme cell death pathway. Nat Rev Immunol, 2(10): p. 735-47. 120. Buferne, M., L. Chasson, M. Grange, A. Mas, F. Arnoux, M. Bertuzzi, P. Naquet, L. Leserman, A.M. Schmitt-Verhulst, and N. Auphan-Anezin, 2015, IFNgamma producing CD8 T cells modified to resist major immune checkpoints induce regression of MHC class I-deficient melanomas. Oncoimmunology, 4(2): p. e974959. 121. Qiu, Z., H. Huang, J.M. Grenier, O.A. Perez, H.M. Smilowitz, B. Adler, and K.M. Khanna, 2015, Cytomegalovirus-Based Vaccine Expressing a Modified Tumor Antigen Induces Potent Tumor-Specific CD8(+) T-cell Response and Protects Mice from Melanoma. Cancer Immunol Res, 3(5): p. 536-46. 122. Belapurkar, P., P. Goyal, and P. Tiwari-Barua, 2014, Immunomodulatory effects of triphala and its individual constituents: a review. Indian J Pharm Sci, 76(6): p. 467-75. 72 123. Gil da Costa, R.M., M.M. Bastos, P.A. Oliveira, and C. Lopes, 2012, Bracken-associated human and animal health hazards: chemical, biological and pathological evidence. J Hazard Mater, 203-204: p. 1-12. 124. Chang, F., S. Syrjanen, Q. Shen, L. Wang, D. Wang, and K. Syrjanen, 1992, Human papillomavirus involvement in esophageal precancerous lesions and squamous cell carcinomas as evidenced by microscopy and different DNA techniques. Scand J Gastroenterol, 27(7): p. 553-63. 125. Hirono, I., Y. Kono, K. Takahashi, K. Yamada, H. Niwa, M. Ojika, H. Kigoshi, K. Niiyama, and Y. Uosaki, 1984, Reproduction of acute bracken poisoning in a calf with ptaquiloside, a bracken constituent. Vet Rec, 115(15): p. 375-8. 126. Latorre, A.O., B.D. Caniceiro, H.L. Wysocki, Jr., M. Haraguchi, D.R. Gardner, and S.L. Gorniak, 2011, Selenium reverses Pteridium aquilinum-induced immunotoxic effects. Food Chem Toxicol, 49(2): p. 464-70. 127. Latorre, A.O., M.S. Furlan, M. Sakai, H. Fukumasu, I.M. Hueza, M. Haraguchi, and S.L. Gorniak, 2009, Immunomodulatory effects of Pteridium aquilinum on natural killer cell activity and select aspects of the cellular immune response of mice. J Immunotoxicol, 6(2): p. 104-14. 128. Gil da Costa, R.M., P.A. Oliveira, M. Vilanova, M.M. Bastos, C.C. Lopes, and C. Lopes, 2011, Ptaquiloside-induced, B-cell lymphoproliferative and early-stage urothelial lesions in mice. Toxicon, 58(6-7): p. 543-9. 129. Adefuye, A. and K. Sales, 2012, Regulation of inflammatory pathways in cancer and infectious disease of the cervix. Scientifica (Cairo), 2012: p. 548150. 130. Wolff, H., K. Saukkonen, S. Anttila, A. Karjalainen, H. Vainio, and A. Ristimaki, 1998, Expression of cyclooxygenase-2 in human lung carcinoma. Cancer Res, 58(22): p. 4997-5001. 131. Sano, H., Y. Kawahito, R.L. Wilder, A. Hashiramoto, S. Mukai, K. Asai, S. Kimura, H. Kato, M. Kondo, and T. Hla, 1995, Expression of cyclooxygenase-1 and -2 in human colorectal cancer. Cancer Res, 55(17): p. 3785-9. 132. Ristimaki, A., N. Honkanen, H. Jankala, P. Sipponen, and M. Harkonen, 1997, Expression of cyclooxygenase-2 in human gastric carcinoma. Cancer Res, 57(7): p. 1276-80. 133. Golijanin, D., J.Y. Tan, A. Kazior, E.G. Cohen, P. Russo, G. Dalbagni, K.J. Auborn, K. Subbaramaiah, and A.J. Dannenberg, 2004, Cyclooxygenase-2 and microsomal prostaglandin E synthase-1 are overexpressed in squamous cell carcinoma of the penis. Clin Cancer Res, 10(3): p. 1024-31. 73 134. Ryu, H.S., K.H. Chang, H.W. Yang, M.S. Kim, H.C. Kwon, and K.S. Oh, 2000, High cyclooxygenase-2 expression in stage IB cervical cancer with lymph node metastasis or parametrial invasion. Gynecol Oncol, 76(3): p. 320-5. 135. McAdam, B.F., F. Catella-Lawson, I.A. Mardini, S. Kapoor, J.A. Lawson, and G.A. FitzGerald, 1999, Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)- 2: the human pharmacology of a selective inhibitor of COX-2. Proc Natl Acad Sci U S A, 96(1): p. 272-7. 136. Hahn, T., I. Alvarez, J.J. Kobie, L. Ramanathapuram, S. Dial, A. Fulton, D. Besselsen, E. Walker, and E.T. Akporiaye, 2006, Short-term dietary administration of celecoxib enhances the efficacy of tumor lysate-pulsed dendritic cell vaccines in treating murine breast cancer. Int J Cancer, 118(9): p. 2220-31. 137. Hermann, M., S. Shaw, E. Kiss, G. Camici, N. Buhler, R. Chenevard, T.F. Luscher, H.J. Grone, and F. Ruschitzka, 2005, Selective COX-2 inhibitors and renal injury in salt-sensitive hypertension. Hypertension, 45(2): p. 193-7. 138. Kosaka, A., T. Ohkuri, and H. Okada, 2014, Combination of an agonistic anti-CD40 monoclonal antibody and the COX-2 inhibitor celecoxib induces anti-glioma effects by promotion of type-1 immunity in myeloid cells and T-cells. Cancer Immunol Immunother, 63(8): p. 847-57. 139. Kvale, D., V. Ormaasen, A.M. Kran, C.C. Johansson, P. Aukrust, E.M. Aandahl, S.S. Froland, and K. Tasken, 2006, Immune modulatory effects of cyclooxygenase type 2 inhibitors in HIV patients on combination antiretroviral treatment. AIDS, 20(6): p. 813-20. 140. Pettersen, F.O., E.A. Torheim, A.E. Dahm, I.S. Aaberge, A. Lind, M. Holm, E.M. Aandahl, P.M. Sandset, K. Tasken, and D. Kvale, 2011, An exploratory trial of cyclooxygenase type 2 inhibitor in HIV-1 infection: downregulated immune activation and improved T cell-dependent vaccine responses. J Virol, 85(13): p. 6557-66. 141. Bzhalava, D., P. Guan, S. Franceschi, J. Dillner, and G. Clifford, 2013, A systematic review of the prevalence of mucosal and cutaneous human papillomavirus types. Virology, 445(1-2): p. 224-31. 142. Hoots, B.E., J.M. Palefsky, J.M. Pimenta, and J.S. Smith, 2009, Human papillomavirus type distribution in anal cancer and anal intraepithelial lesions. Int J Cancer, 124(10): p. 2375-83. 143. Balkwill, F., K.A. Charles, and A. Mantovani, 2005, Smoldering and polarized inflammation in the initiation and promotion of malignant disease. Cancer Cell, 7(3): p. 211-7.