Interferon-Alpha Immunotherapy of Melanoma: Signal Transduction, Gene Transcription, and the Role of Suppressor of Cytokine Signaling Proteins in Immune Cells

INTERFERON-ALPHA IMMUNOTHERAPY OF MELANOMA: SIGNAL TRANSDUCTION, GENE TRANSCRIPTION, AND THE ROLE OF SUPPRESSOR OF CYTOKINE SIGNALING PROTEINS IN IMMUNE CELLS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Jason M. Zimmerer, B.S.

*****

The Ohio State University 2007

Dissertation Committee: Approved by William E. Carson, III, M.D., Advisor

Virginia Sanders, Ph.D. ______Anne VanBuskirk, Ph.D. Advisor Integrated Biomedical Sciences Mathew Ringel, M.D. Graduate Program

ABSTRACT

High dose interferon-alpha-2b (IFN-α-2b) is employed as an adjuvant therapy in

melanoma patients who have undergone surgical resection of high-risk lesions. The

precise molecular targets of IFN-α therapy are unknown, but likely involve signal

transducer and activator of transcription 1 (STAT1) signal transduction within host

immune effector cells. We hypothesized that STAT1-mediated signaling induces

molecular targets important for mediating the anti-tumor effect of exogenously

administered IFN-α. To identify the STAT1-dependent genes, microarray technology

was utilized to characterize the gene expression profile of splenocytes from wild type

(WT) and STAT1-/- mice stimulated with IFN-α. Analysis showed that 25 genes required

STAT1 signal transduction for optimal expression in response to IFN-α (i.e. Gzmb,

Isg20, Ly6c; p < 0.001). Interestingly, human immune cells are also capable of inducing

the homologues of these genes in response to IFN-α. Human PBMCs, CD3+ T cells,

CD56+ natural killer (NK) cells, and CD14+ monocytes each exhibited a distinct and

reproducible transcriptional profile following stimulation with IFN-α by microarray

analysis. Analysis of gene expression within PBMCs from melanoma patients (n = 7)

receiving high-dose IFN-α-2b (20 MU/m2 i.v.) also revealed significant upregulation of

23 genes (including IRF7, TAP1, TNFRSF6, USP18; p < 0.001). A comparison of the

gene expression profile of in vitro IFN-α-stimulated PBMCs from normal donor cells

ii with that of PBMCs obtained from melanoma patients receiving intravenous IFN-α

revealed little overlap. However, within individual patients, the gene expression profile of in vitro IFN-α-stimulated PBMCs was largely predictive of the expression profile of gene expression following IFN-α administration. This indicated that an individual

patient response to IFN-α immunotherapy could be predicted prior to treatment.

The phosphorylation of STAT1 (P-STAT1) within effector cells is highly variable

among melanoma patients following IFN-α immunotherapy and at high doses of IFN-α

P-STAT1 is down-regulated. We hypothesized that high doses of IFN-α administration would yield suboptimal levels of signal transduction and gene transcription and the induction of inhibitors by IFN-α may be the cause. PBMCs collected from metastatic melanoma receiving escalating doses of IFN-α-2b (5 MU/m2 and then 10 MU/m2) exhibited statistically equivalent levels of P-STAT1, P-STAT2 and the induction of interferon stimulated gene (ISG) transcripts. In addition, suppressors of cytokine signaling (SOCS) protein 1 and SOCS3 were induced to a greater degree with the higher dose of IFN-α. This suggests that a negative feedback loop is activated, thus inhibiting the effect of the higher dose. In fact, SOCS1 and SOCS3 are involved in the negative regulation of IFN-α activity including STAT1 activation and ISG transcription in vitro.

The loss of function of SOCS1 and SOCS3 reveals that the anti-tumor effects of IFN-α can be enhanced in the mouse model of malignant melanoma. Impressively, IFN-α treatment eliminated lethal inoculums of melanoma in 70% of SOCS1-deficient mice, whereas all IFN-treated SOCS1-competent mice died. The anti-tumor effects of IFN-α in

iii tumor-bearing SOCS1-deficient mice were markedly inhibited following depletion of

CD8+ T cells. These results indicate that the anti-tumor response of immune effector cells to exogenous IFN-α is regulated by SOCS proteins. Additionally, we show that

SOCS1 was an important regulator of immunosurveillance of developing cancer as

SOCS1-deficient mice were protected from tumor formation.

These reports are the first to characterize the dependence of STAT1 in mediating the transcriptional response of immune cells to IFN-α. In addition, the studies remain the first to define the transcriptional response of immune subsets to IFN-α and to characterize the transcriptional profiles of PBMCs from melanoma patients undergoing

IFN-α immunotherapy. We also determined that microarray analysis of patient PBMCs following in vitro stimulation with IFN-α may be a useful predictor of the individual response to IFN-α immunotherapy. Finally, the optimization of the dose of IFN-α administration and the modulation of SOCS activity may allow for a greater efficacy of

IFN-α immunotherapy.

Dedicated to the women of my life, both past and present: Jennifer, Christine, and

Patricia. You are my life, my guide, and my motivation.

ACKNOWLEDGMENTS

This work would not have been possible were it not for the support of numerous

individuals. First, I would like to thank my mentor William Carson. Dr. Carson has been

a tremendous motivator these last four years and I feel privileged to have worked in his

laboratory. He has been not only a caring physician and a diligent investigator, but a

dedicated teacher to young researchers.

I would also like to thank the members of Carson lab and collaborators for their

support and friendship. Specifically, I would like to recognize Dr. Gregory Lesinski, Sri

Vidya Kondadasula, Volodymyr Karpa, Dr. Brian Becknell, Amy Lehman, Amy

Ruppert, Dr. Michael Radmacher, Dr. James Ihle, Dr. Sushella Tridandapani, Dr.

Michael Caligiuri, Dr. Thomas Olencki, Dr. Kari Kendra, Dr. Michael Walker, and the

CCC Core laboratories. Additionally, I would like to offer my genuine appreciation to

the members of my graduate dissertation committee, namely Virginia Sanders, Anne

Vanbuskirk, and Mathew Ringel for their time and advice.

Finally, I would like to recognize my family for their love and encouragement. I would like to thank my parents, Chris and Michael, for their dedication to my upbringing and their continuous stress of academics. I only hope that I will raise my children with

the same values. Most of all I want to recognize my wife, Jennifer. She has supported

me with love, understanding, and patience during the long hours spent apart due to my

time in the laboratory. She is my center, my heart, my home.

VITA

November 2, 1979...... Born – Dayton, Ohio.

2002...... Bachelors of Science- Biochemistry Ohio Northern University, Ada, OH.

2002 – present...... Ph.D. Candidate, Integrated Biomedical Sciences Graduate Program, The Ohio State University

PUBLICATIONS

Research Publications

1. Lesinski, G.B., M. Anghelina, J. Zimmerer, T. Bakalakos, B. Badgwell, R. Parihar, Y. Hu, G. Abood, C. Magro, J. Durbin and W.E. Carson. The anti-tumor effects of interferon-alpha are abrogated in a STAT1-deficient mouse. Journal of Clinical Investigation 112(2): 170-180, 2003.

2. Lesinski, G.B., B. Badgwell, J. Zimmerer, T. Crespin, Y. Hu, G. Abood and W.E. Carson. Interleukin-12 pre-treatments enhance interferon-alpha-induced Jak-STAT signaling and potentiate the anti-tumor effects of interferon-alpha in a murine model of malignant melanoma. Journal of Immunology 172(12):7368-76, 2004.

FIELDS OF STUDY

Major Field: Integrated Biomedical Sciences Graduate Program

vii

TABLE OF CONTENTS

P a g e

Abstract...... ii

Dedication...... v

Acknowledgments ...... vi

Vita ...... vii

List of Tables...... x

List of Figures ...... xi

List of Abbreviations ...... xiii

Chapters:

1. Introduction...…………………………………………...…………………...... 1

1.1 Interferon-alpha……….....……………………………………...1 1.2 Melanoma…….....……….……………………..………...... 3 1.3 Treatment of melanoma...... 5 1.4 Potential anti-tumor mechanisms of IFN-α…….……….....…....8 1.5 Suppressors of cytokine signaling proteins…….………………10 1.6 Hypothesis………..……………...….………….………………11

2. The role of STAT1 in mediating the transcription of interferon stimulated genes in murine immune cells……………………………………………………………...... 13

2.1 Introduction……………………………..……………………..13 2.2 Materials and Methods………………………………………...15 2.3 Results……………………………………..…………………..18 2.4 Discussion………………………………..……………………19 2.5 Tables and Figures…...…………………..……………………22 3. Gene expression profiling of the immune response to interferon-alpha…...... 28

viii

3.1 Introduction……………………….……………………...…...28 3.2 Materials and Methods……………………...………………...30 3.3 Results.………………………………………………………..34 3.4 Discussion.……………………………………………………39 3.5 Tables and Figures…..…………………..……………………44

4. Interferon-alpha-2b induced signal transduction and gene regulation in patient PBMCs is not enhanced by a dose increase from 5 MU/m2 10 MU/m2……………..…..58

4.1 Introduction………………………………………………...... 58 4.2 Materials and Methods……….….…………………………...60 4.3 Results………………………………………………………..63 4.4 Discussion……………………………………………………67 4.5 Tables and Figures….…………………..……………………71

5. IFN-alpha-induced signal transduction, gene expression, and anti-tumor activity of immune effector cells are negatively regulated by suppressor of cytokine signaling proteins…………………………………………………………………………………...86

5.1 Introduction………………………..…….………………...... 86 5.2 Materials and Methods………………………….…………...89 5.3 Results……………………….…….……………….………..95 5.4 Discussion………………….…….……………..….………104 5.5 Tables and Figures……………………….………..….……109

6. Conclusion……………………………………..…………...…………………...142

Bibliography……………………………….……………………………………………148

LIST OF TABLES

Table Page

2.1 STAT1-Enhanced Genes...... 22

2.2 STAT1-Suppressed Genes...... 23

2.3 STAT1-Independent Genes...... 24

3.1 Gene Regulation in PBMCs following 1hr IFN-α treatment ...... 44

3.2 Gene Regulation in PBMCs following 18hr IFN-α treatment...... 47

3.3 Gene Regulation in T cells following 18hr IFN-α treatment ...... 48

3.4 Gene Regulation in NK cells following 18hr IFN-α treatment...... 49

3.5 Gene Regulation in Monocytes following 18hr IFN-α treatment...... 50

3.6 Gene Down-regulation in Monocytes following 18hr IFN-α treatment .51

3.7 Gene Upregulation 1hr following 20 MU/m2 IFN-α-2b ...... 55

4.1 Linear Response and Non-linear Dose Response in patient PBMCs . . 81

LIST OF FIGURES

Figure Page

2.1 Real Time PCR Validation of Murine Microarray Data ...... 26

3.1 Real Time PCR analysis of select genes identified by microarray analysis of PBMCs following 1 hour in vitro IFN-α stimulation...... 46

3.2 Real Time PCR analysis of select genes identified by microarray analysis of immune subsets following 18 hour in vitro IFN-α stimulation ...... 53

3.3 Real Time PCR analysis of select genes identified by microarray analysis of PBMCs from melanoma patients receiving IFN-α...... 57

4.1 Native and activated forms of STAT1 and STAT2 levels in PBMCs following in vivo escalation of IFN-α-2b administration...... 72

4.2 Ratios of native and activated forms of STAT1 and STAT2 levels in PBMCs following in vivo escalation of IFN-α-2b administration...... 75

4.3 Interferon-stimulated gene expression following IFN-α-2b administration...... 77

4.4 SOCS1 and SOCS3 expression following IFN-α-2b administration.. . . 80

4.5 Real Time PCR Validation of Microarray Data from PBMCs of Patients receiving escalation doses of IFN-α...... 84

5.1 SOCS transcripts are rapidly induced in PBMCs following IFN-α stimulation...... 111

5.2 SOCS transcripts are differentially induced in NK cells and T cells following IFN-α stimulation...... 116

5.3 Differential SOCS expression and STAT1 activation in melanoma patients undergoing IFN-α immunotherapy...... 120 xi

5.4 Over-expression of SOCS1 and SOCS3 protein in Jurkat cells inhibits the response to IFN-α...... 123

5.5 SOCS1- and SOCS3-deficient mice exhibit an augmented response to IFN-α...... 128

5.6 Small inhibitory RNA (siRNA) mediated inhibition of SOCS1 and SOCS3 augments IFN-α-responsiveness in vitro...... 133

5.7 SOCS-deficiency enhances the anti-tumor effect of IFN-A/D in a murine model of malignant melanoma...... 136

5.8 SOCS1-deficiency enhances immunosurveillance of developing melanoma tumors...... 144

xii

LIST OF ABBREVIATIONS

µg Micro-gram

Ab Antibody

ANOVA Analysis of variance

CIS Cytokine inducible SH2-containing protein cRNA Complementary ribonucleic acid

DC Dendritic cell

IFN-α Interferon-alpha

IFN-γ Interferon-gamma i.p. Intraperitoneal

ISG Interferon stimulated gene i.v. Intravenous

Jak Janus kinase

MU/m2 Million units per meters squared

NK Natural killer

P- Phosphorylated

PBMCs Peripheral blood mononuclear cells

PCR Polymerase chain reaction s.c. Subcutaneous

xiii SOCS Suppressors of cytokine signaling

STAT Signal transducer and activator of transcription

WT Wild type

xiv

CHAPTER 1

INTRODUCTION

1.1 Interferon-alpha

Interferon (IFN) was discovered approximately fifty years ago as an antiviral

substance due to its ability to inhibit viral replication (1). Today the interferon family of

proteins includes two main classes of related cytokines: type I IFN and type II IFN. A

sole member makes up the type II IFN that is called IFN-γ (gamma) which binds to the

IFN-γ receptor (IFNGR) complex to elicit a signal within its target cell. Human type I

IFNs comprise a vast group of IFN proteins, designated IFN-α (alpha), IFN-β (beta),

IFN-κ (kappa), IFN-ε (epsilon), IFN-ω (omega), and IFN-ζ (zeta) (2). The IFN-α

proteins can be further divided into 13 subtypes that are called IFNA1, IFNA2, IFNA4,

IFNA5, IFNA6, IFNA7, IFNA8, IFNA10, IFNA13, IFNA14, IFNA16, IFNA17,

IFNA21 (3). All type I IFNs bind to a specific cell surface receptor complex known as

the IFN-α receptor (IFNAR) that consists of IFNAR1 and IFNAR2 chains. Although

Type I IFNs all generally function as antiviral, anti-proliferative, and immunostimulatory

cytokines, the biological significance of having many Type I IFNs is unclear, specifically

the many subtypes of IFN-α (4). However, it has been shown that the various IFN-α

subtypes can interact with the interferon receptor components in different ways and can

activate a number of different signaling pathways (5). It has also been shown that there 1 are large differences in the ability of individual IFN-α proteins to activate natural killer

cells, with IFN-α17 completely lacking this capability (6). Yet, the clinical implications

of these differences remain to be discovered (4).

The receptor for IFN-α is widely expressed on both tumor cells and immune

effector cells and mediates most of its effects via activation of the Janus kinase

(Jak)-signal transducers and activators of transcription (STAT) pathway (7, 8). Upon

binding to its receptor, IFN-α activates Jak1 and tyrosine kinase 2 (Tyk2), which in turn phosphorylate tyrosine residues within the cytoplasmic region of the receptor. These phosphotyrosine residues provide binding sites for STAT1 and STAT2 proteins, latent cytoplasmic transcription factors that are phosphorylated by the Janus kinases (9). The prototypical IFN-α signaling reaction results in the formation of IFN-stimulated gene factor 3 (ISGF3), a DNA-binding complex that consists of STAT1α (or STAT1β),

STAT2, and interferon regulatory factor 9 (IRF9) (10). ISGF3 subsequently translocates

to the nucleus and binds to interferon-stimulated response elements (IRSEs) located in

the promoter regions of IFN-stimulated genes (ISGs) (11). IFN-α can also induce STAT

dimers capable of binding to the IFN-γ-activated site (GAS) element which is also

present in the promoters of ISGs (12). Of the known ISGs, some have only IRSEs or

only GAS elements in their promoters, whereas others have both elements suggesting

multiple STAT complexes might be required for optimal gene expression (13). These

signaling events induce the expression of a variety of immunoregulatory genes and

largely determine the pattern of immune cell activation following exposure to

IFN-α (14-18). 2 A 1980 study by Bart et al. demonstrated promising anti-melanoma effects of

IFN-α in vitro on the murine B16 melanoma cell line (19). These findings led

investigators to study the anti-tumor effects of this cytokine in humans. Since then, the

Eastern Cooperative Oncology Group has conducted several clinical trials on the efficacy

of IFN-α-2b as an adjuvant treatment for high-risk melanoma patients (20). This led to

the FDA approval of high dose IFN-α-2b for the adjuvant treatment of melanoma

patients presenting with thick lesions or node-positive disease. Other IFNs tested to date,

including IFN-β have not proven to be effective (21).

1.2 Melanoma

The skin has 3 layers called the epidermis, dermis, and hypodermis. The top layer, the epidermis, is susceptible to DNA damage that may result in tumor genesis. Of all cancers, skin cancers are the most common and consist nearly 50% of all cancer cases

(22). For most of the skin cancer subtypes the disease is rarely lethal. However, melanoma, a less common form of skin cancer, is a very aggressive and deadly disease.

Melanoma arises from the malignant transformation of melanocytes which are present in the basal layer of the epidermis and produce melanin to protect the skin from ultraviolet radiation.

The American Cancer Society estimated that about 62,190 new melanomas would be diagnosed in the United States in 2006. The number of new melanomas diagnosed in the United States is increasing at a rate faster than any other cancer (23). Among white 3 men and women in the United States, incidence rates for melanoma are about 3% per

year (24). About 7,900 people in the United States were projected to die of melanoma during 2006 and the mortality rate has increased by 50% since 1973 (25). Much of this increase has been in the elderly population, mostly white men. More recently, the death rate from melanoma has leveled off for men and dropped slightly in women.

Although melanoma formation has been linked to ultraviolet radiation exposure,

other risk factors are important. These include fair skin, the presence of numerous nevi

(moles), dysplastic nevi, previous melanoma lesions (recurrence), immune deficiencies,

and family history (26). Around 10% of people with melanoma have a family history of

melanoma which can be related to either family lifestyle or genetic mutations (27). In

families with inherited melanomas, gene mutations that substantially increase the risk of

the development of melanoma are passed from one generation to the next. The genes

most commonly responsible for familial melanoma are involved with the cell cycle check

point (G1) to DNA synthesis (S phase). These genes include the oncogene cyclin

dependent kinase 4 (CDK4) and two tumor suppressors known as p19ARF and p16ink4a

(28, 29). The mutations prevent these genes from doing their normal job of controlling the growth of the cell, leading to cancer genesis. In addition, there are many gene mutations that are not inherited. These include genes such as BRAF, NRAS, NEDD9, and PTEN although few mutations occur in a significant number of melanoma cases (30-

35).

4 Once melanoma has formed, the disease has the potential to become malignant

and invade through the skin into circulation. The most common metastases include lymph node, lung, liver, brain, and intestine (36-39). Melanoma that has escaped confinement of the skin is virtually untreatable and patients die of lesions of the lung or brain, or obstruction of the bowel (40). Due to the aggressive behavior and poor prognosis of melanoma, much of the focus has been on the prevention of the disease.

Adoptions such as the “ABCDE’s of melanoma,” a monthly self examination that is recommended by the American Skin Cancer Foundation, have helped to raise awareness in order to discover the disease at an early stage. Despite this, melanoma remains a prevalent and increasing problem. Although chemotherapies, immunotherapies, and targeted treatments have been investigated, no single agent or combination treatment has attained substantial and sustained anti-tumor activity for a large population of patients (41).

1.3 Treatment of melanoma

Wide local excision is the treatment of choice for primary malignant melanoma and locoregional recurrence. Surgery remains the only modality where a cure is possible and can be quite successful (42, 43). However, metastatic melanoma is incurable and the median survival, at 6 to 9 months, is dismal (44, 45). Systemic therapy with palliative intent is the mainstay of treatment.

5 Dacarbazine has become the standard chemotherapy for metastatic melanoma,

with a response rate of approximately 10-15%, with median response duration of 5 to

6 months and complete response rates of 5% (46). Responses are usually incomplete and

last only a few months (47). Combing this chemotherapy with other agents results in

higher response rates but with increased toxicity and no prolongation of survival (48). In

addition, targeted therapies such as CTLA-4 antibody, Bevacizumab (VEGF inhibitor), and Sorafenib (BRAF inhibitor) are promising agents but have yet to show significant anti-tumor activity in advanced melanoma (49-51).

Currently, immune based therapies show an intriguing potential for generating response. Unlike other malignancies, spontaneous regressions of disease have been observed on occasion in patients with malignant melanoma (52). Careful histologic evaluations have revealed that this process is mediated by activated lymphocytes (53).

Consequently, much effort has centered on the development of immune-based treatments for the therapy of this malignancy, especially in light of the fact that it is relatively chemoresistant (54). Treatment with high-dose interleukin 2 (IL-2) has resulted in

dramatic clinical outcomes, including significant tumor regression in patients with diffuse

metastatic disease. High-dose IL-2 induced a response rate of 16% with 6-8% complete

response in patients with stage IV melanoma (55, 56). In addition, Dudley et al.

performed a study involving lymphodepleting chemotherapy followed by adoptive transfer of autologous tumor reactive lymphocytes in combination with high dose IL-2 for the treatment of patients with stage IV melanoma. Surprisingly, 51% of the treated patients experienced objective clinical responses including an 8% complete response rate

6 (37). In a separate report, combinational immunotherapy with IFN-α and high dose IL-2

exhibited a response rate of 41% (57). Unfortunately, for many of these studies utilizing

IL-2, its use is associated with a unique spectrum of toxicities that prevent its widespread utilization (58).

Recombinant IFN-α is used as a single agent to treat patients with metastatic malignant melanoma and is associated with an overall response rate of 10-20% (20, 59,

60). IFN-α is also known to promote complete regression of the disease in a small subset of patients who generally represent one third of the total responses (61, 62). High-dose

IFN-α is also employed as an adjuvant in patients who have undergone resection of lesions with high-risk of recurrence (nodal disease or primary tumors of Breslow thickness > 4 mm) and remains the only FDA-approved agent for adjuvant therapy (20,

59, 63-65). Common toxicities of interferon include fever, nausea, leukopenia, fatigue, depression, anorexia, and skin rashes (66-68). However, a study performed by Caraceni et al. showed that although anxiety and fatigue were increased in the group treated with interferon-alpha, no sizable impact was seen to the quality of life measures. In addition, the serious toxicities associated with high-dose IFN-α therapy are well documented and can be avoided if the clinician takes care to follow the precise guidelines that have recently been published (69). Do to the inability of other therapies to show a durable response in melanoma, it is unlikely that new immunotherapeutic options will become available in the near future (70-72).

7 1.4 Potential anti-tumor mechanisms of IFN-α

Tumor cells routinely express functional IFN-α receptors and this cytokine can

exert anti-proliferative, anti-angiogenic, and pro-apoptotic effects on tumor cells in vitro

(73-77). IFN-α can also exert multiple effects on T cells and NK cells that enhance their

ability to recognize and destroy malignant cells, thus linking the innate immune response

with the more sustained adaptive-immune response (78). IFN-α induces NK

cell-mediated cytotoxicity and proliferation in vivo and regulates either directly or

indirectly the activity of numerous chemokines and cytokines (79). IFN-α elicits

lymphocyte-activated killer (LAK) activity in vivo and also exerts a profound effect on

cytotoxic T cell responses by stimulating the proliferation, generation, and activation of

existing memory CD8+ cytotoxic T cells (78-80). IFN-α also strongly upregulates tumor

cell expression of MHC Class I and Class II antigens and adhesion molecules such as

ICAM-1 and L-selectin, which are critical for the recognition and destruction of

malignant cells by cytotoxic T cells (8, 81).

Despite IFN-α’s role in both the direct and indirect anti-tumor activity, there has

been some debate as to which process is more important for the reduction of melanoma

tumor burden (82). We have previously demonstrated that ex vivo treatment of patient tumors with clinically relevant concentrations of IFN-α consistently led to activation of

STAT1 and STAT2 (83). More recently, it was observed that some IFN-α-resistant human melanoma cell lines exhibit defects in specific Jak-STAT intermediates, which

8 when reversed, led to the recovery of in vitro sensitivity to IFN-α (84-86). Of note, the most common defect appeared to be the loss of STAT1. However, we recently reported that tumor expression of STAT1 and STAT2 did not correlate with effectiveness of adjuvant IFN-α. We identified a large cohort of high-risk patients with loss of STAT1 in their tumor that exhibited prolonged survival in response to adjuvant IFN-α, while other patients who had normal expression of Jak-STAT proteins recurred after only a few months of IFN therapy (18). Furthermore, we have demonstrated the anti-tumor effects of IFN-α are dependent on STAT1 signaling within immune cells. There is also compelling evidence from other groups to suggest that the immunostimulatory effects of

IFN-α are a critical component of its anti-tumor actions (87-92). Additionally,

Dunn et al. have also shown that endogenously produced IFN-α is required for the prevention of carcinogen-induced tumors and that host immune effector cells are critical targets of IFN-α during the development of protective anti-tumor responses (93). In fact, recent data have shown that the occurrence of autoimmune sequelae and the presence of tumor-infiltrating lymphocytes correlate with clinical response in patients receiving

IFN-α (94, 95). Together, these data suggest that the immunomodulatory actions are critical to the anti-tumor response of this cytokine. However, a careful analysis of gene regulation within human immune effector cells following IFN-α treatment has not been reported.

9 1.5 Suppressors of cytokine signaling

Investigators have identified a family of proteins termed suppressors of cytokine

signaling (SOCS) that negatively regulate Jak-STAT signal transduction (96). The SOCS

family of proteins consists of eight members, including SOCS1-SOCS7 and cytokine

inducible SH2-containing protein (CIS). All SOCS proteins have a central SH2 domain

that allows them to bind to phosphotyrosine residues in cytokine receptors or Jaks, and a

C-terminal SOCS box domain that may function to target SOCS-bound proteins for

proteasomal degradation (97-104). SOCS1 and SOCS3 also contain a kinase inhibitory

region (KIR) that is able to inhibit Jak kinase activity (101, 102, 105).

Transcripts encoding SOCS1, SOCS2, SOCS3, and CIS are often present in cells

at low or undetectable levels, but are rapidly induced by a broad spectrum of cytokines, both in vitro and in vivo. Expression of SOCS genes are regulated by STAT proteins.

Specifically, the SOCS1 promoter contains putative STAT1-, STAT3- and

STAT6-binding sites (106, 107). Recently, Saito et al. reported that SOCS1 mRNA is

eliminated in the absence of STAT1. More specifically, STAT1 was found to act

indirectly by inducing the expression of the interferon regulatory factor-1 (IRF-1)

transcription factor, which in turn stimulates transcription of the SOCS1 gene (107). The

SOCS3 promoter contains a STAT1/STAT3 binding element, while CIS is induced by

STAT5 (108, 109). It is likely that SOCS2 expression is also regulated by the STATs;

however, this remains to be investigated.

10 Once the SOCS species are induced by their respective cytokine, SOCS proteins

can extinguish the signaling pathways that stimulated their production. Therefore, SOCS

proteins act in part of a classical negative feedback loop (96). The expression of SOCS1

and SOCS3 has been shown to mediate potent inhibitory effects on IFN-α-stimulated

signal transduction and gene regulation in several experimental systems (74, 110-112), however, the effect of SOCS expression on exogenous IFN-α in immune cells has yet to be defined in the context of cancer immunotherapy.

1.6 Hypothesis

IFN-α exerts direct anti-proliferative, pro-apoptotic, and anti-angiogenic effects on melanoma cells in vitro, and has distinct immunostimulatory effects that vary according to the immune subset under study (89, 113-115). Unfortunately, the precise molecular targets of exogenously administered IFN-α are unknown. As a result it is not currently possible to identify patients who would have a high likelihood of responding to this treatment. We concluded that STAT1 signal transduction within the host immune effector cells, but not the tumor cells, was critical for mediating the anti-tumor effects of

IFN-α (18). In this regard, we hypothesize that microarray analysis of STAT1-/- and wild type murine splenocytes treated with IFN-α or saline could be utilized to identify candidate molecular targets important for mediating the anti-tumor effect of exogenously administered IFN-α.

11 We have also previously demonstrated a high degree of variability in the

formation of P-STAT1 in patient immune effector cells following IFN-α-2b

immunotherapy and have shown that Jak-STAT signal transduction is down-regulated at

higher dose levels of IFN-α (116). We hypothesize that immune cells from patients

receiving IFN-α immunotherapy would exhibit a gene expression profile that was patient specific. In addition, intermediate doses of IFN-α may be just as effective as higher doses in stimulating the anti-tumor activation of immune cells. This inhibited effect at the higher doses of IFN-α could be due to the induction of inhibitors, such as the SOCS protein family. We hypothesized that the loss of SOCS protein in murine immune effectors would enhance interferon-alpha-induced P-STAT1, transcription of ISGs, and anti-tumor activity. Therefore, modulation of SOCS activity may have beneficial effects in the setting of IFN-α immunotherapy.

CHAPTER 2

THE ROLE OF STAT1 IN MEDIATING THE TRANSCRIPTION OF

INTERFERON STIMULATED GENES IN MURINE IMMUNE CELLS

2.1 Introduction

IFN-α is used as an adjuvant therapy in patients with malignant melanoma following surgical resection of high-risk lesions (lymph node metastases or primary tumor thickness > 4 mm). However, the precise molecular targets of exogenously administered IFN-α are unknown. IFN-α exerts direct anti-proliferative, pro-apoptotic, and anti-angiogenic effects on melanoma cells in culture (113-115), and has potent immunologic actions when administered in vivo (89). The binding of IFN-α to its heterodimeric receptor (IFNAR) activates Janus kinase 1 (Jak1) and Tyrosine kinase 2

(Tyk2) that in turn phosphorylate tyrosine residues on the cytoplasmic region of the

receptor. These phosphotyrosine residues provide docking sites for the signal transducer

and activator of transcription (STAT) family of proteins that are phosphorylated on tyrosine and serine residues by the activated Janus kinases (9). The prototypical IFN-α

signaling reaction results in the formation of a DNA binding complex known as the

interferon-stimulated gene factor 3 (ISGF3) that consists of tyrosine phosphorylated

13 STAT1 (STAT1α or STAT1β), tyrosine phosphorylated STAT2, and a p48 binding

protein, known as interferon regulatory factor 9 (IRF9) (10). This complex then

translocates to the cell nucleus and activates the transcription of IFN-responsive genes

(11).

Our group has previously demonstrated that the anti-tumor effects of IFN-α are critically-dependent on STAT1-mediated signal transduction within host immune cells

(18, 117). In these reports, tumoral expression of STAT1 had no bearing on the ability of

IFN-α to prolong the survival of tumor-bearing mice. In distinct contrast, STAT1-/- mice could not utilize exogenous IFN-α to inhibit the growth of STAT1+/+ melanoma cells.

Thus, STAT1-mediated gene regulation within the host was most important to the anti-tumor effects of IFN-α in this experimental system. There is also compelling evidence from other groups to suggest that the immunostimulatory effects of IFN-α are a critical component of its anti-tumor action (87-92). However, a comprehensive analysis of gene regulation within immune effector cells following IFN-α treatment has not been reported.

To identify genes regulated by STAT1, microarray analysis was utilized to examine the differential gene expression within splenocytes from wild type and

STAT1-deficient mice following in vitro treatment with IFN-α. These studies provide a

14 transcriptional profile of IFN-α-stimulated immune effector cells and the level to which gene expression is dependent on STAT1-mediated signal transduction. These data may aid in understanding the mechanism by which IFN-α exerts its anti-tumor activity.

2.2 Materials and Methods

Reagents. Universal Type I IFN (IFN-A/D, specific activity of 1.1 x 108 U/mg) was

purchased from R&D Systems, Inc. and used in murine experiments (Minneapolis, MN).

Animals. C57BL/6 mice were purchased from Taconic Farms, Inc. (Germantown, New

York). STAT1-/- mice (C57BL/6 background) were generated by homologous

recombination as previously described and housed in a pathogen-free environment (118).

Murine in Vitro Studies. All experiments were performed in compliance with the

guidelines of the Institutional Laboratory Animal Care and Use Committee of The Ohio

State University. Female mice (5-6 weeks of age) were used in all experiments. Spleens

from C57BL/6 and STAT1-/- mice were removed aseptically and dispersed through

70 μM cell strainers. Splenocytes were washed with PBS, pelleted by centrifugation, and

resuspended in RPMI-1640 supplemented with 10% FBS. Previous studies conducted in

our laboratory indicated that IFN-α-induced signal transduction in immune cells was

maximal following stimulation with 104 U/mL (116). In addition, expression of well

15 characterized IFN-α-responsive genes (IFIT2, ISG15) was greater following stimulation with IFN-α for 18 hours versus 6 hours (data not shown). Thus, the majority of in vitro studies employed an 18 hour stimulation. Purified splenocytes were stimulated with either 104 U/ml IFN-A/D or PBS (negative control). Cells were harvested, lysed with

TRIzol reagent (Invitrogen, Carlsbad, CA) and then processed for RNA extraction.

cRNA Preparation and Array Hybridization. Mouse Genome U74Av2 Set GeneChips

(Affymetrix), which query approximately 6,000 murine genes, were used for these analyses. The cRNA was synthesized as suggested by Affymetrix. Briefly, total RNA from cells was prepared in TRIzol (Invitrogen) followed by RNeasy purification (Qiagen,

Valencia, CA). Double stranded cDNA was generated from 8 μg of total RNA using the

Superscript Choice System kit according to the manufacturer’s instructions (Invitrogen).

Biotinylated cRNA was generated by in vitro transcription using the Bio Array High

Yield RNA Transcript Labeling System (Enzo Life Sciences Inc., Farmingdale, NY).

The cRNA was purified using the RNeasy RNA purification kit (Qiagen). cRNA was fragmented according to the Affymetrix protocol and the biotinylated cRNA was hybridized to U133A or U74va2 microarrays (119). The arrays were then scanned

(Affymetrix GMS418) and analyzed (GenePix Pro 4.0) according to Affymetrix protocols.

Data Analysis. Raw data were collected with a confocal laser scanner (Hewlett Packard,

Palo Alto, CA) and probe level data was analyzed using dChip version 1.3 (120). Array

16 normalization was performed using the invariant set procedure. Then, model-based

expression indices (MBEI) were computed using the perfect match only model.

Probe-set level data that was identified as an “array outlier” by dChip was omitted and

considered to be missing data in subsequent analyses. Array quality characteristics

(including percent array outliers, percent present calls and median intensity) were

examined. After MBEI computation and log-transformation of the values, data were

imported into BRB-ArrayTools version 3.22 for subsequent statistical analysis. Probe

sets receiving an Affymetrix “Absent” call for more than 50% of the specimens were

omitted. Univariate paired t-tests were used to make comparisons between classes. A

nominal significance level of 0.001 was employed.

Real Time PCR. Gene expression estimates from the microarray experiments were validated by Real Time PCR for select genes. Following TRIzol extraction and RNeasy purification for microarray analyses, 2 μg of total RNA was reverse transcribed and the resulting cDNA was used as a template to measure gene expression by Real Time PCR using pre-designed primer/probe sets (Assays On Demand; Applied Biosystems, Foster

City, CA) and 2X Taqman Universal PCR Master Mix (Applied Biosystems) according to manufacturer’s recommendations as previously described (121). Pre-designed primer/probe sets for human β-actin were used as an internal control in each reaction well

(Applied Biosystems). Real Time PCR reactions were performed in triplicate in a capped

96-well optical plate. Real Time PCR data was analyzed using the ABI PRISM® 7900

Sequence Detection System (Applied Biosystems).

17 2.3 Results

STAT1-mediated Gene Regulation in Murine Splenocytes. To identify candidate genes within host immune cells involved in mediating the STAT1-dependent immunomodulary effects of IFN-α, the gene expression profile of splenocytes from wild type (WT) and

STAT1-/- mice (n = 3) was examined following treatment with IFN-α (104 U/mL) or PBS

(negative control). Microarray analysis of gene expression indicated that

STAT1-deficiency within the host resulted in impaired or absent expression of many

genes involved in immune function and the response to viral pathogens. From these

studies, three categories of genes were identified based on the importance of STAT1 in

controlling their expression:

STAT1-Enhanced Genes. 25 genes were upregulated greater than 2-fold in response to

IFN-α in WT mice and to a lesser degree in STAT1-/- mice such that the ratio of WT to

knockout (KO) was > 2.0 (Table 2.1). Included in this category were genes involved in

the regulation of T cell adhesion (Ly6c), NK and T cell cytotoxicity (Gzmb), chemotaxis

(Ccl3 and Ccrl2), and several immune response genes (Ifit2, Isg20, Irf7).

STAT1-Suppressed Genes. 20 genes were upregulated greater than 2-fold by IFN-α in

the absence of STAT1 and to a lesser degree in WT mice such that the ratio of KO to WT was > 2.0 (Table 2.2). These included genes encoding negative regulators of Jak-STAT signaling (Socs3), genes involved in the suppression of alloreactive T-cell function

(Arg-1), and genes contributing to chemotaxis (Ccl2, Ccl7, Ccr5).

18 STAT1-Independent Genes. 27 genes were upregulated to a similar degree in response to

IFN-α in WT mice and STAT1-/- mice such that the ratio of WT to knockout (KO)

was < 2.0 and > 0.5 (Table 2.3). Included in this category were genes involved in the

transcriptional regulation (Irf1, Ifi204), class I MHC antigen processing (Psmb9, Tapbp), and genes involved in the ubiquitin cycle (Ube1l, Ube2l6, Zubr1). Of note, the majority

of the genes in this group are expressed to a greater degree in WT mice.

The expression of representative IFN-α-induced genes from each category including

Gzmb and Ly6c (STAT1-enhanced), Ccr5 and Socs3 (STAT1-suppressed), Ifi204 and

Igtp (STAT1-independent) in IFN-treated WT and STAT1-/- splenocytes was validated

via Real Time PCR analysis (Fig. 2.1A-C). In Figure 2.1A, the expression of Gzmb and

Ly6c in response to IFN-α was less in STAT1-/- mice as compared to WT mice.

Conversely, in Figure 2.1B, Ccr5 and Socs3 were induced to a greater degree by IFN-α in STAT1-/- mice. Finally, Ifi204 and Igtp were induced to a similar degree in WT and

STAT1-/- mice (Fig. 2.1C).

2.4 Discussion

The gene expression profile elicited by IFN-α in murine splenocytes and human

immune effectors was investigated using microarray analysis. These studies were conducted in an effort to identify genes that might be instrumental in mediating the

STAT1-dependent anti-tumor effects of IFN-α (18). Murine studies identified a panel of genes whose expression was enhanced by (or suppressed by) STAT1 signal transduction. 19 Mice with genetic deficiencies are an important tool for analyzing the role of specific

transcription factors in the transcriptional response of immune effectors to cytokine

stimulation (122, 123). Previous studies from our laboratory have demonstrated that

STAT1 signaling within mouse immune effector cells is critical to the anti-tumor response of IFN-α (18). Thus, the current finding of STAT1-mediated ISGs likely uncovers numerous genes responsible for the anti-tumor activity of IFN-α. In the present study, defects in STAT1 signal transduction led to the decreased transcription of over 20 genes pertaining to immune cell function, including genes with obvious effects on lymphocyte adhesion and cytotoxicity (e.g. Ly6c and Gzmb). Specifically, granzymes have been found to be essential downstream effector molecules for the cytolysis of virally infected cells (124). Other genes were enhanced by STAT1 signal transduction following

IFN-α including Ifit1, Ifit3, Mx1, Mx2, and Pbef (data not shown). However, these genes did not meet statistical significance in the WT mice treatment samples due to variable gene expression between mice.

Conversely, STAT1 also inhibited the expression of multiple genes, most notably Arg-1, a potent negative alloreactive T cell activity. These enhanced genes could indicate that an alternative IFN-α signaling pathway is augmented in the absence of STAT1 (125).

Gil et al. and Ramana et al. show similar findings when treating WT or STAT1-null mouse bone-marrow-derived macrophages or embryonic fibroblast cell lines treated with

IFN-γ by showing that while many genes were dependent on STAT1 signal transduction,

20 several genes were induced independently of STAT1 expression. In fact, as we show in

Table 2.2, they show that Gadd45g and SOCS3 are upregulated in the absence of

STAT1 (122, 123).

The present report characterizes the transcriptional response of immune cells to IFN-α.

STAT1 deficiency immune splenocytes led to significant alterations in the pattern of

genes that were induced following IFN-α stimulation. Importantly, many of these

STAT1-mediated genes have been previously documented to be expressed in PBMCs

from patients receiving IFN-α administration (126, 127). These genes include Gzmb,

Ifit2, Lgals9, Irf7, Socs3, Ifi27, Ifi30, and Nfil3. Perhaps the genes described as

STAT1-dependent or induced in the absence of STAT1 are clinically relevant to the anti-tumor action of IFN-α. Further analysis of gene regulation in response to exogenous

IFN-α will provide important clues as to its mechanism of action.

21 2.4 Tables and Figures

Geometric mean fold Gene Function KO WT WT/KO Probe set Component of Sp100-rs (Csprs) G-protein coupled receptor 1.3 5 3.85 101845_s_at Granzyme B (Gzmb) Cytolysis by natural killer cells and T cells 2 7.4 3.70 102877_at Schlafen 3 (Slfn3) Negative regulation of cell proliferation 2.2 7.1 3.23 98299_s_at Lymphocyte antigen 6 complex, locus C (Ly6c) Positive regulation of CD8 T cell adhesion 1 3.2 3.21 93077_s_at Calcium/calmodulin-dependent protein kinase II, beta (Camk2b) Enhances T cell proliferation/cytotoxicity 0.7 2.2 3.14 103562_f_at LOC625360 Unknown 1.8 5.4 3.00 102254_f_at Interferon-stimulated protein (Isg20) Antiviral resposne, exonuclease activity 2.7 7.9 2.92 103432_at Sarcoglycan, beta (Sgcb) Membrane organization and biogenesis 1.2 3.4 2.83 93898_at Interferon-induced protein with tetratricopeptide repeats 2 (Ifit2) Immune response 3 8.5 2.83 103639_at Cysteine and glycine-rich protein 1 (Csrp1) Actin cytoskeleton organization and biogenesis 0.8 2.2 2.75 160065_s_at* 2-cell-stage, variable group, member 1 (Tcstv1) Unknown 1.2 3.3 2.75 94727_f_at Fas death domain-associated protein (Daxx) Apoptosis, transcriptional repressor 1.5 4 2.67 96125_at AF067061 Unknown 1.1 2.8 2.54 94749_f_at* LOC673100 Unknown 0.8 2 2.50 160799_at DNA-damage inducible transcript 3 (Ddit3) Regulation of transcription/apoptosis 0.9 2.2 2.44 101429_at Schlafen 4 (Slfn4) Negative regulation of cell proliferation 3.5 8.5 2.43 92315_at 2-cell-stage, variable group, member 3 (Tcstv3) Unknown 1.1 2.6 2.36 96584_f_at Lectin, galactoside-binding, soluble, 3 binding protein (Lgals3bp) Scavenger receptor activity 1.7 4 2.35 97507_at Histocompatibility 2, T region locus 24 (H2-T24) Endogenous antigen presentation 1.2 2.8 2.33 94746_at Chemokine (C-C motif) receptor-like 2 (Ccrl2) Chemotaxis, inflammation 1 2.3 2.30 93617_at Lectin, galactose binding, soluble 9 (Lgals9) Cell adhesion, galactose binding 1.3 2.8 2.16 103335_at Interferon regulatory factor 7 (Irf7) Transcription factor 4.8 10.1 2.11 104669_at* Interferon, alpha-inducible protein (G1p2) Chemotaxis, ubiquitin cycle 5.7 11.9 2.09 98822_at SH3-domain binding protein 2 (Sh3bp2) NK cell-mediated cytotoxicity, TCR-mediated signaling 1.1 2.2 2.00 92975_at

* Significant gene upregulation was observed in multiple probe sets

Table 2.1: STAT1-Enhanced Genes: Augmented upregulation (>2 fold; p < 0.001) in WT splenocytes following IFN-α

Geometric mean fold Gene Function KO WT KO/WT Probe set Arginase 1, liver (Arg1) Alloreactive T cell suppression 7.5 0.7 10.71 93097_at Chemokine (C-C motif) ligand 7 (Ccl7) Chemotaxis, inflammation 11.9 1.2 9.92 94761_at Chemokine (C-C motif) ligand 2 (Ccl2) Chemotaxis, inflammation 8.9 1.4 6.36 102736_at Transforming growth factor, beta induced (Tgfbi) Cell cycle regulation, cellular adhesion 2 0.4 5.00 92877_at Interleukin 18 receptor 1 (Il18r1) Immune response 3.2 0.7 4.57 101144_at Chemokine (C-C motif) receptor 5 (Ccr5) Chemotaxis, inflammation 7.6 1.7 4.47 161968_f_at* Growth arrest and DNA-damage-inducible 45 gamma (Gadd45g) Apoptosis, T-helper 1 cell differentiation 4.9 1.1 4.46 101979_at Suppressor of cytokine signaling 3 (SOCS3) Inhibition of Jak/STAT signaling 5.5 1.3 4.23 162206_f_at* Disabled homolog 2 (Dab2) Regulation of cell cycle 2 0.5 4.00 98045_s_at AW061234 Unknown 2.2 0.6 3.67 103697_at Epidermal growth factor receptor pathway substrate 8 (Eps8) Electron transport 2 0.6 3.33 103222_at Interferon, alpha-inducible protein 27 (Ifi27) Antiviral response 8.3 2.6 3.19 92718_at Interleukin 1 receptor-like 1 (Il1rl1) Immune response 2.6 0.9 2.89 98500_at Interleukin 1 receptor, type II (Il1r2) ATP-binding cassette (ABC) transporter 2.2 0.8 2.75 102658_at Interleukin 2 receptor, alpha chain (Il2ra) T-cell proliferation 2.4 0.9 2.67 101917_at EST sequence Unknown 3.1 1.2 2.58 92779_f_at EST sequence Unknown 2.8 1.1 2.55 92778_i_at Interferon gamma inducible protein 30 (Ifi30) Immune response 2 0.8 2.50 97444_at T-cell specific GTPase (Tgtp) GTPase 9.9 4.6 2.15 102906_at Nuclear factor, interleukin 3, regulated (Nfil3) Cellular survival 2.5 1.2 2.08 102955_at

* Significant gene upregulation was observed in multiple probe sets

Table 2.2: STAT1-Suppressed Genes: Augmented upregulation (>2 fold; p < 0.001) in STAT1 KO splenocytes following IFN-α

Geometric mean fold Gene Function KO WT WT/KO Probe set Torsin family 3, member A (Tor3a) Protein folding, ATP binding 2.1 4.1 1.95 96533_at TRAF-type zinc finger domain containing 1 (TRAFD1) Unknown 1.4 2.7 1.93 103254_at* Interferon gamma induced GTPase (Igtp) Immune response 2.6 5.0 1.92 160933_at Interferon regulatory factor 1 (Irf1) Immune response, transcription factor 1.1 2.1 1.91 102401_at Ubiquitin-activating enzyme E1-like (Ube1l) Ubiquitin cycle, apoptosis 1.4 2.6 1.86 102279_at Tripartite motif protein 21 (Trim21) Ubiquitin-protein ligase activity 1.3 2.4 1.85 102678_at* Ubiquitin-conjugating enzyme E2L6 (Ube2I6) Ubiquitin cycle 1.2 2.2 1.83 102279_at* zinc finger, UBR1 type 1 (Zubr1) Ubiquitin cycle metabolism 1.2 2.2 1.83 104041_at Interferon activated gene 204 (Ifi204) transcriptional co-activator, antiviral 2.4 4.3 1.79 98466_r_at Tryptophanyl-tRNA synthetase (Wars) Protein biosynthesis 1.2 2.1 1.75 98605_at* TAP binding protein (Tapbp) MHC class I molecules linkage to TAP 1.2 2.1 1.75 102689_at LOC673370 Unknown 1.9 3.3 1.74 93779_at Thymidylate kinase family LPS-inducible member (Tyki) Thymidylate kinase activity 2.8 4.7 1.68 103066_at Syntaxin binding protein 3 (Stxbp3) Glut4 translocation 1.2 2.0 1.67 92648_at Protein-tyrosine sulfotransferase 1 (Tpst1) Inflammatory response 1.4 2.2 1.57 103032_at BC094916 Unknown 1.8 2.8 1.56 103615_at AI481105 transcription regulation 2.0 3.1 1.55 102965_at membrane-spanning 4-domains, subfamily A, member 4C (Ms4a4c) Signal transduction 2.0 3.1 1.55 98373_at CD86 antigen (Cd86) T cell stimulation, inflammatory response 1.3 2.0 1.54 102831_s_at EST sequence Unknown 1.3 2.0 1.54 103517_at kinesin family member 9 (Kif9) microtubule motor 1.3 2.0 1.54 161035_at Promyelocytic leukemia (Pml) Regulation of transcription 1.9 2.9 1.53 99015_at Proteosome (prosome, macropain) subunit, beta type 9 (Psmb9) Protein catabolism, processing of class I MHC peptides 1.8 2.5 1.39 93085_at SAM domain and HD domain, 1 (Samhd1) Immune response 1.9 2.6 1.37 103080_at lymphocyte antigen 6 complex, locus A (Ly6a) Cellular adhesion 3.1 2.2 0.71 93078_at C79468 Unknown 0.7 0.4 0.57 95984_at Potassium channel tetramerisation domain containing 12 (Kctd12) Potassium ion transport 0.7 0.4 0.57 104735_at

* Significant gene upregulation was observed in multiple probe sets

Table 2.3: STAT1-Independent Genes: Upregulation (>2 fold; p < 0.001) in WT and STAT1 KO splenocytes following IFN-α

24 Figure 2.1. Real Time PCR Validation of Murine Microarray Data. Real Time PCR

analysis was used to validate the expression of (A) Ly6c and Gzmb (two

STAT1-enhanced genes), (B) Ccr5 and Socs3 (two STAT1-suppressed genes), and (C)

Ifi204 and Igtp (two STAT1-independent genes) in WT and STAT1-/- splenocytes. Data were expressed as the mean fold increase relative to baseline levels (PBS treatment). All real time PCR data were normalized to the level of β-actin mRNA (housekeeping gene).

Error bars denote the standard deviations of triplicate experiments.

25 Figure 2.1

A 25 WT STAT1 KO

10 Fold increase vs. PBS Fold vs. increase 5

0 Ly6c Gzmb

B 12 WT STAT1 KO 10

4 Fold increase vs. PBS vs. increase Fold

0 Ccr5 Socs3

Continued…

26 Figure 2.1 Continued

C 40 WT STAT1 KO

Fold increase vs. PBS vs. Fold increase 10

0 Ifi204 Igtp

CHAPTER 3

GENE EXPRESSION PROFILING OF THE IMMUNE RESPONSE TO

INTERFERON-ALPHA

3.1 Introduction

Surgical treatment of early stage malignant melanoma is frequently curative,

however, the therapeutic options for patients with metastatic disease are limited. IFN-α

has been used both as an adjuvant following the surgical resection of high-risk lesions

(lymph node metastases or primary tumor thickness > 4 mm) and in the advanced disease

setting. The IFN-α-receptor is expressed on melanoma tumor cells as well as on immune

effectors and mediates many of its effects via activation of the Janus kinase (Jak)-signal

transducers and activators of transcription (STAT) pathway. IFN-α exerts direct

anti-proliferative, pro-apoptotic, and anti-angiogenic effects on melanoma cells in culture

and has distinct immunostimulatory effects that vary according to the immune subset

under study (89, 113-115). Unfortunately, the precise molecular targets of exogenously administered IFN-α are unknown. As a result it is not currently possible to identify patients who would have a high likelihood of responding to this treatment.

28 We have examined the role of the Jak-STAT signaling pathway in a murine

model of malignant melanoma using STAT1-deficient mice and STAT1-deficient

melanoma cell lines and found that loss of STAT1 signal transduction within the host

abrogated the anti-tumor effects of IFN-α (18). In contrast, the survival benefits

associated with IFN-α administration were maintained when normal

(i.e. STAT1-competent) mice were challenged with a STAT1-/- murine melanoma cell

line. We concluded that STAT1 signal transduction within the host, but not the tumor

cell, was critical for mediating the anti-tumor effects of IFN-α. Further experiments by

our group and others indicate that the immunostimulatory effects of IFN-α are an

important component of its anti-tumor actions in mice (87-92, 127, 128). In fact, recent

data have shown that the occurrence of autoimmune sequelae and the presence of

tumor-infiltrating lymphocytes correlate with clinical response in patients receiving

IFN-α (94, 95). Together, these data suggest that the immunomodulatory actions are critical to the anti-tumor actions of this cytokine. However, a careful analysis of gene regulation within human immune effector cells following IFN-α treatment has not been reported.

We hypothesized that the gene expression profile of PBMCs from IFN-α treated

patients could be predicted based on in vitro studies of IFN-α stimulated immune cell

subsets. The gene expression profile of IFN-α stimulated peripheral blood mononuclear

cells (PBMCs) and immune cell subsets and PBMCs from melanoma patients receiving

high dose IFN-α-2b (20 MU/m2 i.v.) was evaluated via microarray analysis. We found

that there was only minor overlap in gene expression between IFN-α-stimulated T cells, 29 NK cells, monocytes, and whole PBMCs. As a result of inter-patient variation, the gene

expression profile of normal donors PBMCs following in vitro IFN-α stimulation did not

match that of PBMCs obtained from melanoma patients receiving adjuvant IFN-α.

However, within individual patients, the gene expression profile of PBMCs stimulated

in vitro with IFN-α was largely predictive of the gene expression profile of PBMCs obtained from the same patient following clinical administration of IFN-α.

3.2 Materials and Methods

Reagents. Recombinant human (hu) IFN-α-2b (specific activity of 2 x 108 IU/mg) was

purchased from Schering-Plough, Inc. (Kenilworth, NJ).

Isolation of Immune Subsets. For in vitro assays requiring total PBMCs or immune

subsets, source leukocytes were obtained from healthy adult donors (American Red

Cross, Columbus, OH). PBMCs were isolated by Ficoll-Paque Plus (Amersham

Pharmacia Biotech, Uppsala, Sweden) density gradient centrifugation as previously described (129). Lymphocytes were enriched for individual cell populations (CD3+,

CD56+, and CD14+ cells) by negative selection using RosetteSep reagents (Stem Cell

Technologies, Vancouver, British Columbia). Following isolation, the purity of enriched

cell populations was typically on the order of 95 - 99% as determined by flow cytometry

(data not shown). Purified cells were then cultured in RPMI-1640 media supplemented

30 with 10% human AB serum (Pel-Freez Clinical Systems, Brown Deer, WI) at 37°C with

4 5% CO2 and stimulated with either 10 U/ml IFN-α-2b or PBS for 18 hours. Preliminary

studies indicated that the transcription of IFN-α-responsive genes in PBMCs from normal human donors was equivalent at the one hour and four hour time points following an

in vitro stimulation with 104 U/mL of IFN-α (data not shown). Following incubation,

cells were harvested by centrifugation, resuspended in TRIzol reagent (Invitrogen), and

processed for RNA extraction.

Patients and Blood Samples. Peripheral blood was obtained from melanoma patients

(3 females, 4 males) immediately prior to, and one-hour following intravenous (i.v.)

administration of high dose IFN-α-2b (20 x 106 IU/m2). All samples were obtained at

The Ohio State University following informed consent under an IRB-approved protocol

(OSU 99H0348). PBMCs were isolated from peripheral blood (8 mL) via density

gradient centrifugation with Ficoll-Paque Plus (Amersham Pharmacia Biotech) and

immediately stored in TRIzol reagent (Invitrogen) at -80°C.

cRNA Preparation and Array Hybridization. U133 Plus 2.0 Arrays (Affymetrix, Santa

Clara, CA), which query approximately 47,000 human transcripts were used in these

analyses. The cRNA was synthesized as suggested by Affymetrix. Following lysis of

cells in TRIzol (Invitrogen), mRNA was prepared by RNeasy purification (Qiagen,

Valencia, CA). Double stranded cDNA was generated from 8 μg of total RNA using the

Superscript Choice System according to the manufacturer’s instructions (Invitrogen).

Biotinylated cRNA was generated by in vitro transcription using the Bio Array High 31 Yield RNA Transcript Labeling System (Enzo Life Sciences Inc., Farmingdale, NY).

The cRNA was purified using the RNeasy RNA purification kit (Qiagen). cRNA was

fragmented according to the Affymetrix protocol and the biotinylated cRNA was hybridized to U133A microarrays (119). The arrays were then scanned (Affymetrix

GMS418) and analyzed (GenePix Pro 4.0) according to Affymetrix protocols.

Data Analysis. Raw data were collected with a confocal laser scanner (Hewlett Packard,

Palo Alto, CA) and probe level data were analyzed using dChip version 1.3 [19].

Invariant-set normalization was performed, and only perfect match probes were used in

computing the model-based expression indices (MBEIs). “Array outliers,” identified by

dChip at the probe-set level, were set to missing. The log2(MBEIs) were then calculated

and exported to BRB-ArrayTools v3.2 for further analysis. For each analysis, probe sets

receiving an Affymetrix “Absent” call for more than 50% of the specimens were filtered.

T-tests were used to compare classes where samples were independent. If samples were

dependent, because the same sample was studied at two time points or under two

different conditions, paired t-tests were used in class comparisons. All tests were

two-sided and conducted at a nominal significance level of 0.001. In addition, adjusted

p-values using the Benjamini-Hochberg method were calculated to report probe sets that

met a false discovery rate (FDR) criterion of 10%. Probe sets found to be differentially

expressed and meeting the FDR criterion with at least a two fold change in expression in

the cells of normal human donors were selected for further study. For these a priori tests,

the alpha level was set to 0.05 and p-values were adjusted using Holm’s method to

account for the multiple testing. 32

A random effects model with repeated measures was used to determine if there was similar gene expression profiles between PBMCs treated with IFN-α in vitro and in vivo from the same patient. This test was performed for each of the pre-selected 143 probe sets of interest (representing 65 genes). These probe sets were identified on the basis of preliminary studies from our group as well as previous work from other groups

(Table 3.1, Table 3.7, and Reference 132). In testing if there was a significant difference in expression between two dose levels, paired t-tests were used. Type I error was protected by adjusting p-values using Holm’s method. Two-sided significance levels were set at α = 0.10.

City, CA) and 2X Taqman Universal PCR Master Mix (Applied Biosystems) according to the manufacturer’s recommendations as previously described (121). Pre-designed primer/probe sets for human β-actin were used as an internal control in each reaction well

(Applied Biosystems). Real Time PCR reactions were performed in triplicate in a capped

96-well optical plate. Real Time PCR data was analyzed using the ABI PRISM® 7900

Sequence Detection System (Applied Biosystems).

3.3 Results

Early transcriptional response of IFN-α-stimulated PBMCs from normal donors. In

order to detect the genes that were induced immediately following IFN-α-stimulation, a

1 hour time point was selected. This is also a convenient time to obtain blood from

individuals receiving adjuvant IFN-α in the outpatient setting (see below). The gene

expression profile of normal PBMCs (n = 3) following in vitro treatment with IFN-α was

evaluated via microarray analysis (Table 3.1). This analysis revealed that 21 genes were

significantly upregulated in PBMCs in response to a 1 hour treatment with 104 U/mL

IFN-α (> 2 fold induction; p < 0.001, paired t-test). The expression profile was

characterized by the consistent induction of genes encoding antiviral/immune response

proteins (IFIH1, IFIT1, IFIT3, GBP1), chemo-attractants (CXCL10, CCL8), cytokines

(IL-6), and other species, such as SOCS2 (an inhibitor of growth hormone signaling) and

CASP4 (caspase 4, executor of cell death in response to endoplasmic reticulum stress).

Real Time PCR was employed to validate the expression profiles of three representative

genes (CXCL10, CCL8, IFIT1; Figure 3.1). Transcripts for these genes were induced

approximately 2-4 fold over resting levels following a 1 hour treatment of PBMCs with

IFN-α.

Late transcriptional response of IFN-α-stimulated PBMCs and immune cell subsets from

normal donors. Cell-to-cell interactions and autocrine cytokine stimulation likely

34 influence the transcriptional profile of IFN-α-stimulated PBMCs. The gene expression profile of normal PBMCs (n = 3) following an 18 hour in vitro treatment with IFN-α was

therefore evaluated via microarray analysis (Table 3.2) (130, 131). This analysis

revealed that a total of 26 genes were upregulated in response to an 18 hour treatment

with 104 U/mL IFN-α (> 2 fold induction, p < 0.001; paired t-test). Only one gene was down-regulated in PBMCs under these conditions (EIF3S11, protein biosynthesis

initiation factor). The expression profile was characterized by the consistent induction of

genes encoding antiviral/immune response proteins (IFI16, IFI44, IFIT2, IRF2, ISG20,

LILRA2, OAS2, OASL), regulators of transcription (DRAP2, IRF2, SP110), T cell

activation markers (LY6E), and other species such as CD38, USP18, and MT1H

(Table 3.2). Increased expression of these genes in response to IFN-α was not surprising

given the immunomodulatory effects of Type I IFNs (132, 133). However,

transcriptional regulation of the genes SPTLC2 (sphingolipid biosynthesis), N4BP1

(unknown function), BLVRA (electron transport), and EIF3S11 by IFN-α has not been

previously reported. This data was validated by measuring the expression of several

notable genes expressed in PBMCs (IFIT2, ISG20, LY6E) by Real Time PCR

(Figure 3.2A). Importantly, paired t-test analysis of the microarray results at the 1 hour

and 18 hour time points revealed no overlap of gene expression. Thus, microarray

analyses of the PBMC response to IFN-α-stimulation are markedly dependent on the time point selected for evaluation.

35 Similar in vitro experiments were conducted in CD3+ T cells, CD56+ natural killer

(NK) cells, and CD14+ monocytes (n = 3 for each). The expression of 28 genes was

modulated in the T cell compartment after IFN-α-2b stimulation (27 genes upregulated and 1 down-regulated, p < 0.001, Table 3.3). These genes function in multiple processes including the regulation of apoptosis (TNFSF10), antigen binding (LAG3, MICB), regulation of the NFκB signal cascade (LGASL9, MYD88), transcriptional regulation

(PHF11, SP100), and the antiviral response (IFI44, IFIT1, OAS3). The expression of 32 genes was modulated in the NK cell compartment in response to IFN-α-2b treatment

(30 upregulated genes and 2 down-regulated genes, p < 0.001, Table 3.4). Several of these genes function in transcriptional regulation (IRF7, PML, SP100, TFCP2), as interferon class cytokine receptors (CLRF2), in cell motility (MARCKS), metal ion binding (MT1F, MT1H, MT1X, MT2A), and the antiviral response (IFI27, IFIT5,

IFITM3, ISG20, OAS1, OAS2, OAS3). Finally, the expression of 59 genes was regulated following IFN-α-2b treatment of monocytes (32 upregulated and 26 down-regulated genes, p < 0.001, Table 3.5 and 3.6). Several of these genes function in regulating apoptosis (CUL1, MX1, SGPL1, TFNRSF5), cell motility (PECAM1), antiviral responses (ARHGDIB, GBP1, LCP2, LILRB3, IFI44, IFITM1, ISG20, OAS3,

OASL), T cell activation (CD69), and regulation of transcription (IRF1, PIP5K2B,

STAT1).

These findings were validated by measuring the expression of several notable genes expressed in T cells (IFIT1, LAG3, OAS3), NK cells (IRF7, ISG20, MX2), and monocytes (CD69, ISG20, OASL) by Real Time PCR (Figure 3.2B-D). Of note, just 13 36 of the 27 genes induced in IFN-treated PBMCs were also upregulated in purified T cells

(n = 3), NK cells (n = 3), or monocytes (n = 8). This cell-specific gene expression profile

suggests that each immune cell subset responds to IFN-α in a unique fashion.

Remarkably, the gene expression profile of whole PBMCs at 18 hours is not a composite of the gene expression profiles of the individual immune subsets. This finding implies that IFN-α stimulation of whole PBMCs is characterized by unique cellular interactions

that occur between its constituent subsets.

Gene regulation in PBMCs from melanoma patients receiving high dose IFN-α therapy.

In order to characterize the in vivo immune response to IFN-α, PBMCs were isolated from peripheral venous blood obtained from patients (n = 7) immediately prior to and 1 hour following i.v. administration of IFN-α-2b (20 MU/m2). These patients were

receiving IFN-α as an adjuvant therapy following surgical resection of high-risk

melanoma lesions (lymph node disease or tumor > 4 mm in diameter). Analysis of gene

expression within patient PBMCs demonstrated that 23 genes were upregulated over

baseline values following IFN-α-2b therapy (> 2.0 fold induction, p < 0.001; Table 3.7).

Genes involved in antigen presentation (TAP1), cell adhesion (LGALS3BP), and known

IFN-α-stimulated genes (IFI44, IFIT4, IFITM3, IRF7, ISG20, OASL) were among those induced 1 hour post-IFN-α therapy in patient PBMCs. Real Time PCR was employed to validate the expression profiles of three representative genes (IRF7, OASL, TAP1;

Figure 3.3). Genes that were induced in the PBMCs of melanoma patients receiving adjuvant IFN-α were compared to those induced by in vitro IFN-α treatment of normal

37 PBMCs. Even using a less stringent analysis (α = 0.10), only the IFIT1 and IFIT3 genes

were identified as being upregulated in both experimental groups following in vitro or

in vivo IFN-α treatment. Of note, the genes that were found to be upregulated in vitro in

response to IFN-α were also upregulated in patient PBMCs post-IFN-α-therapy.

However, these genes did not meet statistical significance in the patient treatment samples due to inter-patient variability. Conversely, the genes that were upregulated in patient PBMCs were also induced in PBMCs in vitro. Again, these genes did not meet statistical significance in the normal PBMCs due to sample variability.

Comparison of IFN-α gene regulation in PBMCs stimulated in vitro with IFN-α to that of

PBMCs obtained from the same donor before and after IFN-α therapy. We next investigated the correlation between microarray results generated using PBMCs

stimulated in vitro with IFN-α and those that were generated utilizing PBMCs obtained

from the same patients pre- and post-IFN-α therapy. PBMCs isolated from the peripheral

venous blood of patients (n = 3) immediately prior to and 1 hour following i.v.

administration of IFN-α-2b (20 MU/m2) were used in microarray analysis. An aliquot of the PBMCs collected prior to IFN-α therapy was treated in vitro with PBS or IFN-α

(104 U/mL for 1 hour) and these samples were also evaluated by microarray analysis. In order to compare the microarray results obtained in vitro and in vivo for these 3 patients,

an ANOVA test was employed on 143 pre-selected probe sets representing 65 ISGs (see

Materials and Methods). Evaluation of a relatively small number of pre-selected genes that are regulated by IFN-α markedly increases the statistical power of such an analysis.

38 Analysis of the pre-selected probe sets revealed that over 80% of these ISGs were upregulated to a similar degree in the patient PBMCs treated in vitro with IFN-α and in the PBMCs obtained from these same patients post-IFN-α therapy (data not shown).

Importantly, the in vivo expression profile generated for these 3 melanoma patients was similar to that of the 7 melanoma patients presented in Table 3.7 (i.e. all 23 genes identified in Table 3.7 were also induced in this group of 3 patients). This result demonstrates that microarray analysis of patient PBMCs following in vitro stimulation with IFN-α may be a useful predictor of the individual response to IFN-α immunotherapy.

3.4 Discussion

The present report characterizes the transcriptional response of immune cells to

IFN-α using microarray analysis. We determined the following: 1) the expression profile of IFN-α-stimulated PBMCs was markedly dependent on the duration of cytokine treatment, 2) immune cell subsets exhibited distinct IFN-α-induced gene expression profiles that were not entirely reflective of the overall PBMC profile, 3) PBMCs from melanoma patients showed alternative gene expression profiles following immunotherapy with IFN-α-2b compared to normal donor PBMCs, and 4) microarray analysis of patient

PBMCs following in vitro stimulation with IFN-α may be a useful predictor of the individual in vivo response to IFN-α.

39 Microarray analysis of PBMCs from normal donors identified 21 genes that were upregulated greater than 2 fold in response to in vitro IFN-α treatment at the 1 hour time point and 26 genes at the 18 hour time point (p < 0.001 significance level). Importantly, there was no significant overlap between these two time points. This suggested that the gene expression profile can vary markedly according to the timing of the cytokine stimulus. Ji et al. reported that a total of 516 genes were induced at 3 or 6 hours following in vitro treatment of PBMCs with 200 U/ml of IFN-α (134). Interestingly, they reported only a modest overlap in the gene expression profile between 3 and 6 hour after IFN-α stimulation, which further supports the notion that the transcriptional response to IFN-α varies significantly over time. This group’s use of cells from patients with chronic hepatitis C is an important difference from the current study, since the presence of this chronic viral infection likely enhanced the expression of multiple genes with anti-viral effects. A comparison of our in vitro microarray analysis results with this previous study revealed overlap for only a few genes (i.e. IFIT1, IFI-16, IFIT2, MX2,

IRF7, TNFSF10, IFITM1, STAT1). This difference may also be due to the alternative statistical analysis employed in the study by Ji et al. (SAM analysis with 10% FDR cutoff vs. our dChip with p < 0.001), as well as differences in time points examined, differences in IFN-α dosages, and varying degrees of viral load.

An analysis of immune cell subsets revealed that 13 of the genes that were induced in IFN-α-stimulated PBMCs (OAS1, OAS3, OASL, G1P2, ISG20, IFIT1, IFIT2,

MX1, MX2, VIG1, PBEF, CXCL10, IFI16) were also upregulated in T cells, NK cells or monocytes (as denoted in Tables 3.2B-D by a “**”). However, the lack of significant 40 overlap between NK cells and T cells implied that the functional responses of these two

compartments to exogenous IFN-α were vigorous, yet unexpectedly distinct. A

comparison of these results with those obtained using virally-infected cells treated with

IFN-α (e.g., endothelial, hepatoblastoma, and keratinocyte cell lines) revealed only minor

similarities and underscored the tissue specificity of IFN-α-induced signal transduction

and gene regulation (134-138). Of note, the gene expression profile of IFN-α stimulated monocytes was similar to data from a previous report of IFN-stimulated mononuclear phagocytes that had been pretreated with lipopolysaccharide in vitro (139). “Signature”

genes induced by IFN-α in both reports included GBP1, ISG20, MX1, STAT1, and

OAS3 among others. The monocyte compartment was also unique in that several dozen

genes were significantly down-regulated in response to IFN-α. This implies that there is

significant basal expression of a set of ISGs in monocytes.

This is the first study to employ microarray techniques in the analysis of

transcriptional responses of PBMCs obtained from cancer patients during IFN-α

immunotherapy (126). The pattern of gene regulation was found to be moderately

variable between patients. The level of induction was 2-6 fold for 22 of the 23 genes

listed in Table 3.7, while the level of induction was highly variable for 121 other genes

that did not reach statistical significance (e.g. 0.9-28.2 fold for RGS1, 0.4-22.4 fold for

NR4A2). Whitney et al. utilized microarray analysis to explore the extent of inter-

individual variation in gene expression within unstimulated PBMCs of healthy donors

(n = 75). The greatest degree of individual variation occurred within a cluster of 15

genes know to be IFN-responsive (140). The authors acknowledged that this might be 41 due to subclinical or recent infection in some donors, however this observation was

subsequently documented in another study (141). Furthermore, basal levels of some

cytokines (i.e. IL-4, IL-6, IL-10) can vary significantly among healthy individuals and particularly among patients with advanced malignancy (141, 142). It is possible that the cytokine profile of cancer patients could contribute to the variable gene expression and subsequent response to cytokine immunotherapy. The data from the present study indicate that the response to IFN-α, while generally predictable, can vary widely in amplitude. This may explain the variation in patient responsiveness to the anti-tumor activity of IFN-α and its side effects.

In the present study, IFN-α-induced gene expression profiles were also donor specific in that normal donor PBMCs treated in vitro with IFN-α did not appear to predict the gene expression profile of melanoma patients receiving IFN-α. In contrast, the gene expression profile of PBMCs obtained from patients prior to therapy and treated in vitro with IFN-α closely matched the in vivo expression profile obtained after IFN-α administration in these same patients. These findings support the potential use of microarray technology to predict the individual patient gene response to IFN-α.

The present study demonstrated that the transcriptional profile of

IFN-α-stimulated immune cells is affected by multiple factors including duration of stimulation, cell type, inter-patient variation, and the method of exposure to cytokine

(i.e. in vitro stimulation versus in vivo administration). We also found that in vitro

42 analysis of IFN-α-stimulated PBMCs may be predictive of the in vivo expression profile following IFN-α immunotherapy, however, further investigation of this finding is warranted.

43 3.5 Tables and Figures

Gene Function Fold Affy probe Chemokine (C-X-C motif) ligand 10 (CXCL10) Chemotaxis, stimulates NK cells and monocytes 176.1 204533_at Chemokine (C-C motif) ligand 8 (CCL8) Chemotaxis 122.3 214038_at Interferon-induced protein with tetratricopeptide repeats 1 (IFIT1) Immune response 117.2 203153_at Interferon-induced protein with tetratricopeptide repeats 3 (IFIT3) Immune response 62.2 204747_at Sterile alpha motif domain containing 9-like (SAMD9L) Unknown 44.3 243271_at LOC341720 Unknown 24.6 240287_at Cholesterol 25-hydroxylase (CH25H) Catalytic activity 21.3 206932_at Guanylate binding protein 1 (GBP1) Immune response 13.9* 231577_s_at CD274 antigen (CD274) Immune response, cell proliferation 13.7 227458_at Nuclear receptor coactivator 7 (NCOA7) Transcriptional regulation, nucleic acid metabolism 8.5 225344_at Interferon induced with helicase C domain 1 (IFIH1) Regulation of apoptosis, regulation of translation 5.6 1555464_at FLJ11000 Unknown 4.5 243465_at FLJ10159 Unknown 4.4 218974_at EST sequence Unknown 4.3 237315_at Caspase 4 (CASP4) Induction of apoptosis 3.9 213596_at Interleukin 6 (IL6) B cell and T cell differentiation 3.6 205207_at EST sequence Unknown 3.6 242471_at Suppressor of cytokine signaling 2 (SOCS2) Inhibition of Jak/STAT signaling 2.8 203372_s_at EST sequence Unknown 2.5 243934_at T-cell activation GTPase activating protein (TAGAP) GTPase activator 2.4 234050_at ADP-ribosylation factor-like 8 (ARL8) GTPase 2.1 226345_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001

† Table 3.1: Gene Regulation in PBMCs following 1hr IFN-α treatment

44 Figure 3.1. Real Time PCR analysis of select genes identified by microarray analysis of

PBMCs following 1 hour in vitro IFN-α stimulation. Real Time PCR was used to validate the expression of genes in PBMCs (CXCL10, CCL8, IFIT1). Data were expressed as the mean fold increase relative to baseline levels (PBS treatment). All real time PCR data were normalized to the level of β-actin mRNA. Error bars denote the standard deviations of triplicate experiments.

45 Figure 3.1

Fold increasePBS vs. 1

0 CXCL10 CCL8 IFIT1

Gene Function Fold Affy probe Ubiquitin specific protease 18 (USP18) Ubiquitin-dependent protein catabolism 15.1 219211_at CD38 antigen (p45) (CD38) Apoptosis 8.8 205692_s_at Interferon-induced protein 44 (IFI44) Antiviral response 7.5 214059_at DEAD (Asp-Glu-Ala-Asp) box polypeptide 58 (DDX58) Helicase, Deoxyribonuclease, ubiquitin-protein ligase 7.1 218943_s_at Hect domain and RLD5 (HERC5) Regulation of cyclin dependent protein kinase activity 7 219863_at 2'-5'-oligoadenylate synthetase 2 (OAS2) Immune response, nucleic acid metabolism 6.7 206553_at 2'-5'-oligoadenylate synthetase-like (OASL) Immune response, nucleic acid metabolism 6.2 210797_s_at Lymphocyte antigen 6 complex, locus E (LY6E) T cell differentiation and activation, antiviral response 5.8 202145_at Serine palmitoyltransferase, long chain base subunit 2 (SPTLC2) Unknown 4.6 216202_s_at Likely ortholog of mouse D11lgp2 (LGP2) DNA restriction 4.3 219364_at Interferon stimulated gene 20kDa (ISG20) Immune response, proliferation 3.8* 204698_at Retinoic acid- and interferon-inducible protein (IFIT2) Immune response 3.8 203595_s_at Three prime repair exonuclease 1 (TREX1) DNA repair and replication 3.6 205875_s_at Metallothionein 1H (MT1H) Metal ion binding 3.4 206461_x_at Metallothionein 2A (MT2A) Metal ion binding 3 212185_x_at Phospholipid scramblase 1 (PLSCR1) Phospholipid scramblase 3* 202430_s_at SP110 nuclear body protein (SP110) Regulation of transcription, transcription factor 2.9* 208012_x_at Tripartite motif-containing 21 (TRIM21) Protein ubiquitination 2.9 204804_at Zinc finger, CCHC domain containing 2 (ZCCHC2) Nucleic acid binding 2.7 219062_s_at DR1-associated protein 1 (negative cofactor 2 alpha) (DRAP1) Negative regulation of transcription 2.6 203258_at Biliverdin reductase A (BLVRA) Electron transport 2.4 203771_s_at Interferon regulatory factor 2 (IRF2) Transcription factor, immune response, proliferation 2.3 203275_at Leukocyte immunoglobulin-like receptor 7 (LIR7) Immune response, antigen binding 2.3 211101_x_at Chondroitin 4-O-sulfotransferase 2 (CHST12) dermantan sulfate biosynthesis 2.1 218927_s_at Interferon, gamma-inducible protein 16 (IFI16) Transcriptional repressor, monocyte differentiation 2.1 208966_x_at Nedd4 binding protein 1 (N4BP1) Unknown 2.1 221867_at Eukaryotic translation initiation factor 3 subunit 11 (EIF3S11) Protein biosynthesis 0.5‡ 217719_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001 ‡ > 2 fold downregulation; p < 0.001

† Table 3.2: Gene Regulation in PBMCs following 18hr IFN-α treatment

Gene Function Fold Affy probe Chromosome 1 open reading frame 29 (C1orf29) Unknown 38.3 204439_at Vipirin (cig5) Catalytic activity 30.7 213797_at Interferon-induced protein with tetratricopeptide repeats 1 (IFIT1) Immune response 28.7 203153_at 2'-5'-oligoadenylate synthetase 3 (OAS3) Immune response, nucleic acid metabolism 12.3 218400_at DNA polymerase-transactivated protein 6 (DNAPTP6) Unknown 10.4 222154_s_at 28kD interferon responsive protein (IFRG28) Unknown 9.6 219684_at Interferon-induced protein 44 (IFI44) ** Antiviral response 9.6 214059_at Lectin, galactoside-binding, soluble, 9 (galectin 9) (LGALS9) Positive regulation of NFκB cascade 8 203236_s_at Hect domain and RLD5 (HERC5) ** Regulation of cyclin dependent protein kinase activity 7.5 219352_at A kinase (PRKA) anchor protein 2 (AKAP2) Enzyme binding 7 202760_s_at Tumor necrosis factor (ligand) superfamily, member 10 (TNFSF10) Apoptosis, immune response 5.3 202688_at GPI deacylase (PGAP1) Unknown 5.1 220576_at Metallothionein 1X (MT1X) Metal ion binding 4.5 208581_x_at Likely ortholog of mouse D11lgp2 (LGP2) ** DNA restriction 4.1 219364_at Lymphocyte-activation gene 3 (LAG3) Antigen binding 4.1 206486_at Tripartite motif-containing 5 (TRIM5) Transcription factor, zinc ion binding 4 210705_s_at Nuclear antigen Sp100 (SP100) Regulation of transcription 3.7 202863_at Zinc finger, CCHC domain containing 2 (ZCCHC2) Nucleic acid binding 3.2 219062_s_at Metallothionein 1F (MT1F) Metal ion binding 3.1* 213629_x_at KIAA0082 protein Unknown 2.9 212380_at Protein kinase D2 (PRKD2) Serine/threonine kinase 2.7 209282_at Spermidine/spermine N1-acetyltransferase (SAT) Acyltransferase activity 2.6 203455_s_at Endothelin converting enzyme 1 Unknown 2.4 201749_at PDH fnger protein 11 (PHF11) Regulation of transcription 2.3 221816_s_at Chromosome 20 open reading frame 18 (C20orf18) Protein ubiquitination 2.2 207713_s_at MHC class I polypeptide-related sequence B (MICB) Cell recognition 2.2 206247_at Myeloid differentiation primary response gene (88) (MYD88) Inflammatory response, regulation of NFκB cascade 2.1 209124_at Small nuclear ribonucleoprotein polypeptide N (SNRPN) RNA processsing, splicing 0.3‡ 221974_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001 ‡ > 2 fold downregulation; p < 0.001 ** Also upregulated in normal donor PBMCs

† Table 3.3: Gene Regulation in T cells following 18hr IFN-α treatment

Gene Function Fold Affy probe Interferon, alpha-inducible protein 27 (IFI27) Immune response 12.2 202411_at 2',5'-oligoadenylate synthetase 1 (OAS1) Immune response, nucleic acid metabolism 10.2 205552_s_at Hect domain and RLD 5 (HERC5) Ubiquitin-protein ligase 8.2 219863_at Myxovirus (influenza virus) resistance 2 (MX2) Immune response 7.4 204994_at 2'-5'-oligoadenylate synthetase 3 (OAS3) Immune response, nucleic acid metabolism 7.3 218400_at Likely ortholog of mouse D11lgp2 (LGP2) Helicase 4.9 219364_at Phospholipid scramblase 1 (PLSCR1) Phospholibid scramblase 4.1* 202430_s_at Interferon stimulated gene 20kDa (ISG20) ** Immune response, proliferation 4.1 33304_at 2'-5'-oligoadenylate synthetase 2 (OAS2) ** Immune response, nucleic acid metabolism 4* 204972_at Interferon regulatory factor 7 (IRF7) Transcription factor, antiviral response 3.8 208436_s_at Cytokine receptor CRL2 precursor (CLRF2) Interferon-class cytokine receptor 3.7 214329_x_at Interferon-induced protein with tetratricopeptide repeats 5 (IFIT5) Immune response 3.2 203595_s_at Interferon induced with helicase C domain 1 (IFIH1) Helicase, Deoxyribonuclease 3.2 219209_at KIAA1117 protein Unknown 3 213267_at Metallothionein 1X (MT1X) Metal ion binding 2.7* 204326_x_at Myotubularin related protein 9 (MTMR9) Myotubularin-related protein 2.7 204837_at Nuclear antigen Sp100 (SP100) Regulation of transcription 2.6* 202863_at Metallothionein 1H-like (MT1H) Metal ion binding 2.5 211456_x_at Promyelocytic leukemia (PML) Regulation of transcription 2.5 209640_at Tudor repeat associator with PCTAIRE 2 (TDRD7) Nucleic acid binding 2.5 213361_at Metallothionein 1F (MT1F) Metal ion binding 2.5* 217165_x_at Hypothetical protein from clone 643 (TFCP2) Transcription factor 2.4 209679_s_at Spermidine/spermine N1-acetyltransferase (SAT) Acyltransferase activity 2.4 213988_s_at Hypothetical protein FLJ10260 Unknown 2.2 219885_at Interferon induced transmembrane protein 3 (IFITM3) Immune response 2.2 212203_x_at KIAA0650 protein ATP binding 2.2 212569_at Metallothionein 2A (MT2A) ** Metal ion binding 2.2 212185_x_at Myristoylated alanine-rich protein kinase C substrate (MARCKS) Cell motility 2.1 201670_s_at Core 1 beta-3-galactosyltransferase (C1GALT) Galactosyltransferase 2 219439_at T-complex-associated-testis-expressed 1-like 1 (TCTEL1) Microtubule motor activity, cell cycle 2 201999_s_at Early growth response 3 (EGR3) Regulation of transcription, muscle development 0.4‡ 206115_at Plakophilin 4 (PKP4) Chromatin assembly 0.4‡ 212914_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001 ‡ > 2 fold downregulation; p < 0.001 ** Also upregulated in normal donor PBMCs

† Table 3.4: Gene Regulation in NK cells following 18hr IFN-α treatment

Gene Function Fold Affy probe Interferon stimulated gene 20kDa (ISG20) ** Immune response, proliferation 29.5 204698_at Ubiquitin specific protease 18 (USP18) ** Ubiquitin-dependent protein catabolism 23.4 219211_at Phorbolin 1 (PHRBN) mRNA editing 19.8 210873_x_at Chromosome 1 open reading frame 29 (C1orf29) Electron transport 16.8 204439_at CD69 antigen (CD69) T-cell activation 16 209795_at 2'-5'-oligoadenylate synthetase-like (OASL) ** Immune response, nucleic acid metabolism 14.6* 205660_at Guanylate binding protein 1, interferon-inducible, 67kDa (GBP1) Immune response 12.1 202269_x_at Interferon-induced protein 44 (IFI44) ** Antiviral response 12 214453_s_at Interferon induced transmembrane protein 1 (9-27) (IFITM1) Immune response, negative regulation of proliferation 11.6 201601_x_at Myxovirus (influenza virus) resistance 1, (MX1) Apoptosis, immune response 8.6 202086_at Apolipoprotein L, 3 (APOL3) Inflammatory response, lipid binding/transport 6 221087_s_at Hect domain and RLD5 (HERC5) ** Regulation of cyclin dependent protein kinase activity 5.9 219352_at 2'-5'-oligoadenylate synthetase 3 (OAS3) Immune response, nucleic acid metabolism 5.4 218400_at Torsin family 1, member B (torsin B) (TOR1B) Protein folding 4.7 209593_s_at Interferon regulatory factor 1 (IRF1) Transcription factor, immune response 4.6 202531_at Interleukin 15 receptor, alpha (IL15RA) Proliferation 4.5 207375_s_at A kinase (PRKA) anchor protein 2 (AKAP2) Enzyme binding 4.3 202760_s_at Lymphocyte cytosolic protein 2 (LCP2) Immune response 3.3 205269_at Chromosome 9 open reading frame 91 (C9orf91) Electron transport 3.6 221865_at Nedd4 binding protein 1 (N4BP1) ** Unknown 3.6 32069_at Likely ortholog of rat zinc-finger antiviral protein Unknown 3.4 220104_at Proteasome (prosome, macropain) subunit, beta type, 9 (PSMB9) Immune response, peptidolysis 3.4 204279_at Cullin 1 (CUL1) Apoptosis, cell cycle arrest (G1/S transition) 3.3 207614_s_at Transcription factor ets Unknown 3.3 221680_s_at Myotubularin related protein 9 (MTMR9) Myotubularin-related protein 3.2 204837_at Signal transducer and activator of transcription 1 (STAT1) Transcription factor, JAK-STAT pathway 3 97935_3_at Three prime repair exonuclease 1 (TREX1) ** DNA repair and replication 2.6 34689_at Tumor necrosis factor receptor superfamily, member 5 (TNFRSF5) Apoptosis, inflammatory response, B-cell proliferation 2.5 205153_s_at H.sapiens gene from PAC 747L4 Unknown 2.3 214838_at PHD finger protein 15 (PHF15) Regulation of transcription 2.3 212660_at Leukocyte immunoglobulin-like receptor 7 (LIR7) ** Immune response 2.2 211133_x_at Metallothionein 1F (MT1F) Metal ion binding 2.2 213629_x_at RAB9A, member RAS oncogene family (RAB9A) Protein transport, GTPase mediated signal transduction 2.2 221808_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001 ** Also upregulated in normal donor PBMCs

† Table 3.5: Gene Regulation in Monocytes following 18hr IFN-α treatment

Gene Function Fold Affy probe Biliverdin reductase B (flavin reductase (NADPH)) (BLVRB) Oxireductase 0.5 202201_at Chromosome 10 open reading frame 26 (C10orf26) Electron transport 0.5 202808_at Homo sapiens cDNA FLJ35653 fis, clone SPLEN2013690 Unknown 0.5 65472_at KIAA0232 gene product ATP binding 0.5 212441_at KIAA033333 Unknown 0.5 43511_s_at NADH dehydrogenase (ubiquinone) flavoprotein 1, 51kDa (NDUFV1) Electron transport 0.5 208714_at Ornithine aminotransferase (gyrate atrophy) (OAT) Amino acid metabolism, visual perception 0.5 201599_at Phosphatidylinositol-4-phosphate 5-kinase, type II, beta (PIP5K2B) Receptor signaling protein activity 0.5 201080_at Platelet/endothelial cell adhesion molecule (PECAM1) Cell motility 0.5 208981_at Protein predicted by clone 23733 (HSU79274) Unknown 0.5 204521_at Rho GDP dissociation inhibitor (GDI) beta (ARHGDIB) Immune response, negative regulation of cell adhesion 0.5 201288_at Ribosomal protein (RPL15) Structural constituent of ribosome 0.5 217266_at RNA binding motif protein 8A (RBM8A) mRNA splicing 0.5 213852_at Sphingosine-1-phosphate lyase 1 (SGPL1) Apoptosis, amino acid metabolism 0.5 212321_at Family with sequence similarity 13, member A1 (FAM13A1) Unknown 0.4* 202973_x_at Ribosomal protein L22 (RPL22) Structural constituent of ribosome 0.4 221726_at* Ribosomal protein L3 (HIV-1 TAR RNA binding protein B) (TARBP-B) Structural constituent of ribosome 0.4* 215963_x_at Chromosome 20 open reading frame 104 (C20orf104) Regulation of transcription 0.4 209422_at Dihydropyrimidine dehydrogenase (DPYD) Electron transport 0.4 204646_at F-box and WD-40 domain protein 2 (FBXW2) Ubiquitin-mediated protein degradation 0.4 209630_s_at Heterogeneous nuclear ribonucleoprotein A1 (HNRPA1) mRNA processing 0.4 213356_x_at Hypothetical protein CG003 (13CDNA73) Unknown 0.4 204072_s_at Ribosomal protein L3 (RPL3) Structural constituent of ribosome 0.4 201217_x_at Eukaryotic translation elongation factor 1 gamma (EEF1G) Protein biosynthesis, translation elongation factor 0.4 211927_x_at Translocase of outer mitochondrial membrane 20 homolog (TOMM20) Protein-mitochondiral targeting 0.4 200662_s_at dUTP pyrophosphatase (DUT) DNA replication 0.3 208956_x_at

* Significant gene upregulation was observed in multiple probe sets ‡ > 2 fold downregulation; p < 0.001

‡ Table 3.6: Gene Regulation in Monocytes following 18hr IFN-α treatment

51 Figure 3.2. Real Time PCR analysis of select genes identified by microarray analysis of immune subsets following 18 hour in vitro IFN-α stimulation. Real Time PCR was used to validate the expression of genes in PBMCs (IFIT2, ISG20, LY6E), T cells (IFIT1,

LAG3, OAS3), NK cells (IRF7, ISG20, MX2), and Monocytes (CD69, ISG20, OASL).

Data were expressed as the mean fold increase relative to baseline levels (PBS treatment).

All real time PCR data were normalized to the level of β-actin mRNA. Error bars denote the standard deviations of triplicate experiments.

52 Figure 3.2

40 PBMC

Fold increase vs. PBS Fold increase vs. 10

0 IFIT2 ISG20 LY6E

B 160 T cell

140

120

100

Fold increase vs. PBS 40

0 IFIT1 LAG3 OAS3

Continued… 53 Figure 3.2 Continued

14 NK cell

Foldincrease vs. PBS 4

0 IRF7 ISG20 MX2

140 Monocyte 120

100

Fold increaseFold vs. PBS 40

0 CD69 ISG20 OASL 54

Gene Function Fold Affy probe Interferon-induced protein with tetratricopeptide repeats 3 (IFIT3) Antiviral Response 12 204747_at 2'-5'-oligoadenylate synthetase-like (OASL) Double-stranded RNA binding, DNA binding 6.7* 205660_at Ubiquitin specific protease 18 (USP18) Protein Catabolism 5.8 219211_at Interferon-induced protein 44 (IFI44) Invasive growth, antiviral response 5.5 214453_s_at Interferon regulatory factor 7 (IRF7) Transcription factor, antiviral response 3.7 208436_s_at Nuclear factor, interleukin 3 regulated (NFIL3) Cellular Survival 3.7 203574_at Zinc finger, CCHC domain containing 2 (ZCCHC2) Nucleic acidic binding 3.2 219062_s_at Tumor necrosis factor receptor superfamily, member 6 (TNFRSF6) Apoptosis 3 215719_x_at Likely ortholog of mouse D11lgp2 (LGP2) DNA restriction 2.9 219364_at Metallothionein 1H (MT1H) Metal ion binding 2.8 206461_x_at Interferon stimulated gene 20kDa (ISG20) Cell proliferation, exonuclease activity 2.7* 33304_at GTP cyclohydrolase 1 (GCH1) Neurotransmitter 2.6 204224_s_at Serine palmitoyltransferase, long chain base subunit 2 (SPTLC2) Acyltransferase 2.5 216202_s_at Zinc finger CCCH type domain containing 1 (ZC3HDC1) Nucleic acidic binding 2.5 218543_s_at Hypothetical protein FLJ11286 Unknown 2.4 53720_at Three prime repair exonuclease 1 (TREX1) DNA repair 2.4 205875_s_at Transporter 1, ATP-binding cassette, sub-family B (TAP1) Antigen presentation 2.4 202307_s_at G protein-coupled receptor 43 (GPR43) Rhodopsin-like receptor 2.3 221345_at Interferon induced transmembrane protein 3 (IFITM3) Immune response 2.3 216565_x_at Promyelocytic leukemia (PML) Transcription factor, cell growth 2.2 211012_s_at Lectin, galactoside-binding, soluble, 3 binding protein (LGALS3BP) Cell adhesion, scavenger receptor 2.1 200923_at Metallothionein 1E (MT1E) Metal ion binding 2.1 212859_x_at SCO cytochrome oxidase deficient homolog 2 (SCO2) Electron transport 2.1 205241_at

* Significant gene upregulation was observed in multiple probe sets † > 2 fold upregulation; p < 0.001

† Table 3.7: Gene Upregulation 1hr following 20 MU/m2 IFN-α-2b

55 Figure 3.3. Real Time PCR analysis of select genes identified by microarray analysis of

PBMCs from melanoma patients receiving IFN-α. Real Time PCR was used to validate

the expression of representative genes (IRF7, OASL, TAP1). Data were expressed as the mean fold increase relative to baseline levels (pre-treatment). All real time PCR data were normalized to the level of β-actin mRNA. Patient gene expression estimates were

pooled and error bars denote the standard deviations of 7 melanoma patients.

56 Figure 3.3

10 Fold increaseFold vs. Pretreatment

0 IRF7 OASL TAP1

CHAPTER 4

INTERFERON-ALPHA-2B INDUCED SIGNAL TRANSDUCTION AND GENE

REGULATION IN PATIENT PBMCS IS NOT ENHANCED BY A DOSE

INCREASE FROM 5 MU/M2 TO 10 MU/M2

4.1 Introduction

The receptor for interferon-alpha (IFN-α) is widely expressed on both tumor cells and immune effector cells (7, 8). Binding of IFN-α to its receptor activates Janus kinase

1 (Jak1) and tyrosine kinase 2 (Tyk2), which in turn phosphorylate tyrosine residues within the cytoplasmic region of the IFN-α receptor. These phosphotyrosine residues provide docking sites for signal transducer and activation of transcription 1 (STAT1) and

STAT2, latent cytoplasmic transcription factors that are phosphorylated by the Janus kinases (9). The prototypical IFN-α signaling reaction results in the formation of

IFN-stimulated gene factor 3 (ISGF3), a DNA-binding complex that consists of STAT1α

(or STAT1β), STAT2, and a chaperon protein known as interferon regulatory factor 9

(IRF9) (10). ISGF3 subsequently translocates to the nucleus and binds to interferon-stimulated response elements located in the promoter regions of

58 IFN-responsive genes (11). This signaling event induces the expression of

immunoregulatory genes and largely determines the pattern of immune cell activation

following exposure to IFN-α (14-16).

Recombinant IFN-α is used to treat individuals with metastatic malignant

melanoma and produces clinical responses 10-15% of patients (20, 59, 143). High-dose

IFN-α is also employed as an adjuvant in patients who have undergone resection of high

risk lesions (nodal disease or primary tumors with Breslow thickness > 4 mm) (20, 59,

63-65). However, it has been difficult to determine the optimal dose of IFN-α for

melanoma patients or devise strategies to enhance its anti-tumor effects because its

cellular targets and mechanism of action are largely unknown. Although IFN-α can act

directly on melanoma cells to inhibit proliferation and upregulate the expression of MHC class I antigens, its stimulatory properties on effector cells of the immune system are thought to be most important for its anti-tumor activity (8, 87-92, 127, 128, 144-146).

Our analysis of IFN-α-induced JAK-STAT signal transduction in immune cells revealed a high degree of variability in the activation of STAT1 and STAT2 and reduced phosphorylation of these proteins at higher doses of IFN-α. We therefore hypothesized that intermediate doses of IFN-α would be the most effective in activating STAT proteins and inducing the expression of IFN-stimulated genes in patient PBMCs (116).

In order to test this hypothesis, we analyzed Jak-STAT mediated signal transduction and gene regulation in patients who received escalating doses of IFN-α-2b

59 in the setting of metastatic disease. The phosphorylation of STAT1 and STAT2 was

measured by intracellular flow cytometry and the induction of interferon-stimulated gene

(ISG) transcripts was evaluated by Real Time PCR and microarray analysis. These studies indicated that the PBMC response to 5 MU/m2 IFN-α was superior to that of

10 MU/m2 in the context of the present trial.

4.2 Material and Methods

Patients Blood Samples. Peripheral blood was obtained from melanoma patients

(3 females, 4 males) immediately prior to, and one-hour following subcutaneous (s.c.) administration of IFN-α-2b at 5 MU/m2 or 10 MU/m2. All samples were obtained at The

Ohio State University following informed consent under an IRB-approved protocol

(OSU 99H0348). PBMCs were isolated from peripheral blood (8 mL) via density

gradient centrifugation with Ficoll-Paque Plus (Amersham Pharmacia Biotech) and used

immediately in the assays described below.

Flow cytometric analysis of STAT1 and STAT2. The native and phosphorylated forms of

STAT1 (Tyr701) and STAT2 (Tyr690) were measured using an intracellular flow

cytometric assay as previously described, with modifications (116, 147). Rabbit

anti-human primary antibodies (Ab) (Cell Signaling Technology, Beverly, MA) were

employed in combination with a goat anti-rabbit Alexafluor 488-conjugated secondary

Ab (Molecular Probes, Eugene, OR).

Real Time PCR. Following TRIzol extraction and RNeasy purification, 2 μg of total

RNA was reverse transcribed and the resulting cDNA was used as a template to measure

gene expression by Real Time PCR using pre-designed primer/probe sets (Assays On

Demand; Applied Biosystems, Foster City, CA) and 2X Taqman Universal PCR Master

Mix (Applied Biosystems) according to manufacturer’s recommendations as previously

described (121). Pre-designed primer/probe sets for human β-actin were used as an

internal control in each reaction well (Applied Biosystems). Real Time PCR reactions

were performed in triplicate in a capped 96-well optical plate. Real Time PCR data was

analyzed using the ABI PRISM® 7900 Sequence Detection System (Applied

Biosystems).

cRNA Preparation and Array Hybridization. U133 Plus 2.0 GeneChips (Affymetrix,

Santa Clara, CA), which query approximately 47,000 human transcripts were used for

these analyses. The cRNA was synthesized as suggested by Affymetrix. Briefly, total

RNA from cells was prepared in TRIzol (Invitrogen) followed by RNeasy purification

(Qiagen, Valencia, CA). Double stranded cDNA was generated from 8 μg of total RNA using the Superscript Choice System according to the manufacturer’s instructions

(Invitrogen). Biotinylated cRNA was generated by in vitro transcription using the Bio

Array High Yield RNA Transcript Labeling System (Enzo Life Sciences Inc.,

Farmingdale, NY). The cRNA was purified using the RNeasy RNA purification kit

(Qiagen). cRNA was fragmented according to the Affymetrix protocol and the

61 biotinylated cRNA was hybridized to U133A or U74va2 microarrays (119). The arrays

were then scanned (Affymetrix GMS418) and analyzed (GenePix Pro 4.0) according to

Affymetrix protocols.

Microarray Data Analysis. Raw data were collected with a confocal laser scanner

(Hewlett Packard, Palo Alto, CA) and probe level data was analyzed using dChip version 1.3 (120). Quantile normalization was performed, and only perfect match probes

were used in computing the model-based expression indices (MBEIs). “Array outliers”

identified by dChip at the probe-set level were set to missing. The log2(MBEIs) were

then calculated and exported for further analysis.

A random effects model with repeated measures was used to determine if there

was evidence of a linear or non-linear dose effect on gene expression values. This test

was performed for each of the pre-selected 143 probe sets of interest (representing

65 genes). These probe sets were identified on the basis of preliminary studies from our

group as well as previous work from other groups (126, 132). In testing if there was a significant difference in expression between two dose levels, paired t-tests were used.

Type I error was protected by adjusting p-values using Holm’s method. Two-sided significance levels were set at α = 0.10.

Statistical analysis. Statistical comparisons between treatments for Figures 4.2-4 were performed by using analysis of variance (ANOVA) on the log-transformed data. For

Figures 4.3 and 4.4, random effects ANOVA was used to allow for correlations between 62 replicates from the same patient. Pairwise treatment differences were estimated from the

model; tests were performed using a two-sided α = 0.05 level of significance. All

analyses were performed using SAS v9.1 (SAS Institute, Inc., Cary, NC).

4.3 Results

Native and activated forms of STAT1 and STAT2 in PBMCs following escalation of

IFN-α-2b dose from 5 MU/m2 to 10 MU/m2. Peripheral blood mononuclear cells

(PBMCs) from metastatic melanoma patients (n = 7) were obtained immediately prior to

treatment, 1 hour following the subcutaneous (s.c.) administration of 5 MU/m2 IFN-α-2b

and 1 hour following the s.c. administration of 10 MU/m2 IFN-α-2b two weeks later.

Patients received a total of 6 injections of IFN-α-2b at a dose of 5 MU/m2 prior to dose

escalation. Following procurement, PBMCs were immediately isolated and analyzed for

the native and activated forms of STAT1 (activated residue Tyr701) and STAT2 (activated

residue Tyr690) by flow cytometry (Figure 4.1A-B and 4.1D-E, respectively). For 6 of 7 patients, STAT1 activation (P-STAT1) was not enhanced following the increase in

IFN-α-2b dose from 5 MU/m2 to 10 MU/m2 (Figure 4.1A). Of note, patient No.7, who

exhibited increased signal following administration of IFN-α at 10 MU/m2, experienced

significant psychological effects at this dose and was removed from the trial. Within

individual patients, levels of STAT1 (unphosphorylated form) did not diminish following

administration of IFN-α at 10 MU/m2 except for patient No. 3 (Figure 4.1B). It is not clear why STAT1 levels fell in this patient, however, the patient experienced disease

63 progression, proteinuria, and hypertension and was removed from the trial. Levels of

P-STAT2 in PBMCs were also measured by flow cytometry. In 6 of 7 patients, activation of STAT2 was higher or equivalent at the 5 MU/m2 dose of IFN-α as

compared to the 10 MU/m2 dose (Figure 4.1C). The overall levels of STAT2

(unphosphorylated form) did not decrease following administration of IFN-α at

10 MU/m2, except for patient No. 5 (Figure 4.1D).

Ratio of activated to total STAT1 and STAT2 in PBMCs following escalation of IFN-α-2b

dose from 5 MU/m2 to 10 MU/m2. To determine the efficiency of STAT activation in

PBMCs following IFN-α administration, the ratio of activated STAT protein to the total

level of the STAT protein was evaluated using an ANOVA test (α = 0.05). The use of

these ratios permits a better description of the ability of IFN-α-2b to activate STAT

proteins at each time point and accounts for altered baseline STAT levels. The ratio of P-

STAT1 to total STAT1 in response to s.c. administration of IFN-α-2b for the 5 MU/m2 dose was higher than that for the 10 MU/m2 dose but with borderline significance

(Figure 4.2A; p = 0.0617). Therefore, the activation of STAT1 in response to IFN-α-2b was equivalent at the 5 MU/m2 and 10 MU/m2 doses. The ratio of P-STAT2 to total

STAT2 was also calculated. In these 7 patients, the 5 MU/m2 dose led to a greater activation of STAT2 as compared to the 10 MU/m2 dose of IFN-α-2b (Figure 4.2B;

p = 0.0388).

64 Interferon-stimulated gene expression following IFN-α-2b administration. In order to

characterize the induction of interferon-stimulated genes (ISGs) following s.c.

administration of IFN-α-2b at the two doses, PBMCs from the above patient samples

were analyzed for the induction of specific ISGs by Real Time PCR. As with the activation of STAT1 and STAT2, the expression of IFIT1, IFIT-2, G1P2, and OAS3 was not enhanced following the increase in IFN-α-2b dose from 5 MU/m2 to 10 MU/m2

(Figure 4.3A-D; p > 0.05 for all genes). Thus, the induction of these genes in response to

IFN-α-2b was equivalent at the 5 MU/m2 and 10 MU/m2 doses.

Suppressor of cytokine signaling gene expression following IFN-α-2b administration.

Investigators have identified a family of proteins known as suppressors of cytokine

signaling (SOCS) that negatively regulate Jak-STAT signal transduction (96). The expression of SOCS1 and SOCS3 has been shown to mediate potent inhibitory effects on

IFN-α-stimulated signal transduction and gene regulation in several experimental systems (74, 110-112, 127). However, the effect of exogenous IFN-α on SOCS

expression in resting immune cell subsets has yet to be defined in the context of cancer

immunotherapy. As seen in Figure 4.4A, SOCS1 was induced to a greater degree

following administration of IFN-α at 10 MU/m2 in each patient (p = 0.0002). Similarly,

SOCS3 levels were induced to a greater degree (on average) with the higher dose of

IFN-α (Figure 4.4B; p = 0.0006). This data suggests that higher doses of IFN-α may induce the transcription of genes that negatively regulate the IFN-α response.

65 Gene Regulation in PBMCs from Metastatic Melanoma Patients Receiving Increasing

Doses of IFN-α. PBMCs from 4 melanoma patients undergoing immunotherapy with

escalating doses of IFN-α-2b were evaluated by microarray analysis (Patients #2, 5, 6,

and 7 from above). Sixty one known interferon-regulated genes represented by

143 probe sets were identified for microarray analysis (126, 132). Of these, 36 genes

showed a significant difference in expression after treatment with IFN-α at 10 MU/m2 compared to baseline expression (Holm’s adjusted P < 0.10; Table 4.1). Only five of these 36 genes (ICAM1, IL1B, JUN, JUNB, SOCS3) were down-regulated after treatment with 10 MU/m2 IFN-α. For ICAM1, IL1B, and JUN there was a >2 fold

decrease in expression in response to this dose of IFN-α. The remaining 31 genes were

significantly upregulated after treatment with IFN-α-2b at 10 MU/m2. There was a

>2 fold increase in expression as compared to baseline for 22 of these genes.

Importantly, for the 36 genes showing a significant change in expression following

administration of 10 MU/m2 IFN-α-2b, only one gene was induced (or suppressed) to a greater extent by IFN-α at 10 MU/m2 as compared to 5 MU/m2 (LGALS3BP, 0.99 fold

and 1.46 fold induction at 5 MU/m2 and 10 MU/m2, respectively ; Table 4.1A). This

gene is a carbohydrate binding protein that binds to naïve neutrophils and acts as an

adhesion molecule that can mediate the initial adhesion to the endothelium (148). For

remaining 35 genes, expression levels were similar between the 5 MU/m2 and 10 MU/m2 doses of IFN-α (Table 4.1B, C). In fact 12 genes were equivalently upregulated

(CXCL10, IFI44, IFIT1, IFIT3, IFIT5, MX1, OAS1, OAS2, PML) or down-regulated

(ICAM1, IL1B, JUN) by IFN-α at the 5 MU/m2 and 10 MU/m2 dose levels (non-linear

66 dose response test, p < 0.05; Table 4.1B). These results were validated in patient PBMCs

by measuring the expression of several notable genes (IFIT1, OAS1, CXCL10,

LGALS3BP) by Real Time PCR (Figure 4.5). These Real Time PCR results

demonstrated the expected results as CXCL10, IFIT2, and OASL showed similar

expression levels following both doses of IFN-α, while levels of LGASL3BP transcripts

were higher after therapy with IFN-α 10 MU/m2.

4.4 Discussion

The present study demonstrated that IFN-α-induced signal transduction and gene transcription in patient PBMCs was either greater or statistically equivalent at a dose of

5 MU/m2 as compared to 10 MU/m2. Specifically, levels of P-STAT1 and P-STAT2

were not induced to a greater degree with the higher dose of IFN-α. In addition,

inhibitors of IFN-α signaling, SOCS1 and SOCS3, were induced to a greater degree by

the higher IFN-α dose, suggesting a potential mechanism for the inhibition of signal

transduction at high doses of IFN-α. Although the induction of SOCS transcript was

subtle, it has been shown that low levels of SOCS protein is capable of inhibiting IFN-α

signal transduction (111). It is important to note the order in which the two doses of

IFN-α were administered in this trial. It is possible that if IFN-α at 10 MU/m2 was administered first followed later by the 5 MU/m2 dose that superior results would have

been obtained for the higher dosage. However, the dose escalation schema employed in

this study is likely to be better tolerated by patients.

Tumor cells routinely express functional IFN-α receptors and this cytokine can exert anti-proliferative, anti-angiogenic, and pro-apoptotic effects on tumor cells in vitro

(73-77). We have previously demonstrated that ex vivo treatment of patient tumors with clinically relevant concentrations of IFN-α consistently led to activation of STAT1 and

STAT2 (83). Other groups have observed that some IFN-α-resistant human melanoma cell lines exhibited defects in specific Jak-STAT intermediates, which when reversed, led to the recovery of in vitro sensitivity to IFN-α (84-86). Of note, the most common defect appeared to be the loss of STAT1. However, we have found that tumor expression of

STAT1 and STAT2 did not correlate with effectiveness of adjuvant IFN-α. We identified a large cohort of high-risk patients with loss of STAT1 in their tumor that exhibited prolonged survival in response to adjuvant IFN-α, while other patients who had normal tumor expression of Jak-STAT proteins recurred after just a few months of IFN therapy (128). Using STAT1-deficient mice and STAT1-deficient melanoma tumor cells, our group has conclusively demonstrated the anti-tumor effects of IFN-α are dependent on STAT1 signaling within immune cells and not the cancer cell (18). Based on these data, it is likely that immunologic mechanisms play a key role in the anti-tumor actions of

IFN-α. We propose that continued analysis of the immune response to exogenous IFN-α will lead to important improvements in this treatment regimen.

Currently, there is much debate as to what is the most effective dose of adjuvant

IFN-α therapy for patients with resected melanoma with a high-risk of recurrence (149).

68 Randomized clinical trials have evaluated very low-dose (0.5–1 MU/m2), low dose

(3 MU/m2), intermediate dose (5–10 MU/m2), and high dose IFN-α -2b (> 10 MU/m2)

(63, 65, 69, 150-153). A randomized clinical trial of high dose IFN-α therapy

(20 MU/m2 i.v. per day 5 days/week for four weeks followed by 10 MU/m2 s.c. thrice

weekly for 48 weeks) versus treatment with the GM2-KLH/QS-21 vaccine demonstrated significantly improved relapse free survival (p = 0.015) and overall survival benefit

(p = 0.009) (65). These results imply that high-dose IFN-α has activity as an adjuvant therapy. However, high dose IFN-α therapy is associated with a unique spectrum of toxicities that can be problematic for a subset of patients (69, 150). In an attempt to establish the efficacy of lower doses of IFN-α, Cascinelli et al. conducted a trial in which

patients with high risk lesions were randomized to receive 3 MU of IFN-α-2a s.c. thrice

weekly for three years versus observation alone. However, this dose of IFN-α did not lead to improved overall survival or relapse-free survival (63, 151, 152). In European

Organization for Research and Treatment of Cancer trial number 18952 patients who were treated with 5 MU of IFN-α thrice weekly for 2 years had slightly better distant metastasis-free interval and overall survival rates than those who received no treatment, but this difference was not statistically significant (153). Taken together, these studies indicate that that 10 MU/m2 dose, while possibly effective, is relatively toxic in a subset of patients, whereas lower doses, while better tolerated, appear to be ineffective or marginally effective. The question remains whether indeterminate doses might exert anti-tumor effects and be better tolerated. The present report supports this concept.

69 The pattern of IFN-α-induced intracellular signaling and gene regulation was found to be patient-specific and highly variable. Whitney et al. utilized microarray analysis to explore the extent of inter-individual variation in gene expression within unstimulated PBMCs of healthy donors (n = 75). The greatest degree of individual variation occurred within a cluster of 15 genes know to be IFN-responsive (35). The authors acknowledged this might be attributed to subclinical or recent infection in some donors, however this observation was subsequently documented in other studies (36).

Furthermore, basal levels of some cytokines (IL-4, IL-6, IL-10) can vary significantly among healthy individuals (36) and particularly among patients with advanced malignancy (37). These altered cytokine profiles in cancer patients may contribute to a portion of the variable gene expression that is observed in cancer patients following cytokine immunotherapy.

The present study demonstrated that increasing doses of IFN-α did not mediate enhanced signal transduction and gene stimulation in melanoma patient immune cells.

STAT activation and ISG transcription was variable suggesting that patient-specific dosing may be possible.

70 4.5 Tables and Figures

Figure 4.1. Native and activated forms of STAT1 and STAT2 levels in PBMCs

following in vivo escalation of IFN-α-2b administration. PBMCs from seven melanoma

patients were obtained immediately prior to and 1 hour following administration of

IFN-α (5 MU/m2 s.c.) Two weeks later, patients were administered an escalated dose of

IFN-α (10 MU/m2 s.c.). PBMCs were analyzed freshly for the (A) P-STAT1 (Tyr701),

(B) total STAT1, (C) P-STAT2 (Try690), and (D) STAT2 by flow cytometry. Mean

specific fluorescence (Fsp) is illustrated on the y-axis. Appropriate isotype control

antibodies were used to determine background staining. All flow cytometric data were

derived from at least 10,000 events gated on the lymphocyte populations determined by

light scatter properties (forward scatter vs. side scatter).

71 Figure 4.1 A

12 Pretreatment 2 5 MU/m 10 MU/m 2 10

Fsp (P-STAT1) 4

0 1234567 Patients

250 Pretreatment 5 MU/m 2 10 MU/m 2 200

150

100 Fsp (STAT1)

0 1234567 Patients Continued…

72 Figure 4.1 Continued C

5 Pretreatment 5 MU/m2 2 10 MU/m 4

2 Fsp (P-STAT2)

0 1234567 Patients

140 Pretreatment 5 MU/m2 2 120 10 MU/m

100

60 Fsp (STAT2)

0 1234567 Patients

73 Figure 4.2. Ratios of native and activated forms of STAT1 and STAT2 levels in PBMCs following in vivo escalation of IFN-α-2b administration. To determine the efficiency of

STAT activation, the ratio of activated STAT to total STAT were evaluated. Ratios of

(A) phosphorylated and unphosphorylated STAT1 and (B) STAT2 protein levels as depicted in Figure 4.1 have been place in a ratio format. These ratios account for

IFN-α-2b’s maximum potential for activation of STAT for each time point and limit the bias of altered total STAT levels.

74 Figure 4.2 A

0.4 Pretreatment 5 MU/m2 10 MU/m 2

0.3

0.2

0.1 Fsp (P-STAT1) / (STAT1)

0.0 1234567

B Patient

0.20 Pretreatment 5 MU/m2 10 MU/m 2

0.15

0.10

0.05 Fsp (P-STAT2) / (STAT2)

0.00 1234567 Patient

75 Figure 4.3. Interferon-stimulated gene expression following IFN-α-2b administration.

In parallel to flow cytometry, PBMCs from the seven patient samples were immediately lysed in TRIzol reagent (Invitrogen) to characterize the induction of interferon-stimulated genes. RNA was isolated, converted to cDNA and analyzed for interferon-stimulated genes by Real Time PCR using primers specific for (A) IFIT1, (B) IFIT2, (C) G1P2, and

(D) OAS3. Data were expressed as the mean fold increase relative to baseline levels

(pretreatment). All Real Time PCR data were normalized to the level of β-actin mRNA

(housekeeping gene). Error bars denote the standard deviations of triplicate wells.

76 Figure 4.3 A

160 IFIT1 5 MU/m2 10 MU/m 2 140

120

100

40 Fold expression vs. Pretreatment vs. expression Fold 20

0 1234567 Patient B

200 IFIT2 5 MU/m2 180 10 MU/m 2

160

140

120

100

40 Fold increase vs. Pretreatment Fold increase vs.

0 1234567 Patient Continued…

77 Figure 4.3 Continued C

2 5 MU/m 2 160 G1P2 10 MU/m

140

120

100

40 Fold vs. Pretreatment expression 20

0 1234567 Patient

120 OAS3 5 MU/m 2 110 10 MU/m 2

100

Fold increase vs. Pretreament Fold increase 20

0 1234567 Patient

78 Figure 4.4. SOCS1 and SOCS3 expression following IFN-α-2b administration. In addition to ISG stimulation, PBMCs were characterized for the expression of

(A) suppressors of cytokine signaling (SOCS) 1 and (B) SOCS3 transcript levels before and after IFN-α administration by Real Time PCR. Data were expressed as the mean fold increase relative to baseline levels (pretreatment). All Real Time PCR data were normalized to the level of β-actin mRNA (housekeeping gene). Error bars denote the standard deviations of triplicate wells.

79 Figure 4.4 A

8 2 SOCS1 5 MU/m 2 10 MU/m

2 Fold increase vs. pretreatment

0 1234567 Patients

B 5 SOCS3 2 5 MU/m 2 10 MU/m

1 Fold increase vs. pretreatment increase vs. Fold

0 1234567 Patients

Table 4.1

A) Linear Response: Gene upregulation in patient PBMCs is higher in the 10 MU/m 2 treatment vs. 5 MU/m2 IFN-α-2b administration (P < 0.01, alpha < 0.1, FDR < 10%) Fold Fold Adjusted Gene Function 5MU/m 2 10MU/m 2 Affy probe P (Holm )

Cell adhesion, scavenger receptor Lectin, galactoside-binding, soluble, 3 binding protein (LGALS3BP) 0.99 1.46 200923_at 0.0174017

B) Non-linear Dose Response: Gene upregulation does not correlate directly to dose in patient PBMCs after 5 MU/m2 and 10 MU/m2 IFN-α-2b administration (P < 0.01, alpha < 0.1, FDR < 10%)

Fold Fold Adjusted Gene Function 5MU/m 2 10MU/m 2 Affy probe P (Holm ) Antiviral activity, nucleic acid metabolism 2',5'-oligoadenylate synthetase 1 (OAS1) 5.11 5.25 202869_at* 0.0376 Antiviral activity, nucleic acid metabolism 2',5'-oligoadenylate synthetase 2 (OAS2) 2.74 2.99 204972_at* 0.0734 Chemokine (C-X-C motif) ligand 10 (CXCL10) Chemotaxis, monocyte stimulation 22.13 32.82 204533_at 0.0594 Intercellular adhesion molecule 1 (ICAM1) Cellular adhesion 0.14 0.12 202637_s_at 0.0807 Interferon-induced protein 44 (IFI44) Invasive grow th, antiviral response 12.91 14.80 214059_at 0.0025 Interferon-induced protein with tetratricopeptide repeats 1 (IFIT1) Immune response 24.08 27.48 203153_at* 0.0213 Interferon-induced protein with tetratricopeptide repeats 3 (IFIT3) Immune response 12.34 16.22 204747_at 0.0341 Interferon-induced protein with tetratricopeptide repeats 5 (IFIT5) Immune response 6.08 6.77 203595_s_at* 0.0025 Interleukin 1, beta (IL1B) inflammatory response, cell proliferation 0.10 0.14 205067_at* 0.0287 Myxovirus (influenza virus) resistance 1 (MX1) Antiviral activity 3.16 4.04 202086_at 0.0641 Promyelocytic leukemia (PML) Transcription factor, cell grow th 1.78 1.48 211014_s_at 0.0692 V-jun sarcoma virus 17 oncogene homolog (JUN) Regulation of transcription 0.12 0.21 201464_x_at* 0.0240

* Significant gene upregulation w as observed in multiple probe sets

Continued…

81 Table 4.1 Continued

C) Dose Response: Genes significantly upreguated with Interferon (but do not follow a significant linear or non-linear response) in patient PBMCs after 5 MU/m2 and 10 MU/m2 IFN-α- 2b administration (P < 0.01, alpha < 0.1, FDR < 10%)

Fold Fold Adjust 5MU/ 10MU/ ed P Gene Function m2 m2 Affy probe (Holm)

2',5'-oligoadenylate synthetase 3 (OAS3) Antiviral activity, nucleic acid metabolism 2.06 2.02 218400_at* 0.0218

2'-5'-oligoadenylate synthetase-like (OASL) Antiviral activity, nucleic acid metabolism 4.07 5.09 205660_at* 0.0020

Guanylate binding protein 1, interferon-inducible Immune response (GBP1) 2.02 2.55 202269_x_at 0.0952

Intercellular adhesion molecule 1 (ICAM1) Cellular adhesion 0.18 0.15 202638_s_at* 0.0045

Interferon induced transmembrane protein 1 (IFITM1) Immune response 1.15 1.35 201601_x_at* 0.0629

Interferon induced transmembrane protein 3 (IFITM3) Immune response 1.18 1.31 212203_x_at 0.0765

Interferon induced with helicase C domain 1 (IFIH1) Helicase, Deoxyribonuclease 4.15 5.82 216020_at* 0.0032

Transcription factor, immune response, Interferon regulatory factor 2 (IRF2) proliferation 2.12 2.05 203275_at 0.0461

Transcription factor, antiviral Interferon regulatory factor 7 (IRF7) response 2.25 2.39 208436_s_at 0.0082

Interferon, gamma-inducible protein 16 (IFI16) Transcriptional repressor, monocyte differentiation 2.01 2.27 208966_x_at* 0.0477

Interferon-induced protein 44 (IFI44) Invasive growth, antiviral response 3.33 4.23 214453_s_at 0.0027

Jun B proto-oncogene (JUNB) Regulation of transcription 0.58 0.61 201473_at 0.0495

Likely ortholog of mouse D11lgp2 (LGP2) Helicase 1.91 2.21 219364_at 0.0022

Myxovirus (influenza virus) resistance 2 (MX2) Immune response 1.71 1.89 204994_at 0.0585

N-myc (and STAT) interactor (NMI) Regulation of transcription 1.70 1.90 203964_at 0.0408

Nuclear antigen Sp100 (SP100) Regulation of transcription 2.54 3.14 237426_at 0.0536

Signal transducer and activator of transcription 1 (STAT1) Transcription factor, JAK-STAT pathway 3.43 3.86 232375_at 0.0843

Signal transducer and activator of transcription 2 (STAT2) Transcription factor, JAK-STAT pathway 1.62 1.58 225636_at 0.0408

Suppressor of cytokine signaling 3 (SOCS3) Inhibitor of JAK-STAT pathway 0.69 0.54 206360_s_at 0.0568

Transporter 1, ATP-binding cassette, sub-family B (TAP1) Antigen presentation 1.52 1.81 202307_s_at 0.0534

Tumor necrosis factor (ligand) superfamily, member 10 (TNFSF10) Apoptosis, immune response 4.38 4.85 202688_at* 0.0881

Ubiquitin-dependent protein Ubiquitin specific protease 18 (USP18) catabolism 7.49 14.73 219211_at 0.0006

V-jun sarcoma virus 17 oncogene homolog (JUN) Regulation of transcription 0.17 0.19 201465_s_at 0.0343

* Significant gene upregulation was observed in multiple probe sets

Figure 4.5. Real Time PCR Validation of Microarray Data from PBMCs of Patients

receiving escalation doses of IFN-α. The expression of representative genes was

validated by Real Time PCR analysis including OAS1, IFIT1, CXCL10 (non-linear with

respect to in 5 MU/m2 and 10 MU/m2 doses), and LGALS3BP (higher in 10 MU/m2).

Data were expressed as the mean fold increase relative to baseline levels (pretreatment).

All Real Time PCR data were normalized to the level of β-actin mRNA (housekeeping gene). Error bars denote the standard deviations of triplicate wells.

83 Figure 4.5

50 IFIT1

Fold change vs. Pretreatment Fold change vs. 10

0 55 MU/mMU/m^22 10 10 MU/m^2MU/m2 Interferon-alpha administration

14 OAS1

4 Fold change vs. Pretreatment Fold change vs. 2

0 55 MU/mMU/m^22 10 10 MU/mMU/m^22 Interferon-alpha administration Continued…

84 Figure 4.5 Continued

160 CXCL10

140

120

100

40 Fold change vs. Pretreatment

0 5 MU/m^2MU/m2 10 10 MU/mMU/m^22 Interferon-alpha administration

D 4 LGALS3BP

1 Fold change vs. Pretreatment vs. change Fold

0 55 MU/mMU/m^22 10 10 MU/m MU/m^22 Interferon-alpha administration

CHAPTER 5

IFN-alpha-Induced Signal Transduction, Gene Expression, and Anti-Tumor

Activity of Immune Effector Cells Are Negatively Regulated by Suppressor of

Cytokine Signaling Proteins

5.1 Introduction

Recombinant interferon-alpha (IFN-α is used to treat patients with metastatic malignant melanoma and is associated with an overall response rate of 10-15% (20, 59,

60). High-dose IFN-α is also employed as an adjuvant in patients who have undergone resection of high-risk lesions (nodal disease or primary tumors of Breslow thickness

> 4 mm) (20, 59, 63-65). However, it has been difficult to determine the optimal dose of

IFN-α for melanoma patients or devise strategies to enhance the anti-tumor effects of

IFN-α because its cellular targets and mechanism of action are largely unknown.

Although exogenous administration of IFN-α can act directly on melanoma cells to inhibit proliferation and upregulate the expression of MHC class I antigens, its stimulatory properties on effector cells of the immune system are thought to be critical for its anti-tumor activity (8, 18, 146, 154). Dunn et al. have also shown that

86 endogenously produced IFN-α is required for the prevention of carcinogen-induced

tumors and that host immune effector cells are critical targets of IFN-α during the

development of protective anti-tumor responses (93).

The receptor for IFN-α is widely expressed on both tumor cells and immune

effector cells (7, 8). Binding of IFN-α to its receptor activates Janus kinase 1 (Jak1) and

tyrosine kinase 2 (Tyk2), which in turn phosphorylate tyrosine residues within the

cytoplasmic region of the receptor. These phosphotyrosine residues provide docking

sites for signal transducer and activation of transcription 1 (STAT1) and STAT2, latent

cytoplasmic transcription factors that are phosphorylated by the Janus kinases (9). The

prototypical IFN-α signaling reaction results in the formation of IFN-stimulated gene

factor 3 (ISGF3), a DNA-binding complex that consists of STAT1α (or STAT1β),

STAT2, and interferon regulatory factor 9 (IRF9) (10). ISGF3 subsequently translocates

to the nucleus and binds to interferon-stimulated response elements located in the promoter regions of IFN-responsive genes (11). These signaling events induce the

expression of a variety of immunoregulatory genes and largely determine the pattern of immune cell activation following exposure to IFN-α (14-18). We have previously demonstrated a high degree of variability in the formation of P-STAT1 in patient immune effector cells following IFN-α-2b immunotherapy and have shown that Jak-STAT signal transduction is down-regulated at higher dose levels of IFN-α (116). These data suggested that negative regulatory pathways might influence signal transduction and gene expression in human immune cells following exposure to IFN-α.

Investigators have identified a family of proteins termed suppressors of cytokine signaling (SOCS) that negatively regulate Jak-STAT signal transduction (96). The SOCS family of proteins consists of eight members, including SOCS1-SOCS7 and cytokine inducible SH2-containing protein (CIS). All SOCS proteins have a central SH2 domain that allows them to bind to phosphotyrosine residues in cytokine receptors or Janus kinases (97-102), and a C-terminal SOCS box domain that may function to target

SOCS-bound proteins for proteasomal degradation (103, 104). SOCS1 and SOCS3 also contain a kinase inhibitory region (KIR) that is able to inhibit Jak kinase activity (101,

102, 105). The expression of SOCS1 and SOCS3 has been shown to mediate potent inhibitory effects on IFN-α-stimulated signal transduction and gene regulation in several experimental systems (74, 110-112), however, the effect of exogenous IFN-α on SOCS expression in resting immune cell subsets has yet to be defined in the context of cancer immunotherapy. We hypothesized that SOCS proteins exert a negative effect on

IFN-α-induced immune activity.

We now demonstrate that SOCS1-3 and CIS are rapidly induced in whole

PBMCs, T cells, and NK cells at the transcript and protein level following treatment with

IFN-α. The IFN-α-induced activation of STAT1 and the subsequent regulation of interferon-stimulated genes (ISGs) was significantly reduced in SOCS1 and SOCS3 over-expressing lymphoid cell lines, whereas inhibition of SOCS1 and SOCS3 activity by siRNA knockdown led to an enhanced response. Furthermore, IFN-α-induced signal

88 transduction, gene regulation, and anti-tumor activity were enhanced in SOCS1- and

SOCS3-deficient mice. These results demonstrate that SOCS proteins are critical negative regulators of the immune response to exogenous IFN-α.

5.2 Materials and Methods

Reagents and cell lines. Recombinant human (hu) IFN-α-2b (specific activity of

2 x 108 IU/mg) was purchased from Schering-Plough, Inc. (Kenilworth, NJ). Human growth hormone (huGH) was purchased from Apollo Cytokine Research (San Francisco,

CA). IFN-A/D (specific activity of 1.1 x 108 U/mg, PBL Biomedical Labs, Minneapolis,

MN) was utilized in all murine tumor challenge experiments and was administered via

the i.p. route at a dose of 2 x 104 U per day. IFN-A/D, or Universal Type I IFN, is a human hybrid recombinant type I interferon constructed from recombinant hu-IFN-αA

and hu-IFN-αD. It is active on a wide variety of mammalian cells (PBL Biomedical

Labs). The human melanoma cell lines were gifts from Dr. S. Ferrone (Roswell Park

Cancer Institute, Buffalo, NY) and were cultured in RPMI-1640 with 10% FBS, and antibiotics (155). The Jurkat T cell lymphoma cell line (Clone E6-1) was obtained from the American Type Culture Collection (Manassas, VA). The murine melanoma cell line

JB/MS was obtained from Vincent Hearing (National Cancer Institute, Bethesda,

Maryland, USA) and grown as an adherent monolayer in DMEM supplemented with 10%

FBS, sodium bicarbonate, 4 mM L-glutamine, 1% vitamins, 1% sodium pyruvate, 1%

nonessential amino acids, and antibiotics (156).

89 Antibody for depletion of CD8+ T cells. Rat anti-mouse CD8 antibody (clone 2.43) was

purchased from the National Cell Culture Center (NCCC; Minneapolis, MN). For

depletion of CD8+ T cells, 100 μg antibody was injected i.p. on days -3, -1, +1, +3, and every 4 days thereafter in relation to the tumor challenge. Rat IgG was used as a control.

CD8 depletion was confirmed by flow cytometric analysis of PBMCs obtained from venous blood.

Animals. SOCS1-deficient mice die of overwhelming inflammation unless the endogenous release of IFNγ is eliminated (157). SOCS3-/- is an embryonic lethal

mutation (158). Therefore, SOCS1+/- IFNγ-/- and SOCS3+/- mouse breeding pairs

(C57BL/6, Sv129 background) were obtained from Dr. James Ihle (St. Jude Children’s

Research Institute) and bred to produce their SOCS competent (SOCS1+/+ IFNγ-/- and

SOCS3+/+) and SOCS-deficient (SOCS1+/- IFNγ-/-, SOCS1-/- IFNγ-/-, and SOCS3+/-) counterparts. Genotyping of SOCS1 and SOCS3 mice was performed as previously described (157, 158). Mice of 5-6 weeks of age were used in all experiments. Spleens from male and female mice of each genotype were removed aseptically and mechanically dispersed through 70 μM cell strainers. Splenocytes were washed with PBS and 5%

FBS, pelleted by centrifugation, and resuspended in RPMI-1640 and 10% FBS. All experiments were performed in compliance with the guidelines of the Institutional

Laboratory Animal Care and Use Committee of The Ohio State University

(protocol 2004A0151).

90 Murine tumor models. An intraperitoneal model of murine malignant melanoma was

used to test the anti-tumor effects of IFN-A/D in SOCS-deficient mice (156). Mice were injected i.p. with 106 JB/MS melanoma tumor cells and randomly selected to receive

either PBS or IFN-A/D (2 x 104 U per day, i.p.). Mice were examined daily, and those

exhibiting signs of progressive disease were euthanized via CO2 inhalation. Survival

experiments used at least six mice per group. Because IFN-A/D treatment was effective at protecting against lethal tumor challenge in SOCS1+/- and SOCS1-/- mice, this model was modified to permit the outgrowth of tumors for immunohistochemical analysis. For these studies, mice were injected i.p. on Day 0 with 106 JB/MS melanoma cells.

Beginning on Day 7, mice were treated i.p. for three days with IFN-α (2 x 104 U) or PBS.

Tumors were harvested on Day 10, formalin fixed, embedded in paraffin, and sectioned.

Isolation of immune subsets. Peripheral blood mononuclear cells (PBMCs) were isolated

from source leukocytes of healthy adult donors (American Red Cross, Columbus, OH) or

from the peripheral blood of patients receiving high dose IFN-α-2b (Ohio State

University IRB-approved protocol 99H0348) via density gradient centrifugation with

Ficoll-Paque Plus (Amersham Pharmacia Biotech, Uppsala, Sweden). Lymphocyte subsets were enriched for individual cell populations (CD3+/CD56- T cells and

CD56+/CD3- NK cells) by negative selection with the appropriate Rosette Sep reagents

(Stem Cell Technologies, Vancouver, British Columbia) per the manufacturer’s

recommendations. Enriched cell populations were cultured in RPMI-1640 media

supplemented with 10% human AB serum (HAB, Pel-Freez Clinical Systems, LLC®,

Brown Deer, WI). Cell purity was routinely >95% as determined by flow cytometry. 91

Real Time PCR. Following TRIzol extraction (Invitrogen) and RNeasy purification

(Qiagen), total RNA was quantitated via the RiboGreen RNA Quantitation Kit

(Molecular Probes, Eugene, OR), and reverse transcribed as previously described (159).

The resulting cDNA was used as a template to measure gene expression by Real Time

PCR using pre-designed primer/probe sets (Assays On Demand, Applied Biosystems,

Foster City, CA) and 2X Taqman Universal PCR Master Mix according to the

manufacturer’s recommendations as previously described (160). Pre-designed

primer/probe sets for human β-actin were used as an internal control in each reaction well

(Applied Biosystems). Real Time PCR data was analyzed using the Sequence Detector

software version 1.6.

Flow cytometric analysis of phosphorylated STAT1 (P-STAT1). Phosphorylation of

STAT1 at Tyr701 was measured using an intracellular flow cytometric assay as

previously described, with modifications (116, 147). A rabbit anti-P-STAT1 (Tyr701)

primary antibody (Ab) (Cell Signaling Technology, Beverly, MA) was employed in

combination with a goat anti-rabbit Alexafluor 488-conjugated secondary Ab (Molecular

Probes, Eugene, OR). For assays involving enhanced green fluorescent protein (EGFP)

positive cells, a goat anti-rabbit APC-conjugated secondary Ab was utilized (Santa Cruz

Biotechnology, Santa Cruz, CA).

Immunoprecipitation and immunoblot analysis. Following treatment, cells were harvested and lysed in TN1 lysis buffer (125 mM NaCl, 50 mM Tris pH = 8, 10 mM

92 EDTA, 10 mM Na4P2O7•10H20, 10 mM NaF, 1% Triton X-100, 3 mM Na3VO4, 5 µg aprotinin and leupeptin) and centrifuged at 10,000 rpm. For immunoprecipitation experiments, supernatants were collected and treated with 5 μg of the appropriate antibody, processed per manufacturer recommendations (Abcam, Cambridge, UK), and then subjected to immunoblot analysis as previously described (161).

Design of SOCS constructs. Over-expression of SOCS1, SOCS2, and SOCS3 proteins was achieved using the PINCO retroviral vector as previously described (162). Briefly,

PBMCs from a normal healthy donor were stimulated with IFN-α and RNA was isolated and converted to cDNA for use as a template in a PCR reaction to isolate the human

SOCS1, SOCS2, and SOCS3 genes. Primers were designed to incorporate

1) AT overhangs; 2) 5’ BamHI and 3’ EcoRI restriction sites; 3) A C-terminal myc-tag and 4) a Kozak sequence prior to the initiation codon. SOCS transcripts were amplified using PCR according to the following schema: 94°C for 5 min; 30 cycles of 94°C for

30 sec, 55°C for 30 sec, 72°C for 1 min, followed by a final extension at 72°C for 7 min.

PCR products were gel purified (Qiagen Gel Extraction Kit, Valencia, CA) and ligated into the PCR2.1 vector by TA Cloning according to manufacturer’s recommendations

(Invitrogen). Following sequence verification, plasmids were digested with BamHI and

EcoRI to remove SOCS-encoding dsDNA and these inserts were then ligated into the

BamHI EcoRI-digested PINCO expression plasmid. Ligation reactions were transformed into Top 10 E. coli and plasmid DNA was sequenced with the following primers

(PINCO Fw. 5'-accttacacagtcctgctga-3'; PINCO Rev. 5’-tgaactaatgaccccgtaatt-3’) to

93 verify each SOCS-expressing construct. Following sequence confirmation, each

construct was prepared for virus production by endotoxin-free maxiprep (Qiagen).

Generation of SOCS-expressing retroviral constructs and transduction of Jurkat cells.

SOCS-over-expressing PINCO retroviral constructs were generated by transient

transfection of the Phoenix-Ampho packaging cell line as previously described (162).

The Jurkat cell line was transduced with SOCS-over-expressing retroviral constructs as

previously described by Becknell et al. (162). Infection efficiency was determined by

flow cytometry for EGFP expression. Cells were sorted based on EGFP positivity and

employed as described.

SOCS down-regulation by RNA interference. High-purity siRNA oligonucleotides that

target SOCS1 (5’-ctggttgttgtagcagcttaa-3’) and SOCS3 (5’-tcgggagttcctggaccagta-3’)

sequences were purchased from Qiagen (Valencia, CA). An oligonucleotide that does not match any human genome sequence was used as a control siRNA

(5’-aacacagtggagcgaattcct-3’; Qiagen). Jurkat cells and normal T cells were transfected with siRNA (2 μg) via electroporation using the Nucelofector Amaxa device and

cell-specific nucleofector reagent (Gaithersburg, MD) according to the manufacturer’s

recommendations.

Statistical analysis. In order to estimate the changes in SOCS1-3, CIS, and P-STAT1

over time, linear mixed effects models were applied to the data (Figures 5.1, 2). For each

model, the presence of a significant quadratic trend was assessed, as the changes over

94 time appeared to be nonlinear. Based on the models, point estimates of fold increases

were calculated with 95% confidence intervals. Statistical analyses for Figures 5.3-7

were assessed by analysis of variance (ANOVA); pairwise comparisons were performed

using a two-sided α = 0.05 level of significance. Kaplan-Meier estimates of the survival

function were calculated for the different mouse groups, and log-rank tests were used to

assess group differences in survival. For all analyses, an α = 0.05 level of significance

was used. All analyses were performed using SAS v9.1 (SAS Institute, Cary, NC).

5.3 Results

SOCS1, 2, 3, and CIS are rapidly induced in human PBMCs following in vitro

stimulation with IFN-α. Freshly isolated PBMCs from normal human donors (n = 3)

were treated in vitro with varying concentrations of IFN-α-2b (101-105 U/ml) and tested

for the induction of SOCS transcripts by Real Time PCR (Figure 5.1A-D represent the

combined results of these normal donors). SOCS1-3 and CIS were rapidly induced

(1 hour or less) in normal PBMCs following in vitro stimulation with IFN-α

(Figure 5.1A-D and data not shown). Significant induction of these genes as compared to

PBS-treated cells was observed following treatment with doses of IFN-α as low as

103 U/ml (p < 0.0001), however, maximal expression occurred in response to 104 or

105 U/ml of IFN-α (p < 0.0001). For example, PBMCs treated for 2 hours with

104 U/mL IFN-α expressed high levels of SOCS1 (10.32 ± 2.19 fold induction), SOCS2

(13.10 ± 3.46), SOCS3 (2.79 ± 0.87), and CIS (5.98 ± 2.97) as compared to PBS-treated

95 cells. SOCS transcripts were also reproducibly induced at lower doses of IFN-α

(101-102 U/mL; data not shown) but to a lesser extent. SOCS4-7 and protein inhibitor of activated STAT1 (PIAS1) transcripts were not induced even when high doses of IFN-α were employed (data not shown). IFN-α-induced expression of SOCS transcripts by

PBMCs was transient in nature as levels of SOCS1-3 and CIS reverted to baseline within

2-4 hours of stimulation. PBMCs from these same donors were simultaneously tested for activation of STAT1 using a flow cytometric assay that employs an Ab specific for the phosphorylated form of STAT1 (P-STAT1). Robust phosphorylation of STAT1 was observed within 30 minutes of IFN-α treatment (p < 0.001 vs. PBS-treated PBMCs), however, levels of P-STAT1 had returned to baseline within 2 hours of stimulation

(Figure 5.1E), a finding that was consistent with the expression of multiple SOCS species at this time point. The enhanced expression of SOCS species was confirmed at the protein level by immunoblot analysis. These studies revealed that SOCS1, SOCS2,

SOCS3 and CIS were rapidly induced in PBMCs following exposure to IFN-α

(Figure 5.1F). In each case, there was increased expression of SOCS protein at the

1 hour time point. Maximal induction of SOCS3 and CIS occurred 3 hours post-treatment, whereas SOCS1 and SOCS2 expression peaked at 5 hours and 1 hour, respectively.

SOCS1, 2, 3, and CIS are differentially induced in human T cells and NK cells following in vitro stimulation with IFN-α. To characterize the induction of negative regulators in the immune cell compartments thought to be responsible for mediating the anti-tumor effects of IFN-α (163-165), NK cells and T cells were isolated from fresh PBMCs 96 (n = 3 donors), treated in vitro with 104 U/ml IFN-α for varying periods of time (1, 2, 4,

6 hours), and analyzed for SOCS transcripts by Real Time PCR. SOCS1-3 and CIS

transcripts were significantly upregulated in T cells following incubation with IFN-α as

compared to PBS-stimulated cells (p < 0.001; Figure 5.2A-D). In contrast, only SOCS1

(p = 0.0263), SOCS2 (p = 0.0688), and SOCS3 (p = 0.0031) were upregulated over

baseline in NK cells. Significantly greater induction of SOCS1 (p = 0.0106), SOCS2

(p < 0.0001), and CIS (p < 0.0001) were observed in the T cell compartment as compared

to the NK compartment at the 2 hour time point. SOCS3 transcripts were upregulated to

a similar degree within T cells and NK cells in response to IFN-α (p = 0.1521). T cells

and NK cells were simultaneously evaluated for levels of phosphorylated STAT1 by flow

cytometry (Figure 5.2E). This analysis revealed a significantly greater induction of

P-STAT1 in T cells as compared to NK cells at each time point and a more rapid return to

baseline in T cells beginning at 2 hours (p < 0.0001).

Expression of SOCS transcripts in patient PBMCs post-IFN therapy. To characterize the

expression of SOCS transcripts in response to exogenous IFN-α, peripheral venous blood

was obtained from patients with metastatic melanoma (n = 5) immediately prior to and

1 hour following the first dose of IFN-α-2b (20 MU/m2 i.v.). PCR analysis revealed that

SOCS1, SOCS2, SOCS3 and CIS transcripts were all expressed at the one hour time

point following IFN-α therapy (Figure 5.3A). Of note, there was considerable

inter-patient variation with respect to the induction of SOCS1 (range = 8.6-171.0 fold increase compared to baseline levels), SOCS2 (range = 3.0-9.6 fold increase), SOCS3

(range = 1.1-4.6 fold increase), and CIS (range = 1.2-11.5 fold increase). In addition, 97 freshly isolated PBMCs from normal human donors (n = 5) were treated in vitro with

IFN-α-2b (104 U/ml) and tested for the induction of SOCS transcripts by Real Time PCR

(Figure 5.3B). The induction of SOCS1 in normal donors following in vitro IFN-α

stimulation was also variable. Interestingly, there appeared to be greater induction of

SOCS species in PBMCs following in vivo administration of IFN-α.

Over-expression of SOCS proteins in the Jurkat cell line inhibits the response to IFN-α.

The Jurkat T cell lymphoma cell line was transduced with retroviral constructs expressing

SOCS1, 2, or 3 to further evaluate the role of SOCS in IFN-α-mediated signal transduction and gene expression. As a control, Jurkat cells were also transduced in parallel with the unmanipulated vector (PINCO-EGFP). Cells were then harvested from culture, enriched for EGFP expression by FACS sorting (> 95% pure) and tested for their responsiveness to IFN-α. Increased expression of these SOCS species was confirmed by

Real Time PCR and immunoblot analysis (Figure 5.4A-B). Cells were treated with various concentrations of IFN-α (102-104 U/ml) or PBS for 30 minutes and then analyzed

for the level of P-STAT1 by flow cytometry and downstream ISG transcripts by Real

Time PCR (18 hour time point). Phosphorylation of STAT1 was not appreciably affected

by transduction of Jurkat cells with the empty PINCO-EGFP vector, however, cells that

over-expressed SOCS1 or SOCS3 showed significantly reduced levels of P-STAT1 in

response to IFN-α treatment (p < 0.0001; Figure 5.4C). The induction of IFN-stimulated

genes was also significantly inhibited in cells that over-expressed SOCS1 or SOCS3

(p < 0.045 for all genes; Figure 5.4D). For example, the induction of

98 2’-5’-oligoadenylate synthetase 1 (OAS1) was 2-fold less in SOCS1 over-expressing

Jurkat cells and 4-fold less in SOCS3 over-expressing Jurkat cells as compared to

PINCO-transduced control cell lines (p = 0.0002 and p < 0.0001, respectively). In contrast, over-expression of SOCS2 did not have a significant effect on the activation of

STAT1 or transcription of downstream ISGs. In order to demonstrate that the

PINCO-SOCS2 construct was producing a functional SOCS2 protein, we evaluated the known effects of SOCS2 on growth hormone (GH) induced STAT5 phosphorylation

(166). The 1106 MEL melanoma cell line was used instead of the Jurkat cell line due to

the fact that Jurkat cells have very low levels of STAT5 protein (167). As shown in

Figure 5.4E, over-expression of SOCS2 resulted in a significant decrease in GH-induced

STAT5 phosphorylation. These data confirmed that the SOCS2 vector produces a

functional protein that is capable of inhibiting relevant signaling pathways. Further in

vitro studies were also conducted to analyze the relationship between IFN-α-induced

SOCS transcripts and the transcription of IFN-α responsive genes. Since SOCS

transcripts are rapidly induced, human PBMCs were analyzed for the expression of

SOCS transcripts by Real Time PCR two hours following IFN-α stimulation (104 U/mL).

Expression of G1P2 mRNA was evaluated in these same normal donors at the 4 hour time point, which is when this transcript is maximally induced by IFN-α. As expected, there was a strong inverse correlation between the expression of transcript for SOCS1 and SOCS3 and the induction of G1P2 mRNA (parametric Pearson correlation; p = 0.0200 and p = 0.0218, respectively; data not shown). SOCS2 and CIS mRNA

99 expression did not correlate with G1P2 expression. These results indicate that the

transcriptional response of lymphoid cell lines to IFN-α is negatively regulated by

specific SOCS proteins.

Splenocytes from SOCS-deficient mice exhibit increased responsiveness to IFN-α. The

genotype of SOCS1- and SOCS3-deficient mice was confirmed by PCR and altered

expression of SOCS protein was confirmed by immunoblot analysis (data not shown).

Splenocytes from SOCS1+/+ IFN-γ-/-, SOCS1+/- IFN-γ-/-, and SOCS1-/- IFN-γ-/- mice were

examined for the ability to respond to IFN-α with activation of STAT1 and the

transcription of IFN-α-regulated genes. As seen in Figure 5.5A, the generation of

P-STAT1 in response to IFN-α was markedly increased in SOCS1-/- splenocytes as

compared to SOCS1+/- splenocytes (p = 0.002) and SOCS1+/+ splenocytes (p = 0.004),

while SOCS1+/- splenocytes did not show enhanced activation of STAT1 (p = 0.154 versus SOCS1+/+ splenocytes). The transcription of downstream IFN-stimulated genes

(Gzmb, Ifit2, Mx2) was also significantly enhanced in SOCS1-/- splenocytes as compared

+/- +/+ to splenocytes from SOCS1 and SOCS1 mice (p < 0.05 for all genes; Figure 5.5C).

Of note, splenocytes from SOCS1-deficient and SOCS1-competent mice expressed

identical baseline levels of STAT1 and IFN-α receptor (data not shown). Similar studies were conducted with SOCS3+/+ and SOCS3+/- splenocytes and revealed increased

activation of STAT1 (p = 0.0002, Figure 5.5B) and enhanced transcription of ISGs in

response to IFN-α as compared to SOCS3+/+ splenocytes (p < 0.0001 for all genes;

Figure 5.5D).

100 siRNA-mediated down-regulation of SOCS1 and SOCS3 in Jurkat cells and normal

T cells enhances the response to IFN-α. The Jurkat cell line was transfected with siRNA oligonucleotides targeting SOCS1 or SOCS3 and placed in culture for 16 hours prior to analysis. Control conditions included cells transfected with a control siRNA sequence as well as mock-transfected cells. The siRNA constructs inhibited the expression of SOCS1 and SOCS3 in Jurkat cells at the transcript and protein level (Figure 5.6A and B). Cells were then stimulated with IFN-α (104 U/ml) and analyzed for levels of the interferon-stimulated genes G1P2, OAS1, and IFIT2 at the 18 hour time point. As seen in Figure 5.6C, knockdown of SOCS1 and SOCS3 led to a significant increase in the

transcription of the indicated genes (p < 0.05 for all genes) as compared to the control

conditions. Similar results were obtained following transfection of primary T cells from

normal human donors with siRNA targeting SOCS1 or SOCS3 (data not shown).

SOCS-deficiency augments the anti-tumor effects of IFN-α in vivo. The anti-tumor

effects of IFN-α in the setting of SOCS1-deficiency were studied in a murine model of

malignant melanoma in which JB/MS cells (1 x 106) were injected intraperitoneally (i.p.)

into SOCS1+/- IFN-γ-/- or SOCS1-/- IFN-γ-/- mice (n = 7 mice/group) (38). One day after tumor challenge, mice received daily i.p. injections of IFN-A/D (2 x 104 U) or PBS.

SOCS1+/+ IFN-γ-/- mice and wild-type mice (SOCS1+/+ IFN-γ+/+) served as controls. As

expected, treatment of both tumor-bearing IFN-γ-/- and wild-type mice with IFN-A/D led to a significant improvement in survival as compared to treatment with PBS alone

(Figure 5.7A; p = 0.0027). The anti-tumor effects of IFN-A/D therapy were significantly

enhanced in SOCS1-deficient mice as 57 - 71% of these mice were cured of their tumors, 101 whereas PBS-treated mice all died at 10-14 days (Figure 5.7B, p = 0.0002 for SOCS1+/- mice; Figure 5.7C, p = 0.0002 for SOCS1-/- mice). Of note, there was no evidence that

mice with SOCS deficiencies experienced increased systemic toxicity in response to

IFN-α treatments: These mice behaved normally and histochemical analysis of their visceral organs revealed no increase in tissue damage or inflammatory cell infiltrate (data not shown). Immunohistochemistry of treated tumors revealed increased infiltration of

CD45+ immune cells with IFN-α treatment but there was no difference between

SOCS-competent and SOCS-deficient mice. SOCS1-deficient mice that survived

following IFN-A/D treatment were observed for a total of 90 days during which no

recurrent tumors were identified. These mice were then re-challenged with

1 x 106 JB/MS cells and remained tumor free for 100 days in the absence of additional

IFN-α treatments (identically treated tumor-naïve WT mice rapidly developed lethal disease), suggesting that an adaptive immune response to the tumor had been induced.

Similarly, SOCS3-deficient mice exhibited enhanced survival (median

survival = 29 days) in response to IFN-A/D therapy compared to wild-type littermates

(median survival = 21 days; Figure 5.7D; p = 0.0091). However, all SOCS3-deficient

mice treated with IFN-A/D eventually succumbed to their tumors.

In vivo depletion of CD8+ T cells significantly inhibits the anti-tumor action of

IFN-α of SOCS1-deficient mice. We next examined the role of CD8+ T cells in

mediating the anti-tumor effects of IFN-α in this murine model. SOCS1-competent and

SOCS1-deficient mice were depleted of CD8+ T cells via i.p. injection of an anti-CD8

antibody. Control mice were treated with the appropriate isotype control antibody. Mice 102 were then challenged i.p. with tumor and received daily i.p. injections of IFN-A/D (or

PBS) beginning the next day. Depletion of CD8+ T cells markedly inhibited the

anti-tumor effects of IFN-A/D in SOCS1-deficient mice (Figure 5.7E; p = 0.0002). In

contrast, mice receiving the control antibody exhibited prolonged survival following

therapy with IFN-A/D (p = 0.0002) as observed previously. Depletion of CD8+ T cells also inhibited the effectiveness of IFN-A/D in IFNγ-/- mice (Figure 5.7F; p = 0.0062).

These data suggest that CD8+ T cells play a critical role in mediating the anti-tumor

effects of IFN-α and that this effect is enhanced in SOCS1-deficient mice.

Role of SOCS1 in immunosurveillance of developing tumor. We hypothesized that

SOCS1-deficient mice could mediate the immunosurveillance of small developing

tumors by the regulation of endogenous cytokine responses or other factors. To test this

hypothesis, we first needed to determine the lowest amount of JB/MS melanoma cells

that was necessary to form a competent tumor. As such, JB/MS melanoma cell lines

were injected into WT mice (C57BL/6) with a range of subcutaneous inoculums

(1*100 -1*106 cells). Mice were then observed for tumor formation and we found that

1*104 cells injected subcutaneously was the lowest amount of cells that would

consistently form growing tumors. Thus, IFN-γ-/- and SOCS1-/- IFN-γ-/- mice were

injected with 1*104 JB/MS cells and observed overtime for tumor formation and growth.

Impressively, only 2/7 SOCS1-deficient mice formed tumor while all IFN-γ-/- mice grew

tumor (Figure 5.8).

103 5.4 Discussion

We have previously demonstrated that Jak-STAT signal transduction within host immune cells is critical to the anti-tumor effects of IFN-α in a murine model of malignant melanoma (18). This led us to examine the role of SOCS proteins in regulating the anti-tumor properties of IFN-α-stimulated immune effector cells. The present study demonstrated that SOCS1, SOCS2, SOCS3, and CIS were rapidly induced in a dose-dependent manner by IFN-α in normal lymphocytes at the transcript and protein level. The inhibitory effects of SOCS proteins were highly specific, as over-expression of SOCS1 and SOCS3, but not SOCS2, in the Jurkat T cell lymphoma cell line inhibited

IFN-α-induced activation of STAT1 and the expression of interferon-stimulated genes.

Conversely, siRNA-mediated down-regulation of SOCS1 and SOCS3 in Jurkat cells and

T cells enhanced the transcriptional response to IFN-α. The functional importance of

SOCS proteins in regulating the actions of IFN-α was suggested by the improved ability of SOCS1- and SOCS3-deficient mice to eliminate melanoma tumor cells following treatment with IFN-α.

To date, there have been limited studies investigating the regulation of SOCS family protein expression in IFN-α-stimulated immune effector cells. IFN-α has been shown to induce the expression of SOCS species in megakaryocytes, LPS-stimulated macrophages, in an IL-2 primed human T cell line, and a human chronic myelogenous leukemia cell line (110, 112, 168-171). Over-expression of SOCS proteins has been

104 shown to inhibit cytokine responses in hepatoma cells and activated T cells (111, 172-

176). However the effects of SOCS proteins on unmanipulated human lymphocytes have not been fully explored. We demonstrated that SOCS1, SOCS2, SOCS3, and CIS are differentially induced in both T cells and NK cells. The finding of variable SOCS induction in malignant melanoma patient PBMCs after the administration of high-dose

IFN-α-2b supports the clinical relevance of these findings. This further suggests that the immune response to exogenous cytokines may be influenced by individual differences in the expression of SOCS proteins within lymphocyte populations.

In the present report, inhibition of SOCS expression in T cells had marked effects on IFN-α-induced gene expression. Furthermore, the anti-tumor effects of IFN-α were significantly enhanced in SOCS1-deficient mice: approximately 70% of mice achieved long-term resolution of their tumor-burden while no adverse side-effects were observed.

SOCS1-deficient mice that were tumor free following IFN-α treatment survived a second challenge with JB/MS melanoma cells. SOCS3 deficiency also promoted the anti-tumor effects of IFN-α. These enhanced anti-tumor effects of IFN-α were mediated, at least in part, by CD8+ T cells. Several lines of evidence point to SOCS1 as an important regulator of T cell development and function. Mice with a targeted deficiency of SOCS1 within the T cell compartment exhibit an increased ratio of CD8/CD4 mature thymic cells and a significant increase in the prevalence of CD44hiCD8 memory T cells within the periphery (177). Memory T cells from these mice displayed a 5-fold increased proliferative response to IL-2 and IL-15 as compared to wild type mice. Similarly, the proliferation of T cells from SOCS1-/- IFN-γ-/- mice in response to anti-CD3 Ab was 105 markedly enhanced in the presence of IL-12, whereas this cytokine had minimal effects

on activated T cells from normal mice (177, 178). Recently, Davey et al. discovered that

SOCS1-deficient CD8+ T cells proliferate when transplanted to normal mice and that this

response is driven by IL-15 and self ligands that normally drive homeostatic proliferation in T cell-deficient mice (177). The ability of CD8 depletion to abrogate the pro-survival effects of IFN-α in tumor-bearing, SOCS1-deficient mice suggests that this immune compartment can exert potent anti-tumor actions in response to IFN-α in the absence of

SOCS1 activity.

Importantly, dendritic cell (DC) function is also regulated by SOCS1. SOCS1-/- mice have greater numbers of DCs. Also, these DCs respond more strongly to IL-4 and

IFN-γ and produce higher levels of B cell maturation/differentiation factors such as

Baff/BLys (a member of the TNF ligand superfamily) as compared to normal mice (179).

Shen et al. subsequently demonstrated that down-regulation of SOCS1 in dendritic cells

(DCs) using lentiviral-delivered siRNA led to enhanced production of TNF-α, IL-6, and

IL-2 in response to LPS and IFN-γ treatments. In addition, co-culture of

SOCS1-deficient DCs with antigen-specific CD8+ T cells led to enhanced T cell

proliferation and cytokine secretion as compared to wild type DCs. Immunization with

tryosinase-related protein-2 (TRP2)-pulsed, SOCS1-silenced DCs abrogated the growth

of a TRP2+ B16 melanoma in C57BL/6 mice (180). A similar study confirmed these

results and identified several genes that were expressed to a greater degree in

SOCS1-deficient DCs as compared to SOCS1+/+ DCs (181). Thus, SOCS1-deficiency

106 appears to promote both the processing of tumor antigens by DCs and their recognition

by effector T cells. Analysis of IFN-α-induced DC activity in SOCS-deficient mice is therefore warranted.

Reduced levels of SOCS1 in target cells appear to promote the therapeutic efficacy of IFN-α immunotherapy. Evidence in favor of this concept comes from a manuscript by Roman-Gomez et al. demonstrating that constitutive expression of SOCS1 in PBMCs obtained from chronic myeloid leukemia patients correlated with a significantly shorter progression free survival and poor cytogenetic response to IFN-α therapy (182). Similarly, it has been shown by Imanaka et al. that high level expression of SOCS1 within hepatocytes from chronic hepatitis C patients correlated with a diminished anti-viral response to IFN-α (176). Also, Fenner et al. recently demonstrated that, compared to WT mice, SOCS1-deficient mice are better able to utilize endogenous

IFN-α to clear an otherwise lethal viral infection (183). Although we have not yet linked the phenomenon to endogenous IFN-α, we also show that SOCS1-deficient mice inhibited the formation of tumor. These data are in agreement with our findings and point to a role for SOCS1 in regulating the anti-tumor effects of IFN-α therapy and the outcome of viral infections. Interestingly, selective down-regulation of SOCS1 within the T cell compartment does not lead to an overt inflammatory response, as occurs in

SOCS1-deficient mice that possess a functioning IFN-γ gene (184). Therefore, future targeted therapies that down-regulate SOCS proteins within specific cell types may help to avoid the toxicity associated with global down-regulation of SOCS1.

107 The present study demonstrated that over-expression of SOCS1 and SOCS3 had negative effects on IFN-α-induced Jak-STAT signal transduction and gene regulation, while experimental reductions in SOCS activity enhanced the response to IFN-α.

Reduced expression of SOCS1 or SOCS3 in tumor-bearing mice enhanced the anti-tumor activity of exogenous IFN-α. These findings suggest that modulation of SOCS activity may have beneficial effects in the setting of cytokine-mediated immune phenomena.

SOCS proteins may also play a role in mediating endogenous interferon activity which is known to be important in the immune surveillance for cancer cells (93).

108 5.5 Tables and Figures

Figure 5.1. SOCS transcripts are rapidly induced in PBMCs following IFN-α

stimulation. PBMCs from normal donors were treated with IFN-α (103-105 U/mL) or phosphate-buffered saline (PBS) and SOCS mRNA levels were measured by Real Time

PCR at 4 time points (0.5, 1, 2, 4 hours) using primers specific for (A) SOCS1,

(B) SOCS2, (C) SOCS3, or (D) CIS. Data were expressed as the mean fold increase relative to baseline levels (PBS treatment). All Real Time PCR data were normalized to the level of β-actin mRNA (housekeeping gene). (E) Phosphorylation of STAT1 at

Tyr701 was measured in parallel by flow cytometry. Mean specific fluorescence (Fsp) is illustrated on the y-axis. Appropriate isotype control antibodies were used to determine background staining. All flow cytometric data were derived from at least 10,000 events gated on the lymphocyte populations determined by light scatter properties (forward scatter vs. side scatter). Error bars denote the 95% confidence interval of triplicate experiments (3 donors). Statistically significant results vs. PBS treatment are denoted by an asterisk. PBMCs were isolated from normal donors, stimulated with IFN-α

(104 U/mL) and harvested at various times (1, 3, 5, 7, 24 hours). Following immunoprecipitation, SOCS protein levels were measured by immunoprecipitation and

subsequent immunoblot analysis using antibodies directed against (F) SOCS1, SOCS2,

SOCS3, or CIS. Lysates from a SOCS-over-expressing human melanoma cell line

(HT144) were used as positive controls. Due to the immunoprecipitation protocol levels

of β-actin were measured separately to control for loading. Densitometric data (fold 109 induction) for each condition is shown at the bottom of each lane. Blots shown are representative of separate experiments utilizing PBMCs from three normal donors.

110 Figure 5.1 A 14 SOCS1

0.5 hr * 12 1 hr 2 hr 4 hr *

10 *

6 Fold increase vs. PBS Fold increasevs.

0 103 104 105 Interferon-alpha (U/mL)

25 B SOCS2

0.5 hr * 1 hr 2 hr 20 4 hr

10 * Fold increase PBS vs.

0 103 104 105 Interferon-alpha (U/mL) Continued… 111

C 6 SOCS3 Figure 5.1 Continued

0.5 hr 1 hr 5 2 hr 4 hr * * 4 *

Fold Increase vs. PBS vs. Fold Increase 2

0 103 104 105 Interferon-alpha (U/mL)

D 12 CIS * 0.5 hr 1 hr 10 2 hr 4 hr *

6 *

Fold Increase vs. PBS vs. Fold Increase 4

0 103 104 105 Interferon-alpha112 (U/mL) Continued…

E Figure 5.1 Continued 40

0.5 hr 1 hr 2 hr 30 4 hr *

20 * Fsp (P-STAT1)

0 PBSPBS10 10^33 10 10^44 10 10^55

Interferon-alpha (U/mL)

113

F Figure 5.1 Continued + Ctrl PBS 1h 3h 5h 7h 24h

SOCS1

β-actin

1.0 1.5 1.5 2.3 1.7 1.0 + Ctrl PBS 1h 3h 5h 7h 24h

SOCS2

β-actin

1.0 1.7 0.6 0.6 0.4 0.2 24h + Ctrl 1h 3h 5h 7h PBS

SOCS3

β-actin 1.0 4.8 5.1 2.1 2.7 0.4 + Ctrl 7h 24h PBS 1h 3h 5h

CIS

β-actin

1.0 1.8 2.1 2.1 1.5 0.8

114 Figure 5.2. SOCS transcripts are differentially induced in NK cells and T cells following

IFN-α stimulation. NK cells (CD56+) and T cells (CD3+) were isolated from normal

donors (n = 3) and treated with 104 U/mL IFN-α or PBS. Cells were harvested at 4 time

points (1, 2, 4, 6 hours) for Real Time PCR analyses of (A) SOCS1, (B) SOCS2,

(C) SOCS3, and (D) CIS transcript levels. Data were expressed as the mean fold increase relative to baseline levels (PBS treatment). All real time PCR data were normalized to the level of β-actin mRNA. (E) Flow cytometric analysis of P-STAT1 levels in purified NK and T cell subsets was performed in parallel. Error bars denote the

95% confidence interval of triplicate experiments (3 donors). Statistically significant results vs. PBS treatment are denoted by an asterisk. Significant differences between

T cells and NK cells are denoted by ‡.

115

A Figure 5.2 60 SOCS1

NK cells T cells 50

‡

20 * Fold increase vs. PBS *

0 1246

Time (hours)

B 35 SOCS2 ‡ NK cells * 30 T cells

20 * 15

Fold increase vs. PBS vs. increase Fold 10

0 1246 Time (hours) Continued… 116 C Figure 5.2 Continued 10 SOCS3

NK cells T cells 8 *

PBS * 6

ncrease vs. 4 i

ld o F

0 1246

Time (hours)

D 30 CIS NK cells ‡ T cells 25 *

10 Fold increase vs. PBS

0 1246 Continued… Time (hours) 117 E Figure 5.2 Continued

80 ‡ NK cells * 70 T cells

Fsp (P-STAT1) * 20

0 PBS 0.5 1 2 4 6

Time (hours)

118 Figure 5.3. Differential SOCS expression and STAT1 activation in melanoma patients

undergoing IFN-α immunotherapy. PBMCs from (A) five melanoma patients were obtained immediately prior to and 1 hour following administration of high-dose IFN-α

(20 MU/m2 i.v.) and analyzed for SOCS1-3 and CIS mRNA levels by Real Time PCR. In

addition, freshly isolated PBMCs from (B) normal human donors (n = 5) were treated

in vitro with IFN-α-2b (104 U/ml) and tested for the induction of SOCS transcripts by

Real Time PCR. Data were expressed as the mean fold increase relative to baseline

levels. All real time PCR data were normalized to the level of β-actin mRNA.

119

A Figure 5.3

200 Pt. 1 Pt. 2 150 Pt. 3 Pt. 4 Pt. 5 100

10 Fold increase vs. Pretreatment vs. Fold increase 5

0 SOCS1 SOCS2 SOCS3 CIS

B 8 Donor 1 Donor 2 Donor 3 Donor 4 Donor 5 6

Fold increase vs. PBS Fold increasevs. 2

0 SOCS1 SOCS2 SOCS3 CIS

120 Figure 5.4. Over-expression of SOCS1 and SOCS3 protein in Jurkat cells inhibits the

response to IFN-α. The Jurkat T cell lymphoma cell line was transduced with PINCO

retroviral constructs encoding SOCS1, SOCS2, or SOCS3 protein. Cells transduced with

the empty PINCO vector served as a negative control. Cell populations were routinely

greater than 95% pure post fluorescence-activated cell sorting for fluorescent green

protein. (A) Transcript levels of SOCS1, SOCS2, and SOCS3 in transduced cells as

measured by Real Time PCR. Data are expressed as the mean fold increase relative to

baseline levels. All real time PCR data were normalized to the level of β-actin mRNA in

duplicate experiments. (B) Protein levels of SOCS1, SOCS2, and SOCS3 as measured

by immunoblot analysis of whole cell lysates. These results are representative of

duplicate experiments. (C) Flow cytometric analysis of P-STAT1 formation in

transduced cells following stimulation with IFN-α (102-104 U/mL) or PBS for

30 minutes. (D) Analysis of 2’-5’-oligoadenylate synthetase 1 (OAS1), interferon-stimulated gene 20 (ISG20), and interferon-induced protein with tetratricopeptide repeats 2 (IFIT2) transcript levels in transduced cells following stimulation with IFN-α (104 U/mL) or PBS for 18 hours. Error bars denote the 95%

confidence interval of duplicate experiments. Statistically significant results are denoted

by an asterisk. (E) The 1106 MEL melanoma cell line was transduced with a PINCO retroviral construct encoding the SOCS2 protein. Cells transduced with the empty

PINCO vector served as a negative control. Cells were treated with PBS or human growth hormone (huGH; 100 ng/mL) for 15 minutes. Levels of P-STAT5 were measured by immunoblot analysis. Beta-actin was used as a loading control. SOCS2

121 over-expression was also confirmed by immunoprecipitation and subsequent immunoblot analysis. Non-specific bands (NSB) represent the light chain of the Rabbit anti-SOCS2 antibody used in the immunoprecipitation.

122

A Figure 5.4

Fold Increase vs. PINCO Emtpy Transfection PINCO Emtpy Increase vs. Fold 0 SOCS1 SOCS2 SOCS3 PINCO SOCS Vector

Empty SOCS1 Empty SOCS2 Empty SOCS3

SOCS1 SOCS2 SOCS3

β-actin β-actin β-actin 1.0 2.3 1.0 3.2 1.0 2.1

Continued…

123

C Figure 5.4 Continued

60 PINCO-Empty PINCO-SOCS1 50 PINCO-SOCS2 PINCO-SOCS3

* 30 *

Fsp (P-STAT1) 20 * * * * 10

0 1010^23 10 10^34 10 10^45 Interferon-alpha (U/mL)

124 Continued…

D Figure 5.4 Continued

600 OAS1

500

400

300 *

200 Fold increase vs. PBS 100 *

0 Empty SOCS1 SOCS2 SOCS3

20 ISG20 18

14 12 * 10 * 8

Fold increase vs. PBS increase Fold 4

0 Empty SOCS1 SOCS2 SOCS3

600 IFIT1

500

400 * * 300

200 Fold increase vs. PBS 100

0 Empty SOCS1 SOCS2 SOCS3

125 Continued…

E Figure 5.4 Continued

Pinco Pinco Empty SOCS2 PBS huGH PBS huGH

P-STAT5

β-actin

NSB

SOCS2

126 Figure 5.5. SOCS1- and SOCS3-deficient mice exhibit an augmented response to

IFN-α. (A, B) The formation of P-STAT1 in SOCS-competent and SOCS-deficient murine splenocytes was measured by flow cytometry following stimulation of with

IFN-A/D (104 U/mL) or PBS for 30 minutes. (C, D) Transcription of granzyme B

(Gzmb), Ifit2, and myxovirus resistance 2 (Mx2) genes were measured in splenocytes from SOCS1- and SOCS3-deficient mice following an 18 hour stimulation with IFN-A/D

(104 U/mL) or PBS via Real Time PCR. Data were expressed as the mean fold increase relative to baseline levels (PBS treatment). All real time PCR data were normalized to the level of β-actin mRNA. Error bars denote the 95% confidence interval of duplicate experiments. Statistically significant results are denoted by an asterisk.

127

A Figure 5.5

SOCS1 +/+ * SOCS1 +/- SOCS1 -/- 10

6 Fsp (P-STAT1) 4

0 PBSPBS 10^4104

Interferon-alpha (U/mL)

128 Continued…

B Figure 5.5 Continued

14 SOCS3 +/+ SOCS3 +/- * 12

6 Fsp (P-STAT1) Fsp

0 PBSPBS 10^4104 Interferon alpha (U/mL)

129 Continued…

C Figure 5.5 Continued

180 Gzmb 160 *

140

120

100

Fold increase vs. PBS increase Fold 40

0 SOCS1+/+ SOCS1+/- SOCS1-/-

60 Ifit2 *

20 Fold increase vs. PBS 10

0 SOCS1+/+ SOCS1+/- SOCS1-/-

800 Mx2 *

600

400

200 Fold increase vs. PBS

0 SOCS1+/+ SOCS1+/- SOCS1-/-

130 Continued…

D Figure 5.5 Continued 140 Gzmb * 120

100

40 Fold increase vs. PBS vs. Fold increase

0 SOCS3+/+ SOCS3+/-

80 Ifit2 *

20 Fold increase vs. PBS vs. Fold increase

0 SOCS3+/+ SOCS3+/-

80 Mx2 *

20 Fold increase vs. PBS vs. Fold increase

0 SOCS3+/+ SOCS3+/- 131 Figure 5.6. Small inhibitory RNA (siRNA) mediated inhibition of SOCS1 and SOCS3

augments IFN-α-responsiveness in vitro. Jurkat cells were transfected via

electroporation with siRNA specific to SOCS1 and SOCS3. As a negative control, Jurkat

cells were transfected in parallel with a control siRNA or no siRNA (mock transfection).

(A) Specific down-regulation of SOCS by siRNA was confirmed at the transcript level by

Real Time PCR and (B) at the protein level by immunoprecipitation and immunoblot

analysis of SOCS1 and SOCS3. These results are representative of duplicate experiments. (C) IFN-α-responsiveness at the level of gene transcription was measured in siRNA transfected cell lines via Real Time PCR. Fold increase in ISGs were determined following an 18 hour stimulation with IFN-α (104 U/mL) or PBS (interferon,

alpha-inducible protein (clone IFI-15K) [G1P2], 2’-5’-oligoadenylate synthetase 1

[OAS1], and interferon response factor 7 [IRF7]). Data were expressed as the mean fold

increase relative to baseline levels (PBS treatment). All Real Time PCR data were

normalized to the level of β-actin mRNA. Error bars denote the 95% confidence interval

of duplicate experiments. Statistically significant results are denoted by an asterisk.

132

A Figure 5.6

7 Mock Control siRNA 6 siSOCS1 siSOCS3

2 Fold increase vs. PBS Fold increase vs.

0 SOCS1 SOCS3 SOCS gene

B Mock siControl siSOCS1 siSOCS3

SOCS1

SOCS3

β-actin

133 Continued… Figure 5.6 Continued C G1P2

1000 * 800

600 *

400 Fold increase vs. PBS Fold increase

200

0 Mock siControl siSOCS1 siSOCS3

OAS1

700 * 600 * 500

400

300

Fold increase vs. PBS Fold increase 200

100

0 Mock siControl siSOCS1 siSOCS3

IRF7

6 *

5 *

2 Fold increase vs. PBS Fold increase

0 Mock siControl134 siSOCS1 siSOCS3 Figure 5.7. SOCS-deficiency enhances the anti-tumor effect of IFN-A/D in a murine model of malignant melanoma. (A,) SOCS1+/+ IFN-γ-/-, (B) SOCS1+/- IFN-γ-/-,

on day 0. Wild-type mice served as controls. Beginning on day 1, mice received daily

i.p. injections of PBS or 2 x 104 U of IFN-A/D. As expected, treatment of tumor-bearing

SOCS1+/+ mice with IFN-α led to a significant improvement in survival as compared to

treatment with PBS alone (p = 0.0002; Figure A). The anti-tumor effects of IFN-α

therapy were significantly enhanced in SOCS1-deficient mice as 57 - 71% of these mice

were cured of their tumors (Figure B, C), whereas PBS-treated mice all died at

10-14 days (p = 0.0002 for SOCS1+/- mice; p = 0.0002 for SOCS1-/- mice). Each group contained 7 mice. The anti-tumor effects of IFN-α therapy were significantly enhanced in SOCS3-deficient mice versus SOCS3+/+ mice (p = 0.0091; Figure D). Each group

contained 4 mice.. Prior to tumor inoculation, (E) SOCS1-/- IFN-γ-/- and (F) SOCS1+/+

IFN-γ-/- mice were injected i.p. with 100 μg of a rat anti-mouse CD8 antibody

(clone 2.43) or rat IgG on days -3, -1, +1, +3, and every fourth day thereafter.

CD8+ T cell depletion (defined as < 0.5 % CD8+ cells in the peripheral blood) was confirmed by flow cytometry in all mice prior to tumor challenge. Mice were injected with 1 x 106 JB/MS cells on day 0. Beginning on day 1, mice received daily injections of

PBS or 2 x 104 U of IFN-A/D. The anti-tumor effects of IFN-α therapy were

significantly diminished in CD8 depleted mice as compared to CD8-competent mice,

regardless of SOCS1 genotype (p > 0.0002). Each group contained 7 mice.

135

A Figure 5.7

1.0

0.8

0.6

WT (PBS) WT (IFN)

Survival % Survival 0.4 IFNg-/- (PBS) IFNg-/- (IFN)

0.2

0.0 0 5 10 15 20 25 30 Time (days)

136 Continued…

B Figure 5.7 Continued

1.0 1.0

0.8 0.8

0.6 0.6

Survival % Survival 0.4

Survival % Survival 0.4 IFNg-/- (PBS) IFNg-/- (PBS) IFNg-/- (IFN) IFNg-/- (IFN) IFNg-/-SOCS1+/- (PBS) IFNg-/-SOCS1+/- (PBS) 0.2 IFNg-/-SOCS1+/- (IFN) 0.2 IFNg-/-SOCS1+/- (IFN)

0.00.0 00 1020304050 1020304050 TimeTime (days) (days)

C 1.0 1.0

0.8 0.8

0.6 0.6

IFNg-/- (PBS)

Survival % Survival 0.4 IFNg-/- (PBS) Survival % Survival 0.4 IFNg-/- (IFN) IFNg-/- (IFN) IFNg-/-SOCS1-/- (PBS) IFNg-/-SOCS1-/- (PBS) IFNg-/-SOCS1-/- (IFN) IFNg-/-SOCS1-/- (IFN) 0.2 0.2

0.00.0 00 1020304050 1020304050 TimeTime (days) (days)

137 Continued…

D Figure 5.7 Continued

1.0

0.8

SOCS3+/+ (PBS) 0.6 SOCS3+/+ (IFN) SOCS3+/- (PBS) SOCS3+/- (IFN)

Survival % Survival 0.4

0.2

0.0 0 102030 Time (days)

138 Continued…

E Figure 5.7 Continued

1.0

0.8

0.6

IFNg KO (Isotype Ab; PBS) IFNg KO (Isotype Ab; IFN)

Survival % Survival 0.4 IFNg KO (CD8 Ab; PBS) IFNg KO (CD8 Ab; IFN)

0.2

0.0 0 10203040 Time (days)

1.0

0.8

0.6

IFNg KO SOCS1-/- (Isotype Ab; PBS)

Survival % Survival 0.4 IFNg KO SOCS1-/- (Isotype Ab; IFN) IFNg KO SOCS1-/- (CD8 Ab; PBS) IFNg KO SOCS1-/- (CD8 Ab; IFN) 0.2

0.0 0 10203040 Time139 (days) Figure 5.8. SOCS1-deficiency enhances immunosurveillance of developing melanoma tumors. Beginning on day 1, IFN-γ-/- and SOCS1-/- IFN-γ-/- mice were injected with

1*104 JB/MS cells. Mice were then observed overtime for tumor formation and growth.

Tumor sizes were estimated for each mouse and averaged for each strain of mouse. Error bars denote the standard deviation within each group.

140 Figure 5.8

2000 IFNg-/- 1800 SOCS1-/- IFNg-/-

1600

1400

1200

1000

800

600 Tumor (mm^3) volume 400

200

0 0 5 10 15 20 25 Time (days)

141

CHAPTER 6

CONCLUSION

High dose IFN-α is currently one of a limited number of options for patients with

metastatic melanoma. Additionally, this treatment is the only FDA-approved agent for

patients who have had complete tumor excision and remain at high risk for recurrence.

Therefore, this agent will continue to be offered after surgery to melanoma patients.

Unfortunately, IFN-α can be highly toxic for patients when administered at high doses.

This has reduced enthusiasm for its use in both the metastatic and adjuvant settings (59,

65). However, recent data has correlated clinical response to the occurrence of autoimmune sequelae (94, 95). This data demonstrated that a subset of patients

significantly responded to IFN-α. Furthermore, the work presented herein supports the

use of exogenous IFN-α in the setting of malignant melanoma immunotherapy and

demonstrated the following:

1.) IFN-α induced the expression of many ISGs that were dependent on STAT1

for optimal transcription.

2.) Immune cells from melanoma patients receiving IFN-α exhibited a gene

expression profile that was specific to the individual as well as the immune subset of

interest. 142 3.) Increasing doses of IFN-α did not mediate enhanced signal transduction and

gene transcription in melanoma patient immune cells.

4.) SOCS1 and SOCS3 had negative effects on IFN-α-induced Jak-STAT signal

transduction and gene transcription.

5.) Reduced expression of SOCS1 or SOCS3 in tumor-bearing mice enhanced the

anti-tumor activity of exogenous IFN-α. Additionally, SOCS1-deficiency enhanced the

immunosurveillance of developing tumors.

Despite decades of clinical use of IFN-α, the precise molecular targets are

unknown. However, we have demonstrated that the anti-tumor effects of IFN-α are

dependent on STAT1 signaling within immune cells (18). Because of this, many genes

whose expression has been shown to be regulated by STAT1 are likely to be important in

the anti-tumor response (Chapter 2). Interestingly, granzyme b (Gzmb; cytotoxic factor

secreted by NK cells and T cells), calcium/calmodulin-dependent protein kinase II, beta

(Camk2b; T cell proliferation and activation), lymphocyte antigen 6 complex, locus C

(Ly6c; T cell adhesion), and arginase 1 (Arg-1; suppression of alloreactive T cell

function) are all involved in immune effector responses that may have a role in IFN-α

immunotherapy of melanoma (185-188). Therefore, it is critical to elucidate the role of each STAT1-regulated gene.

We have also previously demonstrated a high degree of variability in the formation of P-STAT1 in patient immune effector cells following IFN-α-2b immunotherapy (116). This may be the result of abnormal basal levels of STAT1, as this 143 leads to alternative signaling through STAT3 (125). Because of this and other factors, there is a high potential for patient variability in response to IFN-α (140). Despite the inter-patient variability, the gene expression profile of in vitro IFN-α-stimulated PBMCs was indicative of the in vivo response following IFN-αimmunotherapy (Chapter 3). Once the prototypical gene expression profile for response has been determined, or the

STAT1-dependent genes important for anti-tumor activity, in vitro treatments may predict the responsiveness to IFN-α.

IFN-α is frequently administered at the highest tolerated dose due to the belief that this is the best method of inducing patient response (189). IFN-α can be given on an outpatient basis, however, the high-dose regimen (10 MU/m2 three times weekly) calls for prolonged periods of therapy where nearly 50% of patients require treatment delays or dose reductions due to constitutional, hepatic, or neurologic symptoms. The serious toxicities associated with high-dose IFN-α therapy are well documented and can be avoided if the clinician takes care to follow the precise guidelines that have recently been published (69). However, the “flu-like” constitutional symptoms caused by the administration of this cytokine are frequently cited as a reason for refusal to accept treatment (69). As a result, a number of low or intermediate dose regimens have come into use (e.g., 3 MU or 9 MU thrice weekly). These schedules, however, have not been selected on the basis of any mechanistic studies or molecular analyses (190). In contrast, our study of the dose escalation of IFN-α showed that STAT activation and ISG transcription was not optimal with 10 MU/m2 IFN-α as compared to the initial 5 MU/m2

144 dose (Chapter 4). These results suggest that lower doses of IFN-α-2b may be as effective

as higher doses in activating the Jak-STAT signal transduction in immune effector cells.

Since each patient responds in a specific manner, this data indicates that an individualized

dose-optimization is needed. Patients receiving IFN-α dose escalation or reduction could

have their immune cells monitored during therapy for IFN-α signal transduction and gene

transcription. The dose at which STAT activation and ISG transcription is the highest

would then be considered a patient’s optimal dose of IFN-α. Optimization of IFN-α

therapy would prevent under-dosing, which would theoretically improve the efficacy of

this cytokine regimen. Over-dosing, which is associated with increased toxicity, would

also be reduced, potentially leading to improved tolerability and enhanced patient

compliance.

One problem associated with dosing IFN-α is that it induces a negative feedback

loop (127). Specifically, as shown in Figure 4.4, SOCS1 and SOCS3 were induced to a

greater degree by the higher IFN-α dose. This suggests a potential mechanism for the

inhibition of signal transduction at high doses of IFN-α. Overall, patient PBMCs had

lower STAT1, STAT2, and an ISG transcript after receiving 10 MU/m2 IFN-α as compared to 5 MU/m2. Targeted therapies against these inhibitor molecules could

enhance the response to IFN-α, allowing for clinicians to utilize lower doses that are

more tolerable for patients.

145 Currently, there have been several studies that have investigated the role of

SOCS1 in cancer therapy (177, 178, 180, 181). These reports have shown that

SOCS1-deficiency within the T cell compartment allows for an increase number of

circulating CD8+ T cells that have an enhanced proliferation response to IL-2, IL-12, and

IL-15 (177, 178). Interestingly, IFN-α treatment of SOCS1-deficent mouse splenocytes

induced the enhanced transcription of the STAT1-enhanced gene, granzyme B,

potentially increasing the cytotoxicity of the T cells (Figure 5.5C). There have also been studies that have shown SOCS1-deficiency within the DC compartment leads to enhanced tumor antigen presentation and it’s recognition by effector T cells (180, 181).

Analysis of our SOCS1-deficient mice (SOCS1-deficient T cell and DC compartments) shows that the anti-tumor effects of IFN-α were significantly enhanced and mediated by

CD8+ T cells (Figure 5.7F). However, the question remains for determining the exact immune cell compartment(s) in which SOCS1-deficiency is sufficient for the enhanced anti-tumor response of IFN-α.

These findings have clear implications for future research. The studies enhanced our knowledge of the immune cell response to IFN-α immunotherapy by showing that intermediate doses may have therapeutic benefit, there is a potential to predict patient response, and SOCS inhibition may improve the efficacy of this cytokine. Melanoma remains an increasingly prevalent disease and current therapies are not satisfactory.

Conventional high-dose immunomodulatory regimens and chemotherapy have failed to improve survival of melanoma patients (191). Recent clinical trials involving IFN-α in

combination with other therapies has led to modest overall responses and no change in 146 the overall survival (192-194). It is becoming increasingly evident that no single

treatment will significantly benefit an entire patient population. In the future, the task

will be to offer a highly individualized therapy plan based on specific molecular genetic

parameters (191). Even so, the “cure” may require an attack on the disease from many fronts such as the direct anti-cancer targeted therapy (e.g. BRAF inhibitors), immune cell activation (e.g. IFN-α), enhancing agents (e.g. SOCS knockdown), and adaptive immunotherapy (autologous T cell response).

147

BIBLIOGRAPHY

1. Isaacs, A., and J. Lindenmann. 1957. Virus interference. I. The interferon. Proc R Soc Lond B Biol Sci 147:258-267.

2. Hardy, M. P., C. M. Owczarek, L. S. Jermiin, M. Ejdeback, and P. J. Hertzog. 2004. Characterization of the type I interferon locus and identification of novel genes. Genomics 84:331-345.

3. Allen, G., and M. O. Diaz. 1994. Nomenclature of the human interferon proteins. J Interferon Res 14:223-226.

4. Foster, G. R., and N. B. Finter. 1998. Are all type I human interferons equivalent? J Viral Hepat 5:143-152.

5. Uze, G., G. Lutfalla, and K. E. Mogensen. 1995. Alpha and beta interferons and their receptor and their friends and relations. J Interferon Cytokine Res 15:3-26.

6. Ortaldo, J. R., R. B. Herberman, C. Harvey, P. Osheroff, Y. C. Pan, B. Kelder, and S. Pestka. 1984. A species of human alpha interferon that lacks the ability to boost human natural killer activity. Proc Natl Acad Sci U S A 81:4926-4929.

7. Navarro, S., O. R. Colamonici, and A. Llombart-Bosch. 1996. Immunohistochemical detection of the type I interferon receptor in human fetal, adult, and neoplastic tissues. Mod Pathol 9:150-156.

8. von Stamm, U., E. B. Brocker, M. von Depka Prondzinski, D. J. Ruiter, P. Rumke, C. Broding, S. Carrel, and F. J. Lejeune. 1993. Effects of systemic interferon- alpha (IFN-alpha) on the antigenic phenotype of melanoma metastases. EORTC melanoma group cooperative study No. 18852. Melanoma Res 3:173-180.

9. Haque, S. J., and B. R. Williams. 1998. Signal transduction in the interferon system. Semin Oncol 25:14-22.

10. Darnell, J. E., Jr., I. M. Kerr, and G. R. Stark. 1994. Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science 264:1415-1421.

148 11. Levy, D., N. Reich, D. Kessler, R. Pine, and J. E. Darnell, Jr. 1988. Transcriptional regulation of interferon-stimulated genes: a DNA response element and induced proteins that recognize it. Cold Spring Harb Symp Quant Biol 53 Pt 2:799-802.

12. Mirkovitch, J., T. Decker, and J. E. Darnell, Jr. 1992. Interferon induction of gene transcription analyzed by in vivo footprinting. Mol Cell Biol 12:1-9. 13. Nguyen, K. B., W. T. Watford, R. Salomon, S. R. Hofmann, G. C. Pien, A. Morinobu, M. Gadina, J. J. O'Shea, and C. A. Biron. 2002. Critical role for STAT4 activation by type 1 interferons in the interferon-gamma response to viral infection. Science 297:2063-2066.

14. Meraro, D., M. Gleit-Kielmanowicz, H. Hauser, and B. Z. Levi. 2002. IFN-stimulated gene 15 is synergistically activated through interactions between the myelocyte/lymphocyte-specific transcription factors, PU.1, IFN regulatory factor-8/IFN consensus sequence binding protein, and IFN regulatory factor-4: characterization of a new subtype of IFN-stimulated response element. J Immunol 168:6224-6231.

15. Hatina, V. J., J. Kralova, and P. Jansa. 1996. Identification of an intragenic interferon-stimulated response element sequence of the mouse class I major histocompatibility complex H-2Kb gene. Exp Clin Immunogenet 13:55-60.

16. Ohmori, Y., and T. A. Hamilton. 1995. The interferon-stimulated response element and a kappa B site mediate synergistic induction of murine IP-10 gene transcription by IFN-gamma and TNF-alpha. J Immunol 154:5235-5244.

17. Herberman, R. B. 1997. Effect of alpha-interferons on immune function. Semin Oncol 24:S9-78-S79-80.

18. Lesinski, G. B., M. Anghelina, J. Zimmerer, T. Bakalakos, B. Badgwell, R Parihar, Y. Hu, B. Becknell, G. Abood, A. R. Chaudhury, C. Magro, J. Durbin, and W. E. Carson, 3rd. 2003. The antitumor effects of IFN-alpha are abrogated in a STAT1- deficient mouse. J Clin Invest 112:170-180.

19. Bart, R. S., N. R. Porzio, A. W. Kopf, J. T. Vilcek, E. H. Cheng, and Y. Farcet. 1980. Inhibition of growth of B16 murine malignant melanoma by exogenous interferon. Cancer Res 40:614-619.

20. Kirkwood, J. M., M. H. Strawderman, M. S. Ernstoff, T. J. Smith, E. C. Borden, and R. H. Blum. 1996. Interferon alfa-2b adjuvant therapy of high-risk resected cutaneous melanoma: the Eastern Cooperative Oncology Group Trial EST 1684. J Clin Oncol 14:7-17.

21. Voelter-Mahlknecht, S., U. Mahlknecht, S. Letzel, and G. Fierlbeck. 2006. Phase 2 Trial of the Continuous IV Administration of Interferon-beta in Patients With Disseminated Malignant Melanoma. Skinmed 5:271-276.

149

22. Madan, V., P. Hoban, R. C. Strange, A. A. Fryer, and J. T. Lear. 2006. Genetics and risk factors for basal cell carcinoma. Br J Dermatol 154 Suppl 1:5-7.

23. Berkwick, M. 1998. Epidemiology: current trends, risk factors, and enviromental concerns. In Cutaneous Melanoma. C. M. Balch, ed, St. Louis, MO, p. 551-565. 24. Kadison, A. S., and D. L. Morton. 2003. Immunotherapy of malignant melanoma. Surg Clin North Am 83:343-370.

25. Jemal, A., R. Siegel, E. Ward, T. Murray, J. Xu, C. Smigal, and M. J. Thun. 2006. Cancer statistics, 2006. CA Cancer J Clin 56:106-130.

26. Lawler, P. E., and S. Schreiber. 1989. Cutaneous malignant melanoma: nursing's role in prevention and early detection. Oncol Nurs Forum 16:345-352.

27. Pho, L., D. Grossman, and S. A. Leachman. 2006. Melanoma genetics: a review of genetic factors and clinical phenotypes in familial melanoma. Curr Opin Oncol 18:173-179.

28. Goldstein, A. M., J. P. Struewing, A. Chidambaram, M. C. Fraser, and M. A. Tucker. 2000. Genotype-phenotype relationships in U.S. melanoma-prone families with CDKN2A and CDK4 mutations. J Natl Cancer Inst 92:1006-1010.

29. Ackermann, J., M. Frutschi, K. Kaloulis, T. McKee, A. Trumpp, and F. Beermann. 2005. Metastasizing melanoma formation caused by expression of activated N-RasQ61K on an INK4a-deficient background. Cancer Res 65:4005-4011.

30. Liu, W., J. W. Kelly, M. Trivett, W. K. Murray, J. P. Dowling, R. Wolfe, G. Mason, J. Magee, C. Angel, A. Dobrovic, and G. A. McArthur. 2006. Distinct Clinical and Pathological Features Are Associated with the BRAF(T1799A(V600E)) Mutation in Primary Melanoma. J Invest Dermatol.

31. Edlundh-Rose, E., S. Egyha Zi, K. Omholt, E. Mansson-Brahme, A. Platz, J. Hansson, and J. Lundeberg. 2006. NRAS and BRAF mutations in melanoma tumours in relation to clinical characteristics: a study based on mutation screening by pyrosequencing. Melanoma Res 16:471-478.

32. Kim, M., J. D. Gans, C. Nogueira, A. Wang, J. H. Paik, B. Feng, C. Brennan, W. C. Hahn, C. Cordon-Cardo, S. N. Wagner, T. J. Flotte, L. M. Duncan, S. R. Granter, and L. Chin. 2006. Comparative oncogenomics identifies NEDD9 as a melanoma metastasis gene. Cell 125:1269-1281.

33. De Smet, C., S. J. Courtois, I. Faraoni, C. Lurquin, J. P. Szikora, O. De Backer, and T. Boon. 1995. Involvement of two Ets binding sites in the transcriptional activation of the MAGE1 gene. Immunogenetics 42:282-290.

150

34. Sigalotti, L., S. Coral, G. Nardi, A. Spessotto, E. Cortini, I. Cattarossi, F. Colizzi, M. Altomonte, and M. Maio. 2002. Promoter methylation controls the expression of MAGE2, 3 and 4 genes in human cutaneous melanoma. J Immunother 25:16-26. 35. Lim, E., J. Browning, D. MacGregor, I. D. Davis, and J. S. Cebon. 2006. Desmoplastic melanoma: comparison of expression of differentiation antigens and cancer testis antigens. Melanoma Res 16:347-355.

36. Young, S. E., S. R. Martinez, and R. Essner. 2006. The role of surgery in treatment of stage IV melanoma. J Surg Oncol 94:344-351.

37. Dudley, M. E., J. R. Wunderlich, J. C. Yang, R. M. Sherry, S. L. Topalian, N. P. Restifo, R. E. Royal, U. Kammula, D. E. White, S. A. Mavroukakis, L. J. Rogers, G. J. Gracia, S. A. Jones, D. P. Mangiameli, M. M. Pelletier, J. Gea-Banacloche, M. R. Robinson, D. M. Berman, A. C. Filie, A. Abati, and S. A. Rosenberg. 2005. Adoptive cell transfer therapy following non-myeloablative but lymphodepleting chemotherapy for the treatment of patients with refractory metastatic melanoma. J Clin Oncol 23:2346-2357.

38. Liang, K. V., S. O. Sanderson, G. S. Nowakowski, and A. S. Arora. 2006. Metastatic malignant melanoma of the gastrointestinal tract. Mayo Clin Proc 81:511-516.

39. Stone, A., J. Cooper, K. L. Koenig, J. G. Golfinos, and R. Oratz. 2004. A comparison of survival rates for treatment of melanoma metastatic to the brain. Cancer Invest 22:492-497.

40. Einhorn, L. H., M. A. Burgess, C. Vallejos, G. P. Bodey, Sr., J. Gutterman, G. Mavligit, E. M. Hersh, J. K. Luce, E. Frei, 3rd, E. J. Freireich, and J. A. Gottlieb. 1974. Prognostic correlations and response to treatment in advanced metastatic malignant melanoma. Cancer Res 34:1995-2004.

41. Alexandrescu, D. T., J. P. Dutcher, and P. H. Wiernik. 2005. Metastatic melanoma: is biochemotherapy the future? Med Oncol 22:101-111.

42. Essner, R., J. H. Lee, L. A. Wanek, H. Itakura, and D. L. Morton. 2004. Contemporary surgical treatment of advanced-stage melanoma. Arch Surg 139:961-966; discussion 966-967.

43. Wong, S. L., and D. G. Coit. 2004. Role of surgery in patients with stage IV melanoma. Curr Opin Oncol 16:155-160.

44. Khan, M. A., S. Andrews, R. Ismail-Khan, P. N. Munster, S. Brem, J. King, D. S. Reintgen, V. K. Sondak, and A. I. Daud. 2006. Overall and progression-free survival in metastatic melanoma: analysis of a single-institution database. Cancer Control 13:211-217.

151 45. Tarhini, A. A., and S. S. Agarwala. 2006. Cutaneous melanoma: available therapy for metastatic disease. Dermatol Ther 19:19-25. 46. Serrone, L., M. Zeuli, F. M. Sega, and F. Cognetti. 2000. Dacarbazine-based chemotherapy for metastatic melanoma: thirty-year experience overview. J Exp Clin Cancer Res 19:21-34.

47. Eggermont, A. M., and J. M. Kirkwood. 2004. Re-evaluating the role of dacarbazine in metastatic melanoma: what have we learned in 30 years? Eur J Cancer 40:1825-1836.

48. Eton, O., S. S. Legha, A. Y. Bedikian, J. J. Lee, A. C. Buzaid, C. Hodges, S. E. Ring, N. E. Papadopoulos, C. Plager, M. J. East, F. Zhan, and R. S. Benjamin. 2002. Sequential biochemotherapy versus chemotherapy for metastatic melanoma: results from a phase III randomized trial. J Clin Oncol 20:2045-2052.

49. Sanderson, K., R. Scotland, P. Lee, D. Liu, S. Groshen, J. Snively, S. Sian, G. Nichol, T. Davis, T. Keler, M. Yellin, and J. Weber. 2005. Autoimmunity in a phase I trial of a fully human anti-cytotoxic T-lymphocyte antigen-4 monoclonal antibody with multiple melanoma peptides and Montanide ISA 51 for patients with resected stages III and IV melanoma. J Clin Oncol 23:741-750.

50. Carson WE, B. J., Shah N, Reddy K, Kefauver C, Leming PD, Kendra K, M Walker. 2003. A phase 2 trial of a recombinant humanized monoclonal anti-vascular endothelial growth factor (VEGF) antibody in patients with malignant melanoma. In Proc Am Soc Clin Oncol, Vol. 22; Abstract 2873, Alexandria, VA.

51. Eisen, T., T. Ahmad, K. T. Flaherty, M. Gore, S. Kaye, R. Marais, I. Gibbens, S. Hackett, M. James, L. M. Schuchter, K. L. Nathanson, C. Xia, R. Simantov, B. Schwartz, M. Poulin-Costello, P. J. O'Dwyer, and M. J. Ratain. 2006. Sorafenib in advanced melanoma: a Phase II randomised discontinuation trial analysis. Br J Cancer 95:581-586.

52. Papac, R. J. 1996. Spontaneous regression of cancer. Cancer Treat Rev 22:395-423.

53. Schneeberger, A., F. Koszik, and G. Stingl. 1995. Immunologic host defense in melanoma: delineation of effector mechanisms involved and of strategies for the augmentation of their efficacy. J Invest Dermatol 105:110S-116S.

54. 2001. Cutaneous melanoma. In Principeles and Practice of Oncology. S. H. VT DeVita, SA Rosenberg, ed. Lipincott, Williams &Wilkens, Philadelphia, PA, p. 1499-1547.

55. Atkins, M. B., M. T. Lotze, J. P. Dutcher, R. I. Fisher, G. Weiss, K. Margolin, J. Abrams, M. Sznol, D. Parkinson, M. Hawkins, C. Paradise, L. Kunkel, and S. A.

152 Rosenberg. 1999. High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17:2105-2116.

56. Rosenberg, S. A., J. C. Yang, D. E. White, and S. M. Steinberg. 1998. Durability of complete responses in patients with metastatic cancer treated with high-dose interleukin-2: identification of the antigens mediating response. Ann Surg 228:307-319.

57. Keilholz, U., C. Scheibenbogen, W. Tilgen, L. Bergmann, E. Weidmann, E. Seither, M. Richter, B. Brado, P. S. Mitrou, and W. Hunstein. 1993. Interferon-alpha and interleukin-2 in the treatment of metastatic melanoma. Comparison of two phase II trials. Cancer 72:607-614.

58. Dutcher, J. 2002. Current status of interleukin-2 therapy for metastatic renal cell carcinoma and metastatic melanoma. Oncology (Williston Park) 16:4-10.

59. Lens, M. B., and M. Dawes. 2002. Interferon alfa therapy for malignant melanoma: a systematic review of randomized controlled trials. J Clin Oncol 20:1818- 1825.

60. Tsao, H., M. B. Atkins, and A. J. Sober. 2004. Management of cutaneous melanoma. N Engl J Med 351:998-1012.

61. Kirkwood, J. M., M. S. Ernstoff, C. A. Davis, M. Reiss, R. Ferraresi, and S. A. Rudnick. 1985. Comparison of intramuscular and intravenous recombinant alpha-2 interferon in melanoma and other cancers. Ann Intern Med 103:32-36.

62. Gogas, H., J. Ioannovich, U. Dafni, C. Stavropoulou-Giokas, K. Frangia, D. Tsoutsos, P. Panagiotou, A. Polyzos, O. Papadopoulos, A. Stratigos, C. Markopoulos, D. Bafaloukos, D. Pectasides, G. Fountzilas, and J. M. Kirkwood. 2006. Prognostic significance of autoimmunity during treatment of melanoma with interferon. N Engl J Med 354:709-718.

63. Kirkwood, J. M., J. G. Ibrahim, V. K. Sondak, J. Richards, L. E. Flaherty, M. S. Ernstoff, T. J. Smith, U. Rao, M. Steele, and R. H. Blum. 2000. High- and low- dose interferon alfa-2b in high-risk melanoma: first analysis of intergroup trial E1690/S9111/C9190. J Clin Oncol 18:2444-2458.

64. Creagan, E. T., R. J. Dalton, D. L. Ahmann, S. H. Jung, R. F. Morton, R. M. Langdon, Jr., J. Kugler, and L. J. Rodrigue. 1995. Randomized, surgical adjuvant clinical trial of recombinant interferon alfa-2a in selected patients with malignant melanoma. J Clin Oncol 13:2776-2783.

65. Kirkwood, J. M., J. G. Ibrahim, J. A. Sosman, V. K. Sondak, S. S. Agarwala, M. S. Ernstoff, and U. Rao. 2001. High-dose interferon alfa-2b significantly prolongs

153 relapse-free and overall survival compared with the GM2-KLH/QS-21 vaccine in patients with resected stage IIB-III melanoma: results of intergroup trial E1694/S9512/C509801. J Clin Oncol 19:2370-2380.

66. Dorval, T., T. Palangie, M. Jouve, E. Garcia-Giralt, L. Israel, E. Falcoff, D. Schwab, and P. Pouillart. 1986. Clinical phase II trial of recombinant DNA interferon (interferon alpha 2b) in patients with metastatic malignant melanoma. Cancer 58:215- 218.

67. Sertoli, M. R., M. G. Bernengo, A. Ardizzoni, I. Brunetti, A. Falcone, M. G. Vidili, M. P. Cusimano, A. Appino, G. Doveil, C. Fortini, and et al. 1989. Phase II trial of recombinant alpha-2b interferon in the treatment of metastatic skin melanoma. Oncology 46:96-98.

68. Kefford, R. F. 2003. Adjuvant therapy of cutaneous melanoma: the interferon debate. Ann Oncol 14:358-365.

69. Kirkwood, J. M., C. Bender, S. Agarwala, A. Tarhini, J. Shipe-Spotloe, B. Smelko, S. Donnelly, and L. Stover. 2002. Mechanisms and management of toxicities associated with high-dose interferon alfa-2b therapy. J Clin Oncol 20:3703-3718.

70. Legha, S. S., S. Ring, O. Eton, A. Bedikian, A. C. Buzaid, C. Plager, and N. Papadopoulos. 1998. Development of a biochemotherapy regimen with concurrent administration of cisplatin, vinblastine, dacarbazine, interferon alfa, and interleukin-2 for patients with metastatic melanoma. J Clin Oncol 16:1752-1759.

71. Ridolfi, R., V. Chiarion-Sileni, M. Guida, A. Romanini, R. Labianca, A. Freschi, G. Lo Re, R. Nortilli, S. Brugnara, P. Vitali, and O. Nanni. 2002. Cisplatin, dacarbazine with or without subcutaneous interleukin-2, and interferon alpha-2b in advanced melanoma outpatients: results from an Italian multicenter phase III randomized clinical trial. J Clin Oncol 20:1600-1607.

72. Sabel, M. S., and V. K. Sondak. 2002. Melanoma vaccines: breakthrough or bust? Cancer Invest 20:1114-1116.

73. Baker, P. K., A. R. Pettitt, J. R. Slupsky, H. J. Chen, M. A. Glenn, M. Zuzel, and J. C. Cawley. 2002. Response of hairy cells to IFN-alpha involves induction of apoptosis through autocrine TNF-alpha and protection by adhesion. Blood 100:647-653.

74. Brender, C., M. Nielsen, K. Kaltoft, G. Mikkelsen, Q. Zhang, M. Wasik, N. Billestrup, and N. Odum. 2001. STAT3-mediated constitutive expression of SOCS-3 in cutaneous T-cell lymphoma. Blood 97:1056-1062.

154 75. Chin, Y. E., M. Kitagawa, W. C. Su, Z. H. You, Y. Iwamoto, and X. Y. Fu. 1996. Cell growth arrest and induction of cyclin-dependent kinase inhibitor p21 WAF1/CIP1 mediated by STAT1. Science 272:719-722.

76. Selleri, C., T. Sato, L. Del Vecchio, L. Luciano, A. J. Barrett, B. Rotoli, N. S. Young, and J. P. Maciejewski. 1997. Involvement of Fas-mediated apoptosis in the inhibitory effects of interferon-alpha in chronic myelogenous leukemia. Blood 89:957- 964.

77. Zhou, Y., S. Wang, B. G. Yue, A. Gobl, and K. Oberg. 2002. Effects of interferon alpha on the expression of p21cip1/waf1 and cell cycle distribution in carcinoid tumors. Cancer Invest 20:348-356.

78. Biron, C. A. 2001. Interferons alpha and beta as immune regulators--a new look. Immunity 14:661-664.

79. Brassard, D. L., M. J. Grace, and R. W. Bordens. 2002. Interferon-alpha as an immunotherapeutic protein. J Leukoc Biol 71:565-581.

80. Parronchi, P., S. Mohapatra, S. Sampognaro, L. Giannarini, U. Wahn, P. Chong, E. Maggi, H. Renz, and S. Romagnani. 1996. Effects of interferon-alpha on cytokine profile, T cell receptor repertoire and peptide reactivity of human allergen-specific T cells. Eur J Immunol 26:697-703.

81. Martin-Henao, G. A., R. Quiroga, A. Sureda, J. R. Gonzalez, V. Moreno, and J. Garcia. 2000. L-selectin expression is low on CD34+ cells from patients with chronic myeloid leukemia and interferon-a up-regulates this expression. Haematologica 85:139- 146.

82. Kirkwood, J. M., and M. S. Ernstoff. 1990. Role of interferons in the therapy of melanoma. J Invest Dermatol 95:180S-184S.

83. Carson, W. E. 1998. Interferon-alpha-induced activation of signal transducer and activator of transcription proteins in malignant melanoma. Clin Cancer Res 4:2219-2228.

84. Kaplan, D. H., V. Shankaran, A. S. Dighe, E. Stockert, M. Aguet, L. J. Old, and R. D. Schreiber. 1998. Demonstration of an interferon gamma-dependent tumor surveillance system in immunocompetent mice. Proc Natl Acad Sci U S A 95:7556-7561.

85. Pansky, A., P. Hildebrand, E. Fasler-Kan, L. Baselgia, S. Ketterer, C. Beglinger, and M. H. Heim. 2000. Defective Jak-STAT signal transduction pathway in melanoma cells resistant to growth inhibition by interferon-alpha. Int J Cancer 85:720-725.

86. Wong, L. H., I. Hatzinisiriou, R. J. Devenish, and S. J. Ralph. 1998. IFN-gamma priming up-regulates IFN-stimulated gene factor 3 (ISGF3) components, augmenting

155 responsiveness of IFN-resistant melanoma cells to type I IFNs. J Immunol 160:5475- 5484.

87. Belardelli, F., I. Gresser, C. Maury, and M. T. Maunoury. 1982. Antitumor effects of interferon in mice injected with interferon-sensitive and interferon-resistant Friend leukemia cells. II. Role of host mechanisms. Int J Cancer 30:821-825.

88. Belardelli, F., I. Gresser, C. Maury, and M. T. Maunoury. 1982. Antitumor effects of interferon in mice injected with interferon-sensitive and interferon-resistant Friend leukemia cells. I. Int J Cancer 30:813-820.

89. Belardelli, F., M. Ferrantini, E. Proietti, and J. M. Kirkwood. 2002. Interferon-alpha in tumor immunity and immunotherapy. Cytokine Growth Factor Rev 13:119-134.

90. Gresser, I., C. Maury, C. Carnaud, E. De Maeyer, M. T. Maunoury, and F. Belardelli. 1990. Anti-tumor effects of interferon in mice injected with interferon- sensitive and interferon-resistant Friend erythroleukemia cells. VIII. Role of the immune system in the inhibition of visceral metastases. Int J Cancer 46:468-474.

91. Gresser, I., T. Kaido, C. Maury, D. Woodrow, J. Moss, and F. Belardelli. 1994. Interaction of IFN alpha/beta with host cells essential to the early inhibition of Friend erythroleukemia visceral metastases in mice. Int J Cancer 57:604-611.

92. Fallarino, F., and T. F. Gajewski. 1999. Cutting edge: differentiation of antitumor CTL in vivo requires host expression of Stat1. J Immunol 163:4109-4113.

93. Dunn, G. P., A. T. Bruce, K. C. Sheehan, V. Shankaran, R. Uppaluri, J. D. Bui, M. S. Diamond, C. M. Koebel, C. Arthur, J. M. White, and R. D. Schreiber. 2005. A critical function for type I interferons in cancer immunoediting. Nat Immunol 6:722-729.

94. Koon, H., and M. Atkins. 2006. Autoimmunity and immunotherapy for cancer. N Engl J Med 354:758-760.

95. Moschos, S. J., H. D. Edington, S. R. Land, U. N. Rao, D. Jukic, J. Shipe-Spotloe, and J. M. Kirkwood. 2006. Neoadjuvant treatment of regional stage IIIB melanoma with high-dose interferon alfa-2b induces objective tumor regression in association with modulation of tumor infiltrating host cellular immune responses. J Clin Oncol 24:3164- 3171.

96. Alexander, W. S. 2002. Suppressors of cytokine signalling (SOCS) in the immune system. Nat Rev Immunol 2:410-416.

97. Endo, T. A., M. Masuhara, M. Yokouchi, R. Suzuki, H. Sakamoto, K. Mitsui, A. Matsumoto, S. Tanimura, M. Ohtsubo, H. Misawa, T. Miyazaki, N. Leonor,

156 T. Taniguchi, T. Fujita, Y. Kanakura, S. Komiya, and A. Yoshimura. 1997. A new protein containing an SH2 domain that inhibits JAK kinases. Nature 387:921-924.

98. Masuhara, M., H. Sakamoto, A. Matsumoto, R. Suzuki, H. Yasukawa, K. Mitsui, T. Wakioka, S. Tanimura, A. Sasaki, H. Misawa, M. Yokouchi, M. Ohtsubo, and A. Yoshimura. 1997. Cloning and characterization of novel CIS family genes. Biochem Biophys Res Commun 239:439-446.

99. Narazaki, M., M. Fujimoto, T. Matsumoto, Y. Morita, H. Saito, T. Kajita, K. Yoshizaki, T. Naka, and T. Kishimoto. 1998. Three distinct domains of SSI-1/SOCS- 1/JAB protein are required for its suppression of interleukin 6 signaling. Proc Natl Acad Sci U S A 95:13130-13134.

100. Yoshimura, A., T. Ohkubo, T. Kiguchi, N. A. Jenkins, D. J. Gilbert, N. G. Copeland, T. Hara, and A. Miyajima. 1995. A novel cytokine-inducible gene CIS encodes an SH2-containing protein that binds to tyrosine-phosphorylated interleukin 3 and erythropoietin receptors. Embo J 14:2816-2826.

101. Nicholson, S. E., T. A. Willson, A. Farley, R. Starr, J. G. Zhang, M. Baca, W. S. Alexander, D. Metcalf, D. J. Hilton, and N. A. Nicola. 1999. Mutational analyses of the SOCS proteins suggest a dual domain requirement but distinct mechanisms for inhibition of LIF and IL-6 signal transduction. Embo J 18:375-385.

102. Sasaki, A., H. Yasukawa, A. Suzuki, S. Kamizono, T. Syoda, I. Kinjyo, M. Sasaki, J. A. Johnston, and A. Yoshimura. 1999. Cytokine-inducible SH2 protein-3 (CIS3/SOCS3) inhibits Janus tyrosine kinase by binding through the N-terminal kinase inhibitory region as well as SH2 domain. Genes Cells 4:339-351.

103. Larsen, L., and C. Ropke. 2002. Suppressors of cytokine signalling: SOCS. Apmis 110:833-844.

104. Wormald, S., and D. J. Hilton. 2004. Inhibitors of cytokine signal transduction. J Biol Chem 279:821-824.

105. Yasukawa, H., H. Misawa, H. Sakamoto, M. Masuhara, A. Sasaki, T. Wakioka, S. Ohtsuka, T. Imaizumi, T. Matsuda, J. N. Ihle, and A. Yoshimura. 1999. The JAK-binding protein JAB inhibits Janus tyrosine kinase activity through binding in the activation loop. Embo J 18:1309-1320.

106. Naka, T., M. Narazaki, M. Hirata, T. Matsumoto, S. Minamoto, A. Aono, N. Nishimoto, T. Kajita, T. Taga, K. Yoshizaki, S. Akira, and T. Kishimoto. 1997. Structure and function of a new STAT-induced STAT inhibitor. Nature 387:924-929.

157 107. Saito, H., Y. Morita, M. Fujimoto, M. Narazaki, T. Naka, and T. Kishimoto. 2000. IFN regulatory factor-1-mediated transcriptional activation of mouse STAT- induced STAT inhibitor-1 gene promoter by IFN-gamma. J Immunol 164:5833-5843.

108. Auernhammer, C. J., C. Bousquet, and S. Melmed. 1999. Autoregulation of pituitary corticotroph SOCS-3 expression: characterization of the murine SOCS-3 promoter. Proc Natl Acad Sci U S A 96:6964-6969.

109. Matsumoto, A., M. Masuhara, K. Mitsui, M. Yokouchi, M. Ohtsubo, H. Misawa, A. Miyajima, and A. Yoshimura. 1997. CIS, a cytokine inducible SH2 protein, is a target of the JAK-STAT5 pathway and modulates STAT5 activation. Blood 89:3148-3154.

110. Sakai, I., K. Takeuchi, H. Yamauchi, H. Narumi, and S. Fujita. 2002. Constitutive expression of SOCS3 confers resistance to IFN-alpha in chronic myelogenous leukemia cells. Blood 100:2926-2931.

111. Song, M. M., and K. Shuai. 1998. The suppressor of cytokine signaling (SOCS) 1 and SOCS3 but not SOCS2 proteins inhibit interferon-mediated antiviral and antiproliferative activities. J Biol Chem 273:35056-35062.

112. Brender, C., M. Nielsen, C. Ropke, M. H. Nissen, A. Svejgaard, N. Billestrup, C. Geisler, and N. Odum. 2001. Interferon-alpha induces transient suppressors of cytokine signalling expression in human T cells. Exp Clin Immunogenet 18:80-85.

113. Maellaro, E., L. Pacenti, B. Del Bello, M. A. Valentini, P. Mangiavacchi, C. De Felice, P. Rubegni, P. Luzi, and C. Miracco. 2003. Different effects of interferon- alpha on melanoma cell lines: a study on telomerase reverse transcriptase, telomerase activity and apoptosis. Br J Dermatol 148:1115-1124.

114. Bauer, J. A., B. H. Morrison, R. W. Grane, B. S. Jacobs, E. C. Borden, and D. J. Lindner. 2003. IFN-alpha2b and thalidomide synergistically inhibit tumor-induced angiogenesis. J Interferon Cytokine Res 23:3-10.

115. Garbe, C., K. Krasagakis, C. C. Zouboulis, K. Schroder, S. Kruger, R. Stadler, and C. E. Orfanos. 1990. Antitumor activities of interferon alpha, beta, and gamma and their combinations on human melanoma cells in vitro: changes of proliferation, melanin synthesis, and immunophenotype. J Invest Dermatol 95:231S-237S.

116. Lesinski, G. B., S. V. Kondadasula, T. Crespin, L. Shen, K. Kendra, M. Walker, and W. E. Carson, 3rd. 2004. Multiparametric flow cytometric analysis of inter-patient variation in STAT1 phosphorylation following interferon Alfa immunotherapy. J Natl Cancer Inst 96:1331-1342.

158 117. Badgwell, B., G. B. Lesinski, C. Magro, G. Abood, A. Skaf, and W. Carson, 3rd. 2004. The antitumor effects of interferon-alpha are maintained in mice challenged with a STAT1-deficient murine melanoma cell line. J Surg Res 116:129-136.

118. Durbin, J. E., R. Hackenmiller, M. C. Simon, and D. E. Levy. 1996. Targeted disruption of the mouse Stat1 gene results in compromised innate immunity to viral disease. Cell 84:443-450.

119. Moschella, F., B. Bisikirska, A. Maffei, K. P. Papadopoulos, D. Skerrett, Z. Liu, C. S. Hesdorffer, and P. E. Harris. 2003. Gene expression profiling and functional activity of human dendritic cells induced with IFN-alpha-2b: implications for cancer immunotherapy. Clin Cancer Res 9:2022-2031.

120. Li, C., and W. H. Wong. 2003. The analysis of gene expression data: methods and software. DNA-Chip Analyzer (dChip).

121. Tokuhiro, S., R. Yamada, X. Chang, A. Suzuki, Y. Kochi, T. Sawada, M. Suzuki, M. Nagasaki, M. Ohtsuki, M. Ono, H. Furukawa, M. Nagashima, S. Yoshino, A. Mabuchi, A. Sekine, S. Saito, A. Takahashi, T. Tsunoda, Y. Nakamura, and K. Yamamoto. 2003. An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis. Nat Genet 35:341- 348.

122. Gil, M. P., E. Bohn, A. K. O'Guin, C. V. Ramana, B. Levine, G. R. Stark, H. W. Virgin, and R. D. Schreiber. 2001. Biologic consequences of Stat1-independent IFN signaling. Proc Natl Acad Sci U S A 98:6680-6685.

123. Ramana, C. V., M. P. Gil, Y. Han, R. M. Ransohoff, R. D. Schreiber, and G. R. Stark. 2001. Stat1-independent regulation of gene expression in response to IFN-gamma. Proc Natl Acad Sci U S A 98:6674-6679.

124. Mullbacher, A., P. Waring, R. Tha Hla, T. Tran, S. Chin, T. Stehle, C. Museteanu, and M. M. Simon. 1999. Granzymes are the essential downstream effector molecules for the control of primary virus infections by cytolytic leukocytes. Proc Natl Acad Sci U S A 96:13950-13955.

125. Qing, Y., and G. R. Stark. 2004. Alternative activation of STAT1 and STAT3 in response to interferon-gamma. J Biol Chem 279:41679-41685.

126. Zimmerer, J. M., Lesinski, G.B., Kornacker, K., Shen, L., Liyanarachichi, S., Durbin, J., Kendra, K., Walker, M. and W.E. Carson. 2003. Gene expression profiling of the response to interferon-alpha immunotherapy. In American Association for Cancer Research Annual Meeting, Orlando, FL. Abstract #4679.

159 127. Zimmerer, J. M., Lesinski, G.B., Kondadasula, S.V., Becknell, B., and W.E. Carson. 2006. Suppressors of cytokine signaling proteins 1 and 3 regulate the immune response to interferon-alpha. In American Association for Cancer Research Annual Meeting, Washington, D.C. Abstract #5560.

128. Lesinski, G. B., D. Valentino, E. M. Hade, S. Jones, C. Magro, A. R. Chaudhury, M. J. Walker, and W. E. Carson, 3rd. 2005. Expression of STAT1 and STAT2 in malignant melanoma does not correlate with response to interferon-alpha adjuvant therapy. Cancer Immunol Immunother.

129. Parihar, R., J. Dierksheide, Y. Hu, and W. E. Carson. 2002. IL-12 enhances the natural killer cell cytokine response to Ab-coated tumor cells. J Clin Invest 110:983-992.

130. Honda, K., H. Yanai, H. Negishi, M. Asagiri, M. Sato, T. Mizutani, N. Shimada, Y. Ohba, A. Takaoka, N. Yoshida, and T. Taniguchi. 2005. IRF-7 is the master regulator of type-I interferon-dependent immune responses. Nature 434:772-777.

131. Takaoka, A., H. Yanai, S. Kondo, G. Duncan, H. Negishi, T. Mizutani, S. Kano, K. Honda, Y. Ohba, T. W. Mak, and T. Taniguchi. 2005. Integral role of IRF-5 in the gene induction programme activated by Toll-like receptors. Nature 434:243-249.

132. Ji, X., R. Cheung, S. Cooper, Q. Li, H. B. Greenberg, and X. S. He. 2003. Interferon alfa regulated gene expression in patients initiating interferon treatment for chronic hepatitis C. Hepatology 37:610-621.

133. Schlaak, J. F., C. M. Hilkens, A. P. Costa-Pereira, B. Strobl, F. Aberger, A. M. Frischauf, and I. M. Kerr. 2002. Cell-type and donor-specific transcriptional responses to interferon-alpha. Use of customized gene arrays. J Biol Chem 277:49428-49437.

134. Zhu, H., H. Zhao, C. D. Collins, S. E. Eckenrode, Q. Run, R. A. McIndoe, J. M. Crawford, D. R. Nelson, J. X. She, and C. Liu. 2003. Gene expression associated with interferon alfa antiviral activity in an HCV replicon cell line. Hepatology 37:1180- 1188.

135. Wang, X., Z. H. Yuan, L. J. Zheng, F. Yu, W. Xiong, J. X. Liu, G. X. Hu, and Y. Li. 2004. Gene expression profiles in an hepatitis B virus transfected hepatoblastoma cell line and differentially regulated gene expression by interferon-alpha. World J Gastroenterol 10:1740-1745.

136. Geiss, G. K., V. S. Carter, Y. He, B. K. Kwieciszewski, T. Holzman, M. J. Korth, C. A. Lazaro, N. Fausto, R. E. Bumgarner, and M. G. Katze. 2003. Gene expression profiling of the cellular transcriptional network regulated by alpha/beta interferon and its partial attenuation by the hepatitis C virus nonstructural 5A protein. J Virol 77:6367- 6375.

160 137. Chang, Y. E., and L. A. Laimins. 2000. Microarray analysis identifies interferon-inducible genes and Stat-1 as major transcriptional targets of human papillomavirus type 31. J Virol 74:4174-4182.

138. Gomez, D., and N. C. Reich. 2003. Stimulation of primary human endothelial cell proliferation by IFN. J Immunol 170:5373-5381.

139. Stroncek, D. F., C. Basil, D. Nagorsen, S. Deola, E. Arico, K. Smith, E. Wang, F. Marincola, and M. C. Panelli. 2005. Delayed polarization of mononuclear phagocyte transcriptional program by type I interferon isoforms. J Transl Med 3:24.

140. Whitney, A. R., M. Diehn, S. J. Popper, A. A. Alizadeh, J. C. Boldrick, D. A. Relman, and P. O. Brown. 2003. Individuality and variation in gene expression patterns in human blood. Proc Natl Acad Sci U S A 100:1896-1901.

141. Boldrick, J. C., A. A. Alizadeh, M. Diehn, S. Dudoit, C. L. Liu, C. E. Belcher, D. Botstein, L. M. Staudt, P. O. Brown, and D. A. Relman. 2002. Stereotyped and specific gene expression programs in human innate immune responses to bacteria. Proc Natl Acad Sci U S A 99:972-977.

142. Lauerova, L., L. Dusek, M. Simickova, I. Kocak, M. Vagundova, J. Zaloudik, and J. Kovarik. 2002. Malignant melanoma associates with Th1/Th2 imbalance that coincides with disease progression and immunotherapy response. Neoplasma 49:159-166.

143. Legha, S. S. 1986. Interferons in the treatment of malignant melanoma. A review of recent trials. Cancer 57:1675-1677.

144. Salmon, S. E., B. G. Durie, L. Young, R. M. Liu, P. W. Trown, and N. Stebbing. 1983. Effects of cloned human leukocyte interferons in the human tumor stem cell assay. J Clin Oncol 1:217-225.

145. Giacomini, P., A. Aguzzi, S. Pestka, P. B. Fisher, and S. Ferrone. 1984. Modulation by recombinant DNA leukocyte (alpha) and fibroblast (beta) interferons of the expression and shedding of HLA- and tumor-associated antigens by human melanoma cells. J Immunol 133:1649-1655.

146. Parangi, S., M. O'Reilly, G. Christofori, L. Holmgren, J. Grosfeld, J. Folkman, and D. Hanahan. 1996. Antiangiogenic therapy of transgenic mice impairs de novo tumor growth. Proc Natl Acad Sci U S A 93:2002-2007.

147. Fleisher, T. A., S. E. Dorman, J. A. Anderson, M. Vail, M. R. Brown, and S. M. Holland. 1999. Detection of intracellular phosphorylated STAT-1 by flow cytometry. Clin Immunol 90:425-430. 148. Nieminen, J., C. St-Pierre, and S. Sato. 2005. Galectin-3 interacts with naive and primed neutrophils, inducing innate immune responses. J Leukoc Biol 78:1127-1135.

161

149. Schuchter, L. M. 2004. Adjuvant interferon therapy for melanoma: high-dose, low-dose, no dose, which dose? J Clin Oncol 22:7-10.

150. Sleijfer, S., M. Bannink, A. R. Van Gool, W. H. Kruit, and G. Stoter. 2005. Side effects of interferon-alpha therapy. Pharm World Sci 27:423-431.

151. Cascinelli, N., F. Belli, R. M. MacKie, M. Santinami, R. Bufalino, and A. Morabito. 2001. Effect of long-term adjuvant therapy with interferon alpha-2a in patients with regional node metastases from cutaneous melanoma: a randomised trial. Lancet 358:866-869.

152. Hancock, B. W., K. Wheatley, S. Harris, N. Ives, G. Harrison, J. M. Horsman, M. R. Middleton, N. Thatcher, P. C. Lorigan, J. R. Marsden, L. Burrows, and M. Gore. 2004. Adjuvant interferon in high-risk melanoma: the AIM HIGH Study--United Kingdom Coordinating Committee on Cancer Research randomized study of adjuvant low-dose extended-duration interferon Alfa-2a in high-risk resected malignant melanoma. J Clin Oncol 22:53-61.

153. Eggermont, A. M., S. Suciu, R. MacKie, W. Ruka, A. Testori, W. Kruit, C. J. Punt, M. Delauney, F. Sales, G. Groenewegen, D. J. Ruiter, I. Jagiello, K. Stoitchkov, U. Keilholz, and D. Lienard. 2005. Post-surgery adjuvant therapy with intermediate doses of interferon alfa 2b versus observation in patients with stage IIb/III melanoma (EORTC 18952): randomised controlled trial. Lancet 366:1189-1196.

154. Nathan, P. D., and T. G. Eisen. 2002. The biological treatment of renal-cell carcinoma and melanoma. Lancet Oncol 3:89-96.

155. Lesinski, G. B., B. Badgwell, J. Zimmerer, T. Crespin, Y. Hu, G. Abood, and W. E. Carson, 3rd. 2004. IL-12 pretreatments enhance IFN-alpha-induced Janus kinase- STAT signaling and potentiate the antitumor effects of IFN-alpha in a murine model of malignant melanoma. J Immunol 172:7368-7376.

156. Berkelhammer, J., R. W. Oxenhandler, R. R. Hook, Jr., and J. M. Hennessy. 1982. Development of a new melanoma model in C57BL/6 mice. Cancer Res 42:3157-3163.

157. Marine, J. C., D. J. Topham, C. McKay, D. Wang, E. Parganas, D. Stravopodis, A. Yoshimura, and J. N. Ihle. 1999. SOCS1 deficiency causes a lymphocyte-dependent perinatal lethality. Cell 98:609-616.

158. Marine, J. C., C. McKay, D. Wang, D. J. Topham, E. Parganas, H. Nakajima, H. Pendeville, H. Yasukawa, A. Sasaki, A. Yoshimura, and J. N. Ihle. 1999. SOCS3 is essential in the regulation of fetal liver erythropoiesis. Cell 98:617-627.

162 159. Lengyel, P. 1993. Tumor-suppressor genes: news about the interferon connection. Proc Natl Acad Sci U S A 90:5893-5895.

160. Carr, D. J., J. Chodosh, J. Ash, and T. E. Lane. 2003. Effect of anti-CXCL10 monoclonal antibody on herpes simplex virus type 1 keratitis and retinal infection. J Virol 77:10037-10046.

161. Gregorieff, A., S. Pyronnet, N. Sonenberg, and A. Veillette. 2000. Regulation of SOCS-1 expression by translational repression. J Biol Chem 275:21596-21604.

162. Becknell, B., R. Trotta, J. Yu, W. Ding, H. C. Mao, T. Hughes, T. Marburger, and M. A. Caligiuri. 2005. Efficient infection of human natural killer cells with an EBV/retroviral hybrid vector. J Immunol Methods 296:115-123.

163. Azogui, O., M. F. Avril, A. Margulis, M. Guillard, B. Caillou, and M. Prade. 1991. Tumor-infiltrating CD3- NK cells are more effective than CD3+ T cells in killing autologous melanoma cells. J Invest Dermatol 97:425-429.

164. Xu, D., P. Gu, P. Y. Pan, Q. Li, A. I. Sato, and S. H. Chen. 2004. NK and CD8+ T cell-mediated eradication of poorly immunogenic B16-F10 melanoma by the combined action of IL-12 gene therapy and 4-1BB costimulation. Int J Cancer 109:499-506.

165. Wargo, J. A., L. Y. Schumacher, B. Comin-Anduix, V. B. Dissette, J. A. Glaspy, W. H. McBride, L. H. Butterfield, J. S. Economou, and A. Ribas. 2005. Natural killer cells play a critical role in the immune response following immunization with melanoma- antigen-engineered dendritic cells. Cancer Gene Ther 12:516-527.

166. Horvat, S., and J. F. Medrano. 2001. Lack of Socs2 expression causes the high- growth phenotype in mice. Genomics 72:209-212.

167. Srahna, M., L. A. Van Grunsven, J. E. Remacle, and P. Vandenberghe. 2006. CTLA-4 interacts with STAT5 and inhibits STAT5-mediated transcription. Immunology 117:396-401.

168. Wang, Q., Y. Miyakawa, N. Fox, and K. Kaushansky. 2000. Interferon-alpha directly represses megakaryopoiesis by inhibiting thrombopoietin-induced signaling through induction of SOCS-1. Blood 96:2093-2099.

169. Ishida, N., K. Oritani, M. Shiraga, H. Yoshida, S. Kawamoto, H. Ujiie, H. Masaie, M. Ichii, Y. Tomiyama, and Y. Kanakura. 2005. Differential effects of a novel IFN- zeta/limitin and IFN-alpha on signals for Daxx induction and Crk phosphorylation that couple with growth control of megakaryocytes. Exp Hematol 33:495-503.

163 170. Li, Z., D. Metze, D. Nashan, C. Muller-Tidow, H. L. Serve, C. Poremba, T. A. Luger, and M. Bohm. 2004. Expression of SOCS-1, suppressor of cytokine signalling-1, in human melanoma. J Invest Dermatol 123:737-745.

171. Crespo, A., M. B. Filla, S. W. Russell, and W. J. Murphy. 2000. Indirect induction of suppressor of cytokine signalling-1 in macrophages stimulated with bacterial lipopolysaccharide: partial role of autocrine/paracrine interferon-alpha/beta. Biochem J 349:99-104.

172. Egwuagu, C. E., C. R. Yu, M. Zhang, R. M. Mahdi, S. J. Kim, and I. Gery. 2002. Suppressors of cytokine signaling proteins are differentially expressed in Th1 and Th2 cells: implications for Th cell lineage commitment and maintenance. J Immunol 168:3181-3187.

173. Yu, C. R., R. M. Mahdi, S. Ebong, B. P. Vistica, I. Gery, and C. E. Egwuagu. 2003. Suppressor of cytokine signaling 3 regulates proliferation and activation of T- helper cells. J Biol Chem 278:29752-29759.

174. Vlotides, G., A. S. Sorensen, F. Kopp, K. Zitzmann, N. Cengic, S. Brand, R. Zachoval, and C. J. Auernhammer. 2004. SOCS-1 and SOCS-3 inhibit IFN-alpha- induced expression of the antiviral proteins 2,5-OAS and MxA. Biochem Biophys Res Commun 320:1007-1014.

175. Brand, S., K. Zitzmann, J. Dambacher, F. Beigel, T. Olszak, G. Vlotides, S. T. Eichhorst, B. Goke, H. Diepolder, and C. J. Auernhammer. 2005. SOCS-1 inhibits expression of the antiviral proteins 2',5'-OAS and MxA induced by the novel interferon- lambdas IL-28A and IL-29. Biochem Biophys Res Commun 331:543-548.

176. Imanaka, K., S. Tamura, K. Fukui, N. Ito, S. Kiso, Y. Imai, T. Naka, T. Kishimoto, S. Kawata, and Y. Shinomura. 2005. Enhanced expression of suppressor of cytokine signalling-1 in the liver of chronic hepatitis C: possible involvement in resistance to interferon therapy. J Viral Hepat 12:130-138.

177. Davey, G. M., R. Starr, A. L. Cornish, J. T. Burghardt, W. S. Alexander, F. R. Carbone, C. D. Surh, and W. R. Heath. 2005. SOCS-1 regulates IL-15-driven homeostatic proliferation of antigen-naive CD8 T cells, limiting their autoimmune potential. J Exp Med 202:1099-1108.

178. Eyles, J. L., D. Metcalf, M. J. Grusby, D. J. Hilton, and R. Starr. 2002. Negative regulation of interleukin-12 signaling by suppressor of cytokine signaling-1. J Biol Chem 277:43735-43740.

179. Hanada, T., H. Yoshida, S. Kato, K. Tanaka, K. Masutani, J. Tsukada, Y. Nomura, H. Mimata, M. Kubo, and A. Yoshimura. 2003. Suppressor of cytokine

164 signaling-1 is essential for suppressing dendritic cell activation and systemic autoimmunity. Immunity 19:437-450.

180. Shen, L., K. Evel-Kabler, R. Strube, and S. Y. Chen. 2004. Silencing of SOCS1 enhances antigen presentation by dendritic cells and antigen-specific anti-tumor immunity. Nat Biotechnol 22:1546-1553.

181. Hanada, T., K. Tanaka, Y. Matsumura, M. Yamauchi, H. Nishinakamura, H. Aburatani, R. Mashima, M. Kubo, T. Kobayashi, and A. Yoshimura. 2005. Induction of hyper Th1 cell-type immune responses by dendritic cells lacking the suppressor of cytokine signaling-1 gene. J Immunol 174:4325-4332.

182. Roman-Gomez, J., A. Jimenez-Velasco, J. A. Castillejo, F. Cervantes, M. Barrios, D. Colomer, A. Heiniger, and A. Torres. 2004. The suppressor of cytokine signaling-1 is constitutively expressed in chronic myeloid leukemia and correlates with poor cytogenetic response to interferon-alpha. Haematologica 89:42-48.

183. Fenner, J. E., R. Starr, A. L. Cornish, J. G. Zhang, D. Metcalf, R. D. Schreiber, K. Sheehan, D. J. Hilton, W. S. Alexander, and P. J. Hertzog. 2006. Suppressor of cytokine signaling 1 regulates the immune response to infection by a unique inhibition of type I interferon activity. Nat Immunol 7:33-39.

184. Chong, M. M., A. L. Cornish, R. Darwiche, E. G. Stanley, J. F. Purton, D. I. Godfrey, D. J. Hilton, R. Starr, W. S. Alexander, and T. W. Kay. 2003. Suppressor of cytokine signaling-1 is a critical regulator of interleukin-7-dependent CD8+ T cell differentiation. Immunity 18:475-487.

185. Shresta, S., J. W. Heusel, D. M. Macivor, R. L. Wesselschmidt, J. H. Russell, and T. J. Ley. 1995. Granzyme B plays a critical role in cytotoxic lymphocyte-induced apoptosis. Immunol Rev 146:211-221.

186. Lin, M. Y., T. Zal, I. L. Ch'en, N. R. Gascoigne, and S. M. Hedrick. 2005. A pivotal role for the multifunctional calcium/calmodulin-dependent protein kinase II in T cells: from activation to unresponsiveness. J Immunol 174:5583-5592.

187. Jaakkola, I., M. Merinen, S. Jalkanen, and A. Hanninen. 2003. Ly6C induces clustering of LFA-1 (CD11a/CD18) and is involved in subtype-specific adhesion of CD8 T cells. J Immunol 170:1283-1290.

188. Holan, V., J. Pindjakova, M. Krulova, A. Neuwirth, J. Fric, and A. Zajicova. 2006. Production of nitric oxide during graft rejection is regulated by the Th1/Th2 balance, the arginase activity, and L-arginine metabolism. Transplantation 81:1708- 1715.

165 189. Sabel, M. S., and V. K. Sondak. 2003. Is there a role for adjuvant high-dose interferon-alpha-2b in the management of melanoma? Drugs 63:1053-1058.

190. Rini, B. I., S. Halabi, J. Taylor, E. J. Small, and R. L. Schilsky. 2004. Cancer and Leukemia Group B 90206: A randomized phase III trial of interferon-alpha or interferon- alpha plus anti-vascular endothelial growth factor antibody (bevacizumab) in metastatic renal cell carcinoma. Clin Cancer Res 10:2584-2586.

191. Hauschild, A., and U. R. Kleeberg. 2006. [Adjuvant therapy of melanoma. From non-specific immune stimulants into the future]. Hautarzt 57:764-772.

192. Kaufmann, R., K. Spieth, U. Leiter, C. Mauch, P. von den Driesch, T. Vogt, R. Linse, W. Tilgen, D. Schadendorf, J. C. Becker, G. Sebastian, S. Krengel, L. Kretschmer, C. Garbe, and R. Dummer. 2005. Temozolomide in combination with interferon-alfa versus temozolomide alone in patients with advanced metastatic melanoma: a randomized, phase III, multicenter study from the Dermatologic Cooperative Oncology Group. J Clin Oncol 23:9001-9007.

193. Kivela, T., S. Suciu, J. Hansson, W. H. Kruit, M. S. Vuoristo, O. Kloke, M. Gore, M. Hahka-Kemppinen, L. M. Parvinen, E. Kumpulainen, Y. Humblet, and S. Pyrhonen. 2003. Bleomycin, vincristine, lomustine and dacarbazine (BOLD) in combination with recombinant interferon alpha-2b for metastatic uveal melanoma. Eur J Cancer 39:1115- 1120.

194. Agarwala, S. S., and J. M. Kirkwood. 2003. Temozolomide in combination with interferon alpha-2b in patients with metastatic melanoma: a phase I dose-escalation study. Cancer 97:121-127.

166