Gene Expression Profiling in Hematopoietic Stem Cell Transplantation

Material Information

Gene Expression Profiling in Hematopoietic Stem Cell Transplantation
Buzzeo, Matthew Peter
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource

Thesis/Dissertation Information

Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences
Committee Chair:
Reddy, Vijay S.
Committee Members:
Liu, Chen
Schultz, Gregory S.
Graduation Date:


Subjects / Keywords:
Blood ( jstor )
Bone marrow ( jstor )
Cells ( jstor )
Cytokines ( jstor )
Diseases ( jstor )
Gene expression ( jstor )
Interleukins ( jstor )
Molecules ( jstor )
Receptors ( jstor )
Stem cells ( jstor )
Medicine -- Dissertations, Academic -- UF
allograft, gcsf, gvhd, hsct, microarray, pbsct
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Medical Sciences thesis, M.S


Allogeneic peripheral blood stem cell transplantation (PBSCT) is the treatment of choice for high-risk hematological malignancies. In PBSCT, genetically compatible donors are administered granulocyte-colony stimulating factor (G-CSF) to mobilize stem cells from their bone marrow. Despite a significantly higher number of T cells contained in the G-CSF-mobilized stem cell allograft as compared to a bone marrow allograft, the risk of acute-graft-versus-host disease (aGVHD) is not increased. This effect is attributed to the ability of G-CSF to exert immunosuppressive effects on the allograft predominantly through increased tolerogenic dendritic cell (DC) counts, which polarize donor na?ve T cells toward a T helper-type 2 (Th2) phenotype. Specifically, Th2 cells and other donor-derived tolerogenic cells (e.g., monocytes, regulatory T cells) decrease aGVHD risk by increasing levels of immunosuppressive cytokines such as interleukin-4 (IL-4) and IL-10, thereby antagonizing post-transplant alloreactive events. Additional roles for allograft composition in influencing the development of aGVHD are not well described, however, and few definitive markers exist for increased aGVHD risk or aGVHD progression to advanced and potentially fatal stages. Emerging high-throughput technologies such as DNA microarrays can be used as powerful tools for addressing these problems. ( ,,,,,,, )
Our study used microarray technology to report the gene expression profile of the G-CSF-mobilized allograft and to discover novel transcriptional markers that may predict development of aGVHD. We enrolled 5 donors and 8 recipients (4 control, 4 aGVHD) for a University of Florida IRB-approved study and collected peripheral blood after patients had given informed consent. For donors, blood was collected before and after G-CSF administration; for recipients, blood was collected at the time of engraftment -- a total of 18 samples. For each sample, RNA was purified from whole blood leukocytes and used to generate cRNA which was subsequently hybridized onto Affymetrix GeneChip® Human Genome U133 Plus 2.0 microarrays. Significant changes in gene expression were assessed by controlling the false discovery rate (FDR) between 5% and 10%. Additionally, relevant immune cells—including monocytes, granulocytes, lymphocytes and dendritic cells—were enumerated via complete blood count (CBC) with differential and flow cytometry.
In donors, although G-CSF drastically increased allograft immune cell content, adaptive immune response genes were significantly down-regulated, including antigen presentation, T cell activation and cytolytic effector response pathways. In recipients, several novel, putative transcriptional markers of aGVHD were discovered, such as the pro-inflammatory cytokine IL-27 and the autoimmune regulator gene AIRE. In conclusion, this study offers insight into the applications of microarray technology toward understanding and predicting outcomes of hematopoietic stem cell transplantation.
Statement of Responsibility:
by Matthew Peter Buzzeo

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Matthew Peter Buzzeo. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:


This item has the following downloads:

Full Text




2 2007 Matthew Peter Buzzeo


3 To my family: Christine, Robert A. and Robert W. Buzzeo


4 ACKNOWLEDGMENTS I wish to thank Christina Cline (Shands Bone Marrow Transplant Unit) for data management support and Jose Iturraspe (Hematopa thology Associates, LLC) for flow cytometry technical assistance. I am grateful to Dr. Marina-Telonis Scott (ICBR Microarray Core) for training me in technical aspects of microarray anal ysis, and Drs. Jie Yang and George Casella (Department of Statistics) for assistance with st atistical analyses. I thank Maxine Rushing, Ann Yarborough and Jennifer Frazier (D ivision of Hematology/Oncology) for their assistance with fiscal and clerical matters. I thank members of the Reddy-Lab for thei r support and Masters Program Coordinators Dr. Donna Duckworth and Joyce Conners for their diligence in keeping me and the rest of the M.S. students on tr ack. Finally, I thank my supervisory committee members: Drs. Vijay Reddy (Chair), Greg Schultz and Chen Liu for their guidance. This study was funded in part by the University of Florida Shands Cancer Center.


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........6 LIST OF FIGURES................................................................................................................ .........7 ABSTRACT....................................................................................................................... ..............8 CHAPTER 1 GENERAL BACKGROUND................................................................................................... 10 1.1 Hematopoietic Stem Cell Transplantation........................................................................ 10 1.1.1 Donor Selection......................................................................................................11 1.1.2 Stem Cell Mobilization with G-CSF...................................................................... 12 1.1.3 Recipient Conditioning and Transplantation.......................................................... 12 1.2 Graft-versus-Host Disease................................................................................................13 1.3 Immunomodulatory Pr operties of G-CSF........................................................................ 16 1.4 Microarray Analysis and Gene Expression Profiling in Medicine................................... 18 1.4 Current Study Aims......................................................................................................... .19 2 MATERIALS AND METHODS............................................................................................ 21 2.1 Enrollment of Subjects..................................................................................................... 21 2.2 Sample Acquisition and Processing.................................................................................. 21 2.3 Complete Blood Counts and Flow Cytometry..................................................................22 2.4 Microarray Anlaysis........................................................................................................ .23 3 RESULTS...............................................................................................................................24 3.1 Effects of G-CSF on Allograft Composition.................................................................... 24 3.2 Relevant Transcriptional Responses to G-CSF................................................................24 3.3 Gene Expression Profile of aGVHD.................................................................................25 4 DISCUSSION.........................................................................................................................38 4.1 Microarray and Flow Cytometric Analysis of the PBSCT Allograft............................... 38 4.2 Toward a Molecular Signature of aGVHD.......................................................................43 5 CONCLUSIONS AND FUTURE AIMS............................................................................... 46 LIST OF REFERENCES...............................................................................................................48 BIOGRAPHICAL SKETCH.........................................................................................................55


6 LIST OF TABLES Table page 3-1 Donor Blood Counts and Flow Cytometr y: Pre vs. Post GCSF Mobilization......................28 3-2 Immune-Related Transcriptiona l Response to G-CSF Mobilization.....................................29 3-3 Characteristics of aGVH D and aGVHD-Free Recipients......................................................33 3-4 Immune-Related Transcriptiona l Response in aGVHD at Engraftment................................34 3-5 Potential Transcriptional Markers of aGVHD Progression....................................................36


7 LIST OF FIGURES Figure page 1-1 Ferrara Model of aGVHD Pathophysiology......................................................................14 1-2 G-CSF as an I mmunological Modulator............................................................................16 3-1 Genome-Wide Transcriptional Response to G-CSF..........................................................28 3-2 Gene Expression Profile of aGVHD..................................................................................33


8 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science GENE EXPRESSION PROFILING IN HEMATOPOIETIC STEM CELL TRANSPLANTATION By Matthew Peter Buzzeo August 2007 Chair: Vijay Reddy Major: Medical Sciences Allogeneic peripheral blood stem cell transplant ation (PBSCT) is the treatment of choice for high-risk hematological malignancies. In PBSCT, genetically compatible donors are administered granulocyte-colony stimulating fact or (G-CSF) to mobilize stem cells from their bone marrow. Despite a significantly higher number of T cells contained in the G-CSFmobilized stem cell allograft as compared to a bone marrow allograft, the risk of acute-graftversus-host disease (aGVHD) is not increased. This effect is attrib uted to the ability of G-CSF to exert immunosuppressive effects on the allograf t predominantly through increased tolerogenic dendritic cell (DC) counts, which polarize donor nave T cells toward a T helper-type 2 (Th2) phenotype. Specifically, Th2 cells and other donor-derived tole rogenic cells (e.g., monocytes, regulatory T cells) decrease a GVHD risk by increasing levels of immunosuppressive cytokines such as interleukin-4 (IL-4) and IL-10, thereby antagonizing post-transplant alloreactive events. Additional roles for allograft composition in in fluencing the development of aGVHD are not well described, however, and few definitive mark ers exist for increased aGVHD risk or aGVHD progression to advanced and potentially fatal stages. Emerging high-throughput technologies such as DNA microarrays can be used as pow erful tools for addre ssing these problems.


9 Our study used microarray technology to report the gene expression profile of the G-CSFmobilized allograft and to disc over novel transcriptional marker s that may predict development of aGVHD. We enrolled 5 donors and 8 recipients (4 control, 4 aGVHD) for a University of Florida IRB-approved study and collected periph eral blood after patients had given informed consent. For donors, blood was collected before and after G-CSF administration; for recipients, blood was collected at the time of engraftment a total of 18 samples. For each sample, RNA was purified from whole blood leukocytes and used to generate cRNA which was subsequently hybridized onto Affymetrix GeneChip Human Genome U133 Plus 2.0 microarrays. Significant changes in gene expression were assessed by cont rolling the false discovery rate (FDR) between 5% and 10%. Additionally, relevant immune cellsincluding monoc ytes, granulocytes, lymphocytes and dendritic cellswere enumer ated via complete blood count (CBC) with differential and flow cytometry. In donors, although G-CSF drastically increased allograft immune cell content, adaptive immune response genes were significantly down -regulated, including anti gen presentation, T cell activation and cytolytic effector response pathways. In recipients, several novel, putative transcriptional markers of aGVHD were discovere d, such as the pro-inflammatory cytokine IL27 and the autoimmune regulator gene AIRE. In conclusion, this study offers insight into the applications of microarray technology toward understanding and predic ting outcomes of hematopoietic stem cell transplantation.


10 CHAPTER 1 GENERAL BACKGROUND 1.1 Hematopoietic Stem Cell Transplantation The concept of exploiting hematopoietic cells of the bone marrow to reconstitute deficient immune systems has been in place for approximately 50 years (Thomas et al. 1957). The principle governing early studies in experimental bone marrow tr ansplantation (BMT) was that patients afflicted with hematological maligna ncies could be admini stered high doses of chemotherapy and/or radiation and subsequently be infused with BM harvested from the self (autologous) or from a genetically compatible do nor (allogeneic). In years that followed, allogeneic BMT became an increasingly accepte d treatment for hematological cancers and inherited immune deficiencies. In addition to the anti-malignancy effects afforded by donorderived immune cells (which we re beneficial in eliminating quiescent cancers cells that had escaped chemoradiation treatment) donor-derived hematopoietic stem cells (HSC) could engraft, differentiate, and re-establish th e hosts immunological repertoire. Eventually it was discovered that certain growth factors c ould efficiently mobilize stem cells from the marrow into the peripheral circulation (Gianni et al. 1989). This discovery spurred a revolution in the field of stem cell transplantation. Gradually, donor m obilization with granulocyte-colony stimulating factor (G-CSF) followed by peripheral blood st em cell transplantation (PBSCT) began to supersede BMT as the standard of care, owing la rgely to easier access to donor stem cells, in addition to more efficient engraftment a nd immune reconstitution kinetics (Storek et al. 2001). Although originally id entified as a growth f actor for neutrophils a nd used clinically to boost granulocyte counts during myelosuppression, G-CSF has a secondary effect of increasing hematopoietic progenitor counts. This effect is attributed to the action of neutrophil proteases flooding the marrow and weakening interactions between stromal cells and stem cells, thus


11 promoting stem cell departure from the ma rrow microenvironment a nd into the greater circulation (Levesque et al. 2002). Large numbers of pluripotent HSCs contained within G-CSFmobilized stem cell products f acilitate a rebirth of hematopoi esis and immune function in previously immune-compromised hosts. Years af ter its discovery, the use of G-CSF-mobilized PBSCT remains the treatment of choice to combat high-risk hematological malignancies and in some instances is a cancer cure. However, su ch tremendous benefits are sometimes offset by severely destructive consequences, such as susc eptibility to opportunist ic infections, cancer relapse and graft-versus-host disease (GVHD). Thus, continued research from a clinical, translational and basic sciences perspective will be needed to fully understand mechanisms behind these complications and to develop strategi es for preventing, treati ng or eliminating them. 1.1.1 Donor Selection The first step in the process of PBSCT is id entification of a suitable donor. In some cases, the nature of the malignancy may call for an auto logous transplant, especially for patients with non-Hodgkins lymphoma (Philip et al. 1995), multiple myeloma (Attal et al. 1996) or for certain autoimmune diseases. Much of the fo cus here, however, shall concern the use of allogeneic PBSCT for the treatment of hematological malignancies, such as acute leukemia. Allogeneic donor selection take s place principally in consid eration of the human leukocyte antigen (HLA) genotype. Ideally, do nor and recipient are at to be matched at all six HLA loci: two each for HLA-A, HLA-B and HLA-DR 1. The HLA genes are extraordinarily polymorphic, thus the probability of finding a co mpatible donor is rare, unless donor and patient are related as siblings. If so, there is a 25% chance for a complete HLA match as the HLA genes are inherited as haplotypes. Donor-recipient related PBSCT carries minimal risk of graft rejection because there will be co mpatibility between donor and reci pient cells in terms of their major histocompatibility (MHC) antigens.


12 If an HLA-matched sibling donor is not availa ble, the other alternat ive is to locate a matched unrelated donor (MUD) through a nationa l volunteer donor regi stry. Although MUD donors are compatible based on th eir HLA genotype, disparities in other polymorphic genes such as minor histocompatibility antigens may increase the risk of allograft intolerance. 1.1.2 Stem Cell Mobilization with G-CSF After a suitable HLA-matched donor is identif ied, recombinant human G-CSF (filgrastim) is administered over a course of 4 to 5 days. The absolute number of stem cells mobilized is quantified based on CD34 expression and is dos e dependent. Generally, a dose of 10g/kg/day is optimal and yields as much as 0.99% CD34+, or 61 CD34+ cells per microliter of peripheral blood (Lane et al 1995). The resultant stem cell produc t is harvested from the peripheral circulation in a manner that is decidedly less invasive than a conventional bone marrow aspiration. Although G-CSF has mini mal toxicity, bone pain is the most common side effect and there is a risk of otherwise h ealthy donors experiencing acute in flammatory reactions that can damage pulmonary (Arimura et al 2005) and vascular (Lindemann & Rumberger, 1993) tissues. Patients with preexisting autoimmune disorders such as rheumatoid arthritis may experience temporary disease aggravation (Stricker & Goldbe rg, 1996). The effects of G-CSF are transient and usually within 2 weeks pos t-administration, white blood counts return to normal. 1.1.3 Recipient Conditioning and Transplantation Recipient pre-transplant conditi oning is crucial for maximizi ng the benefits of PBSCT. The first benefit is a reduction in tumor mass or leukemia blast generation by the action of high dose cytotoxic chemotherapy and/or radiati on. Secondly, ablation of the recipients immunological repertoire leads to generalized immune suppressi on, which prevents acute graft rejection, thus eliminating the need for prophylactic administrati on of immunosuppressants as in solid organ transplantation. Finally, the condi tioning regimen clears the marrow space allowing


13 for homing and engraftment of donor-derived HSCs and other blood elements. After PBSCT infusion, there is a variable time to engraftmenta term used to describe the period after transplantation when absolute neutrophil c ount (ANC) exceeds 500 cells/L. The period in between the conditioning regimen and engraftmen t is a precarious time in which generalized immune suppression and ne utropenia renders the host vulnerable to opportunistic bacterial, viral and fungal infections. 1.2 Graft-versus-Host Disease The major complication associated with all ogeneic PBSCT is graf t-versus-host disease (GVHD) which is subdivided in to two categories depending on the timing of onset and the pathological nature. Acute GVHD (aGVHD) occurs within 100 days post-transplant and is clinically manifest predominantly in the skin, liver and gastrointestinal (GI) tract; whereas chronic GVHD (cGVHD) develops gr eater than 100 days post-tran splant, and may assume the guise of systemic autoimmune diseases such as scleroderma or Sjorgens syndrome. The pathophysiological mechanisms distinguishing aGVHD from cGVHD are qu ite clear (Iwasaki, 2004). aGVHD is due to donor T cells becoming activated by host dendri tic cells (DC) in response to pro-inflammatory cytokines released during the conditioning regimen. The result is alloreactive dono r-derived CD4+ T helper-type 1 (Th1) cells maturing and subsequently imposing cytolytic effector responses on host tiss ues, resulting in multi-organ damage. Poor clinical prognoses are imminent pa rticularly in cases where tissue damage is severe and disease progression is recalcitrant to immunosuppressive therapies. In cGVHD, the initial conditioning regimen causes thymic damage which allows autoreactive donor-derived T cells to forgo negative selection in the thymus. This in turn leads to CD4+ T helper-type 2 (T h2) cells assisting autoreactive B cells in the produc tion of antibodies against host (self) antigens. Thus, the distinguishing features of acu te vs. chronic GVHD are whether host immune dysregulation is


14 driven by Th1 or Th2 responses. The remainde r of the discussion will, however, focus on aGVHD. The classical model of aGVHD pathophysiol ogy (Figure 1-1) is divided into three phases: conditioning regimen, donor T cell activati on and cytolytic effector responses (Reddy & Ferrara, 2003). Figure 1-1. Ferrara Model of aGVHD pathophysio logy. Host myeloablative conditioning (1) initiates the release of pro-inflammatory cy tokines from the host tissues that activate resident DC. After transplantation, activat ed host DC present a lloantigens to nave donor T cells (2), which differentiate into T h1 cells and then mature into cytotoxic-T lympocytes (CTL). Damage to the host skin, liver and gastrointestinal tract is driven by cytolytic effector responses (3) mediated through Fas/FasL and perforin/granzyme pathways of apoptosis. Phase 1 of aGVHD begins even before transpla ntation, when the recipi ent is treated with myeloablative chemotherapy and/or total body ir radiation. The conditioning regimen benefits the host by preventing allograft intolerance and promoting leukemicide; however, conditioning inevitably results in some degree of tissue damage. Epithelial damage to the GI tract compromises the innate barrier between the microbe-rich gut and the surrounding tissues and


15 circulation. Ulceration of the GI mucosa cause s leakage of microbial constituents such as lipopolysaccharide (LPS) and CpG. These micr obe-associated molecules act through toll-like receptors (TLR) on host immune cells to induce th e synthesis and release of pro-inflammatory cytokines such as tumor necrosis factor(TNF), interleukin-1 (IL-1), IL-6 and IL-12 from monocytes and macrophages. In addition to direc tly promoting inflammation of the tissues, this cytokine storm also leads to activation of hos t DC. DC activation is marked primarily by an up-regulation of MHC, co-stimul atory molecules and adhesion proteins that facilitate interactions with nave T cells in the secondary lym phatic tissues. Phase 2 is characterized by the presentation of host antigens to donor T cells. At this point, T cells have been activated to secrete high leve ls of IL-2 which in turn promotes increased expression of the IL-2 receptor. IL-2 is a pot ent autocrine growth factor for T cells and is required for maturation into effector CD4+ or CD8+ T cells. In addition to IL-2, other proinflammatory cytokines such as interferon(IFN) peak during this period and act to amplify alloreactive T cell proliferation. Phase 3 is generally the period when aGVHD be comes clinically manifest. Damage to host tissues is mediated by allor eactive natural killer (NK), CD4+ and CD8+ cytotoxic T lymphocytes (CTLs) (Schmaltz et al. 2001; van den Brink & Burakoff, 2002). CD8+ and NK cells promote host cell death by releasing perforin granules which causes pore formation in host cell membranes. This is then followed by releas e of proteolytic granzymes into the pores which cleave and activate host cell caspases, lead ing to apoptosis. Alternatively, CD4+ effector cells express high levels of Fas ligand (FasL) th at engage the Fas recep tor (Fas) on host cells, resulting in activation of the re ceptor-mediated apoptotic pathway. These mechanisms are also important in negatively regulating aGVHD, as the perforin/granzymeand Fas/FasL-mediated


16 pathways of apoptosis are implicat ed in fratricide, or the controlled elimin ation of alloreactive CD8+ cells (Maeda et al. 2005). 1.3 Immunomodulatory Properties of G-CSF In addition to being a potent stimulator of stem cell mobilization, a growing body of evidence suggests that G-CSF ha s potent immunomodulatory effect s (Frankze, 2006). The term stem cell product is somewhat misleading as G-CSF-mobilized allografts contain many other cells in addition to HSCs, incl uding granulocytes, T cells, monoc ytes and DCs. G-CSF acts on these various cells and influen ces their proliferation, differen tiation and cytokine production (Figure 1-2). Figure 1-2. G-CSF as an immunol ogical modulator. G-CSF is used clinically to boost white blood cell counts and drastically increases granulocyte (neutrophil) counts in myelosuppressed patients. In HSCT, G-CSF is used to mobilize donor stem cells but also has potent immunomodulatory effect s on graft composition. G-CSF increases plasmacytoid DC (DC2) counts which then polarize nave T cells toward Th2. Additionally, G-CSF increases regulatory T cell counts (Tregs) and monocytes. Tregs, Th2 cells and monocytes secrete hi gh levels of immunos uppressive cytokines such as IL-4, IL-10 and TGF. Despite a 1-log greater number of T cells contained in the G-CSF-mobilized stem cell product as compared to a BM harv est, the incidence of aGVHD is not increased; although risk of


17 cGVHD may be higher (Korbling et al. 1995). This effect is generally ascribed to a polarization of donor T cells towards Th2 responses that suppr ess post-transplant allo reactive events through production of immunosuppressive cy tokines such as interleukin-4 (IL-4), IL-10 and transforming growth factor-beta (TGF). Mechanistically, this shift in immune res ponses can be explained by: (a) T cells respond to G-CSF by up-regulation of GATA-3 which in duces polarization of T cells toward a Th2 phenotype (Zhu et al. 2006); (b) G-CSF induces the mobili zation of tolerogenic DC subtypes, namely plasmacytoid DC (DC2) which resemble immature DC and polarize T cells towards Th2 (Arpinati et al. 2000); (c) elevated number s of monocytes contained in the G-CSF-mobilized allograft suppress T cell activation through the produc tion of IL-10 (Fraser et al. 2006); (d) GCSF promotes an increase in CD4+CD25+ T regulatory (Treg) cells that promote post-transplant tolerance (Rutella et al. 2002). The beneficial effects of G-CSF (both in terms of immune reconstitution kinetics and seeming protection from aGVHD) are mediated by effects on the donor and not on the recipient (Reddy et al. 2000). This is further supported by the f act that G-CSF is not required posttransplant in order for recipien ts to engraft. In HLA hapl otype-mismatched PBSCT, posttransplant administration of GCSF can engender detrimental c onsequences including delayed immune reconstitution and increased susceptibilit y to viral and fungal infectionsthe latter complication associated with reduced Th1 cellu lar responses and overall immune dysregulation (Volpi et al. 2001). Thus, it appears that the role of G-CSF in affec ting transplant outcomes is through its influence on graft composition. Given the ability of G-CSF to promote Th2 respon ses, there is interest in utilizing it as an adjuvant treatment for Th1/T cell-mediated autoi mmune diseases such as Type I diabetes and


18 Crohns disease (Kared et al. 2005; Korzenik & Dieckgraefe, 2005). However, G-CSF may exacerbate autoimmune diseases that are al ready driven by Th2 responses and autoantibody production, as in systemic lupus erythema tosus and rheumatoid arthritis (Lawlor et al. 2004). Indeed, the clinical similariti es between cGVHD and certain Th2 autoimmune diseases suggests that G-CSF may increase the risk of cGVHD th rough the polarization of donor-derived T cells toward Th2 (Banovic et al. 2005). 1.4 Microarray Analysis and Gene Ex pression Profiling in Medicine DNA microarray technology has th e potential to offer revolut ionary insights into the complex nature of biological processes and mech anisms of human disease. Instead of studying one particular gene and how it relates to disease, a DNA mi croarray can allow access to the entire human genome in a single experiment; th e only requisite material being sub-microgram amounts of RNA. The information garnered from such experiments can have useful clinical applications including the molecular classification and staging of disease a nd in the discovery of prognostic indicators for disease progression (Cobb et al. 2005). The chemistry behind DNA microarray tec hnology is extraordinarily complex, yet experiments are undertaken with relative ease an d can be completed in a two-day period. A state-of-the-art microarray chip can measur e the expression of over 34,000 well-characterized human genes, which represents the entire transc riptionally active genome. When comparisons are made between multiply sampled groups, volumi nous amounts of data are generated. Indeed, the amount of data gained from a microarray analysis is both a bl essing and a curse, and the true challenge lies in being able to ex tract useful and biologically releva nt information from that data. This is achieved through employi ng rigorous statistical analyses. Originally, reporting the fold change (FC) in expression for each gene was considered sufficient, especially in experimental situati ons where the control and experimental tissue


19 samples had large differences in biological activ ity, such as in a normal vs. disease state or before vs. after drug treatment. This method is generally effective in pointing out genes that exhibit large changes in expres sion; however, less strongly in duced or repressed genes are filtered out of the data set. In response to th is problem, an algorithm was developed to organize genes into clusters based on th eir level of expression (Eisen et al. 1998). In this way, interrelationships between genes and their tr anscriptional regulati on can be elucidated. Although a cluster analysis is a useful way to extract meaningful biological information from microarray data, a major trade-off is an i gnorance of statistical sign ificance. Microarrays are based on multiple hypothesis testing, which in its elf can lead to increases in Type I errors and the generation of false-positives if conventional thresholds of significance, such as P<0.05, are used (Allison et al. 2006). To overcome this problem of false-positive discovery, an algorithm was devised to control the false discovery rate (FDR), usually at a maximum of 10% (Reiner et al. 2003). Thus, if 1000 genes are reported as significant based on a 10% FDR, 100 genes would, in theory, be erroneously reported. Today, utilizing the FDR algorithm is the most widely accepted method of assigning statisti cal significance to gene s uncovered by microarray analysis. 1.4 Current Study Aims G-CSF is commonly used to mobilize blood stem cells and as a growth factor to promote granulocyte counts during myelos uppression. We hypothesized that its effects on the immune system were counterproductive due to overa ll immune suppression. Our study employed a comprehensive flow cytometric and microarray an alysis of peripheral blood leukocytes from GCSF-mobilized donors in order to describe the immunomodulatory effects of G-CSF in a broad context, based on changes in immune cell counts and genome-wide gene expression patterns.


20 Secondly, we analyzed the gene expression profiles of aGVHD and aGVHD-free patients to generate a molecular signature of early-mid phase aGVHD. The rationale for this study was that few definitive markers exist for predic ting aGVHD and that utilizing microarray technology would allow for the discovery of several putative transcriptional markers of disease progression. Thus, our hypothesis was that si gnificant differences in leukoc yte gene expression patterns between aGVHD and aGVHD-free patients would be observed at the time of engraftment, yet before a diagnosis of clinical aGVHD.


21 CHAPTER 2 MATERIALS AND METHODS 2.1 Enrollment of Subjects This study was approved by the University of Florida Institutional Re view Board. From November 2005 to August 2006, peripheral blood sa mples were collected from donors (n=5) and recipients (n=8) of allogeneic PBSCT after they had given informed consent. Donor blood was collected at baseline and again afte r administration of rHu-G-CSF (Neupogen, Amgen, Thousand Oaks, CA) for five days at 10g/kg/day, which coincide d with the time of stem cell product harvesting. Transplant recipients received myeloabla tive conditioning and PBSCT and provided peripheral bloo d at engraftment (ANC 500). The 10 donor samples and 8 recipient samples were used for microarray analysis, for a total of 18 microarray chips. The 8 recipients were grouped according on their GVHD status (Table 2-1). Those recipien ts that developed GVHD within 100 days post-transplant were assigned to the aGVHD group (n=4). In all ca ses, aGVHD was diagnosed within 5-10 days post-engraftment. GVHD-free patients (n=4 ) were assigned to the control group. 2.2 Sample Acquisition and Processing Peripheral blood for microarray analysis was processed immediately after collection in order to reduce ex-vivo effects of altering the gene expres sion profile. Blood was centrifuged at 1,200 rpm for 10 minutes and the plasma was rem oved and stored at -20C. The remaining erythrocyte and buffy coat fract ions were added to 40 mL of ammonium chloride Buffer EL (Qiagen, Valencia, CA) for 15 min at 4C to ly se erythrocytes. The remaining leukocyte fraction was concentrated by cen trifugation, and the cells were washed in 15 mL of Buffer EL again to lyse residual erythrocytes.


22 Leukocyte RNA was purified using the RNeasy mi ni kits (Qiagen, Valencia, CA). RNA concentration was determined on a NanoDr op ND-1000 Spectrophotometer and quality was assessed by capillary electrophoresis on the Ag ilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). Biotinylated cRNA was generate d from 2.5 g of total cellular RNA. For each sample, 15g of fragmented cRNA was hybridized onto Affymetrix GeneChip Human Genome U133 Plus 2.0 microarrays (Affyme trix, Santa Clara, CA) for 16 hours at 45C. Arrays were washed and stained according to standard A ffymetrix protocol, and then scanned on an Affymetrix GeneChip Scanner 3000. Image analys is was performed with Affymetrix GeneChip Operating Software v1.2. 2.3 Complete Blood Counts and Flow Cytometry Total white blood cell counts, absolute ne utrophil counts and monocyte counts were generated using a COULTER AcT diff (Beckman-Coulter, Fullerton, CA). Other cell phenotypes (HSC, T cells, B cells, NK cells, DC) were quantified using a BD FACSCalibur 4color flow cytometer (Becton Dickinson, Mountai n View, CA). The following gives the cell types measured and their corresponding surface markers: HSC (CD34+); T cell (CD3+/CD4+/CD8+; B cell (CD20+); NK cell (CD56+CD16+); Treg (CD25+CD4+). Dendritic cell counts were analyzed as de scribed previously (Reddy et al. 2004). In brief, DC were negative for lineage markers (CD3, CD14, CD16, CD19, CD20, CD56) and positive for HLA-DR. DC1 were CD11c+, and DC2 were CD123+. Absolute number of ci rculating cells (cells/mm3) was calculated by multiplying the percentage of cells with the total number of white blood cells per microliter.


23 2.4 Microarray Anlaysis The data discussed in this publication have been deposited in NCBIs Gene Expression Omnibus (GEO, o/) and are accessible through GEO Series accession numbers GSE7400 (Donor G-CSF st udy) and GSE7510 (Recipient aGVHD study). We analyzed 54,613 probe sets on each microarray. For donor microarrays, the probe signal intensities were first log transformed and then an analysis of covariance (ANCOVA) adjusting for difference of each donor (e.g, age, gender) was carried out to determine if the expression level significantly changed after mobilization with G-CSF. Significant changes in gene expression were assessed by controlling the FDR at 5%. For recipient microarrays, the probe signal inte nsities were first log transformed and then an ANCOVA adjusting for type of transplant (e.g., MUD, related) was carried out to assess if the expression level significantly differed among th e aGVHD and control groups. Each p-value was calculated using a pooled error estimate to gain more power for hypothesis testing. Differential gene expression among two groups was assessed by c ontrolling the false disc overy rate (FDR) at 10%. The NetAffx Analysis Center available thr ough the Affymetrix website was used for probe set annotation assistance. In addition, information gathered from independent literature searches for each significan t gene was used for assigni ng biological and immunological significance. Changes in gene e xpression are reported as the log2 fold change (FC) of the signal intensity. Unsupervised hierarch ical cluster analyses were gene rated using dChip 2006 software.


24 CHAPTER 3 RESULTS 3.1 Effects of G-CSF on Allograft Composition Mean blood counts and a panel of cell phenot ypes that were enumerated via flow cytometry before (pre-G-CSF) and after (postG-CSF) administration of G-CSF are reported (Table 3-1). A threshold of P<0.05 was used for assigning statistical significance. G-CSF administration for five days induced consistent ly significant increases in total white blood cell count (WBC) (7.1 vs. 41.4 x 103/mm3, P=0.011), absolute neutrophil count (ANC) (4.4 vs. 36 x 103/mm3, P=0.013) and monocytes (262 vs. 1,672 x 103/mm3, P=0.008). CD34+ counts were not analyzed in pre-G-CSF donors; however, pos t-G-CSF blood stem cell products contained appreciable amounts of CD34+ stem cells (mean 216 cells/mm3). In measuring cells of the lymphoid lineage, T (CD3+), B (CD20+) and NK (CD3-/CD56+/CD16+) cells, all significantly increased after G-CSF (P=0.004; P=0.031; P=0.039) with T cells increasing 7.5-fold (1.6 vs. 11.8 x 103/mm3). CD4+, CD8+ and Treg counts were significan tly increased (P=0.004; P=0.013; 0.019), with CD4+ cells predominating. Finally, DC1 a nd DC2 were significantly increased (P=0.014; P=0.015). When comparing the rela tive increases of DC2 vs. DC1, G-CSF mobilization preferentially incr eased DC2 counts, as the rati o of DC2:DC1 was significantly increased (0.46 vs. 1.69, P=0.01). 3.2 Relevant Transcriptional Responses to G-CSF Of 54,675 probe sets, 4,501 showed significant ch anges in expression between the paired pre-G-CSF and post-G-G-CSF samples, based on a 5% FDR (Figure 3-1). Annotations derived from the NetAffx database were used to categorize 459 probe sets (corresponding to 335 unique genes) as relevant to stem cell mobilization and the immune response.


25 An independent literature search for each ge ne was carried out in order to assess its specific biological contribution to stem cell mobilization or imm une function. A representative table is reported (Table 3-2). The complete data set is available online through GEO accession number GSE7400. Based on a 5% FDR microarray analysis of gene expression, we observed that stem cell mobilization and function was highly favored af ter G-CSF, as many genes involved in HSC mobilization (PLAUR) and function mediated by Wn t signaling (LEF1, TCF-7) were significant. Similarly, many genes specific for neutrophil ac tivation, adhesion and transendothelial migration were induced, including proteas es, (MMP-9), cathepsins (C TSC), complement receptors (CR5A1) and adhesion molecules (CD47). In a ddition to pro-granulopoietic gene expression activity, several key components of the inflammato ry immune response were activated, including signaling via the IL-1 receptor to produce TNF. Despite this, the cytokine and chemokine profile appears to indicate overa ll immune suppression through up-regulat ion of Th2 cytokine receptors (IL-10R IL-4R ), down-regulation of Th1 cytokine receptors (IL-2R IL-12 ) and downregulation of pro-inflammatory chemokines (CCL1, CCL2, CCL5) and chemokine receptors (CCR5, CCR7). In summary, there appears to be an overwhelming antagonism of adaptive immune responses after G-CSF, as evident by strong repression of antigen presentation, T cell activation and cytoly tic effector response gene expression. 3.3 Gene Expression Profile of aGVHD Of the 54,675 probe sets measured, 1,658 show ed significant changes in expression between the aGVHD and control groups, based on a 10% FDR (Figure 3-2). Of these, 65 probe sets corresponding to 57 different genes were de signated as relevant to the immune response (Table 3-4). The complete data set is av ailable online through GEO accession number GSE7510.


26 Down-regulated genes in early aGVHD include d activation-induced cytidine deaminase (AICD) and terminal deoxynucleot idyltransferase (TdT), which ar e important for generation of lymphocyte receptor diversity. Certain pro-infl ammatory genes such as toll-interleukin-1 domain containing adaptor protein (TIRAP), interferon16 (IFNA16), advanced glycosylation end product-specific receptor (AGER) and compon ents of the IL-1 recep tor/NF-kappaB pathway (IRAK1BP1, NFKB1) were suppressed. Genes e xpressed primarily by antigen presenting cells that are important in stimulating T cell responses including the cysteiny l leukotriene receptor 2 (CYSLTR2), MHC I (HLA-C) and inducible T-ce ll co-stimulator ligand (ICOSLG) were downregulated. CD244an NK cell activating receptor and marker for cytotoxic CD8+ T cells was down-regulated as well, in addition to chem okine-ligand 1 (CCL1) and the novel cytokine interleukin-27 (IL27). Significantly up-regulated genes include many pro-inflammatory cytokine and cytokine receptor genes, such as IL-2, IL-2R IL-22R TNF and lymphotoxin beta (LTB). Clustering adjacent to TNF and LTB is ARTS-1, a re gulator of TNF receptor shedding. The CCAAT/enhancer binding protein (CEBPA) was induced, further indicating activation of proinflammatory cytokine synthesis. Up-regulated genes directly imp licated in T cell activation and cell-mediated toxicity include mast cell tr yptase (TPSAB1), program med cell death ligands (PDCD1LG2), complement component 3a recepto r 1 (C3AR1) and autoimmune regulator gene (AIRE). Strongly induced or repressed ge nes were designated as genes that may potentially serve as transcriptional markers for aGVHD (Table 3-5). Some of the most strongly up-regulated genes in our aGVHD patients included steroid 5 alpha reductase 2like (SRD5A2L), mirror-image polydactyly 1 (MIPOL1), members of the cadherin family of cell adhesion proteins (PCDHB5,


27 PCDH9), zinc and ring finger 3 (ZNRF3) and th e DNA-damage responsive ataxia telangiectasia mutated (ATM) gene. However, anothe r cadherin-family gene, protocadherin subfamily B7 (PCDHGB7) was strongly repressed. The only definitively immune-related gene to show up on this analysis was activati on-induced cytidine deaminase (AICD), which was suppressed.


28 -3.0 -2.0 -1.0 0 1.0 2.0 3.0Table 3-1. Donor Blood Counts and Flow Cyto metry: Pre vs. Post G-CSF Mobilization Cell Type Pre-G-CSF* Pre-G-CSF* FC P WBC ( x 103) 7.1 2.4 41.4 22.4 5.80.011 ANC ( x 103) 4.4 1.8 36 22 8.20.013 Monocytes 262 67 1,672 771 6.40.008 CD34+ (HSC) 216 195 CD3+ (T cells) 1,558 611 11,752 4,860 7.50.004 CD4+ 256 147 1,811 1,265 7.10.005 CD8+ 338 815 815 764 2.40.013 CD3CD56+CD16 (NK cells) 971 240 7,175 3,149 7.40.039 CD20+ (B cells) 529 380 3,883 2,538 7.30.031 CD25+CD4+ (Treg) 64 47 920 658 14.40.019 LinHLA-DR+ CD123+ (DC1) 15 3 71 37 4.70.014 LinHLA-DR+ CD11c+ (DC2) 6 5 126 84 21.00.015 DC2:DC1 0.46 0.31 1.69 0.79 3.70.010 Reported as the mean standard deviation. Fold change (FC) generated by computing the ratio of post-G-CSF count to pre-G-CSF count. A_Pre-G C_Pre-G B_Pre-G D_Pre-G E_Pre-G B_Post-G E_Post-G D_Post-G A_Post-G C_Post-G Figure 3-1. Genome-wide transcri ptional response to G-CSF. A microarray analysis was performed on Prevs. Post-G-CSF-mobili zed peripheral blood leukocytes for five allogeneic PBSCT donors (A-E). Shown is the unsupervised hierarchical cluster analysis of genes and samples. After G-CSF-mobilization, 4,501 probe sets were significant based on a 5% FDR. 1,569 were increased in expression; whereas 2,932 were decreased.


29 Table 3-2. Immune-Relat ed Transcriptional Response to G-CSF Mobilization Gene Symbol Gene Title FC* Function CD177 CD177 molecule 630.41 neutrophil marker MMP8 matrix metallopeptidase 8 (neutrophil collagenase) 145.65 proteolysis ELA2 elastase 2, neutrophil 60.26 proteolysis DEFA4 defensin, alpha 4, corticostatin 50.00 defense response MPO myeloperoxidase 22.39 defense response CTSG cathepsin G 15.36 proteolysis CD24 CD24 molecule 14.57 humoral immune response MMP9 matrix metallopeptidase 9 9.53 proteolysis IRAK3 interleukin-1 receptor-associate d kinase 3 7.24 cy tokine signal transduction IL18R1 interleukin 18 recep tor 1 6.26 proinflammatory cytokine CST7 cystatin F (leukocystatin) 5.07 immune response CXCL16 chemokine (C-X-C motif) ligand 16 4.74 chemotaxis IL10RB Interleukin 10 receptor, beta 4.18 Th2 cell development LTB4R Leukotriene B4 r eceptor 4.13 inflammatory response TLR5 toll-like receptor 5 3.72 inflammatory response CD14 CD14 molecule 2.91 inflammatory response MMP25 matrix metallopeptidase 25 2.74 inflammatory response CKLF chemokine-like factor 2.68 chemotaxis IL4R interleukin 4 receptor 2.57 Th2 cell development NCF4 neutrophil cytosolic factor 4, 40kDa 2.53 immune response CSF3R colony stimulating factor 3 receptor 2.44 signal transduction CTSC cathepsin C 2.44 proteolysis TNF tumor necrosis factor 2.21 immune response NFKBIA nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha 2.19 regulation of NFkappaB import into nucleus TOLLIP toll interacting protein 2.16 inflammatory response IL17RA interleukin 17 recepto r A 2.08 cell surface receptor linked signal transduction IRAK2 interleukin-1 receptor-associa ted kinase 2 2.02 inflammatory response MYD88 myeloid differentiation primary response gene (88) 1.93 inflammatory response NFAM1 NFAT activating protein with ITAM motif 1 1.91 inflammatory response TLR4 toll-like receptor 4 1.89 inflammatory response JAK2 Janus kinase 2 (a protein tyrosine kinase) 1.89 cytokine signaling FKBP1A FK506 binding protein 1A, 12kDa 1.89 inflammatory response


30 Gene Symbol Gene Title FC Function SYK Spleen tyrosine kinase 1.66 lymphocyte activation IRAK4 interleukin-1 receptor-associate d kinase 4 1.59 cy tokine signaling STAT5B signal transducer and activator of transcription 5B 1.44 cytokine signaling IRAK1BP1 Interleukin-1 recepto r-associated kinase 1 binding protein 1 0.70 I-kappaB kinase/NFkappaB cascade LTB Lymphotoxin beta (TNF superfamily, member 3) 0.62 innate immune response IL27RA interleukin 27 receptor, alpha 0.56 Th1 cell development ILF2 interleukin enhancer binding factor 2, 45kD 0.54 regulation of transcription STAT1 signal transducer and activator of transcription 1, 91kDa 0.53 cytokine signaling DNTT deoxynucleotidyltransferase, terminal 0.50 DNA replication CD86 CD86 molecule 0.50 T cell activation IGKC /// IGKV1-5 immunoglobulin kappa constant /// immunoglobulin kappa variable 1-5 0.48 antigen presentation (MHC I) CD74 CD74 molecule, major histocompatibility complex, class II invariant chain 0.48 antigen presentation TAP2 transporter 2, ATP-binding cassette, sub-family B (MDR/TAP) 0.48 antigen presentation (MHC I) HDAC9 histone deacetylase 9 0.46 B cell activation HLA-DMB major histocompatibility complex, class II, DM beta 0.46 antigen presentation ARTS-1 type 1 tumor necrosis factor receptor shedding aminopeptidase regulator 0.45 antigen presentation (MHC I) IL2RA interleukin 2 receptor, alpha 0.45 regulation of T cell proliferation PSME2 proteasome (prosome, m acropain) activator subunit 2 (PA28 beta) 0.43 immune response SLA2 Src-like-adaptor 2 0.39 T cell activation CD79B CD79b molecule, immunoglobulin -associated beta 0.39 B cell receptor signaling CCR5 chemokine (C-C motif) receptor 5 0.39 chemotaxis HLA-DRA major histocompatibility complex, class II, DR alpha 0.38 antigen presentation HLA-DRB1 /// HLA-DRB3 major histocompatibility complex, class II, DR beta 1//DR beta 3 0.38 antigen presentation NFATC2IP nuclear factor of ac tivated T-cells, cytoplasmic, calcineurin-dependent 2 interacting protein 0.38 protein modification HLA-DMA major histocompatibility complex, class II, DM alpha 0.37 antigen presentation SLAMF1 signaling lymphocytic activation molecule family member 1 0.36 lymphocyte activation CD7 CD7 molecule 0.36 T cell activation SOCS2 suppressor of cytokine signaling 2 0.35 negative regulation of cytokine signaling HLA-DPA1 major histocompatibility complex, class II, DP alpha 1 0.34 antigen presentation CD8B CD8b molecule 0.34 T cell activation IL12RB1 interleukin 12 receptor, beta 1 0.32 Th1 cell development CD22 /// MAG CD22 molecule /// myelin associated glycoprotein 0.32 B cell marker GATA3 GATA binding protein 3 0.32 Th2 development CD160 CD160 molecule 0.31 cytolytic effector cell marker HLA-DPB1 major histocompatibility complex, class II, DP beta 1 0.30 antigen presentation Table3 2.Continued


31 Gene Symbol Gene Title FC Function HLA-DOB major histocompatibility complex, class II, DO beta 0.30 antigen presentation PRF1 perforin 1 (pore forming protein) /// perforin 1 (pore forming protein) 0.28 cytolytic effector response LCK lymphocyte-specific protein tyrosine kinase 0.28 T cell activation FYN FYN oncogene related to SRC, FGR, YES 0.27 T cell activation HLA-DQB1 major histocompatibility complex, class II, DQ beta 1 0.27 antigen presentation POU2AF1 POU domain, class 2, associating factor 1 0.27 regulation of transcription XCL2 chemokine (C motif) ligand 2 0.26 chemotaxis GZMB granzyme B (granzyme 2, cytotoxic T-lymphocyteassociated serine esterase 1) 0.25 cytolytic effector response KLRG1 killer cell lectin-like receptor subf amily G, member 1 0.25 cellular defense response FASLG Fas ligand (TNF superfamily, member 6) 0.25 apoptosis GZMH granzyme H (cathepsin G-like 2, protein h-CCPX) 0.25 cytolytic effector response KLRD1 killer cell lectin-like r eceptor subfamily D, member 1 0.25 cell surface receptor linked signal transduction CCR7 chemokine (C-C motif) receptor 7 0.24 chemotaxis CRTAM cytotoxic and regulatory T cell molecule 0.24 cytolytic effector response STAT4 signal transducer and activator of transcription 4 0.24 Th1 cell development BCL2 B-cell CLL/lymphoma 2 0.23 anti-apoptosis CD3E CD3e molecule, epsilon (CD3-TCR complex) 0.22 T cell activation IL7R interleukin 7 receptor 0.22 T cell proliferation CD3D CD3d molecule, delta (CD3-TCR complex) 0.22 T cell activation LAT linker for activation of T cells 0.22 T cell activation NOD3 NOD3 protein 0.21 T cell activation KLRB1 killer cell lectin-like r eceptor subfamily B, member 1 0.21 cell surface receptor linked signal transduction KLRF1 killer cell lectin-like r eceptor subfamily F, member 1 0.21 cell surface receptor linked signal transduction IL32 interleukin 32 0.21 pro-inflammatory cytokine GNLY granulysin 0.20 defense response CD2 CD2 molecule /// CD2 molecule 0.20 T cell activation IL2RB interleukin 2 receptor, beta /// interleukin 2 receptor, beta 0.19 T cell proliferation MS4A1 membrane-spanning 4-domains, subfamily A, member 1 0.19 B cell activation GZMA granzyme A (granzyme 1, cytotoxic T-lymphocyteassociated serine esterase 3) 0.19 cytolytic effector response LCK lymphocyte-specific protein tyrosine kinase 0.18 T cell activation CCR6 chemokine (C-C motif) receptor 6 0.17 chemotaxis CD8A CD8a molecule /// CD8a molecule 0.17 T cell activation CXCL5 chemokine (C-X-C motif) ligand 5 0.17 chemotaxis Table3 2.Continued


32 Gene Symbol Gene Title FC Function CD28 CD28 molecule 0.17 regulation of T cell proliferation KLRC1 /// KLRC2 killer cell lectin-like recept or subfamily C, member 1 // member 2 0.17 cellular defense response GZMK granzyme K (granzyme 3; tryptase II) /// granzyme K (granzyme 3; tryptase II) 0.16 cytolytic effect response CD96 CD96 molecule 0.16 NK/T cell adhesion NCR3 natural cytotoxicity triggering receptor 3 0.14 cytolytic effector response IGH@ /// IGHG1 /// IGHM immunoglobulin heavy locus /// immunoglobulin heavy constant gamma 1 // mu 0.08 antigen presentation (MHC I) Expressed as the log2 fold change (F C) in signal intensity, from preto post-G-CSF (FC>1, upregulated; FC<1, down-regulated) Table3 2.Continued


33 -3.0 -2.0 -1.0 0 1.0 2.0 3.0 Table 3-3. Characteristics of aGVHD and aGVHD-Free Recipients Patient ID Age Disease Type of Transplant Date of Engraftment aGVHD Date of aGVHD 1 41 AML MUD PBSCT 5/26/2006 no 2 42 AML MUD PBSCT 4/3/2006 no 3 52 AML Related PBSCT 5/15/2006 no 4 52 AML Related PBSCT 4/17/2006 no 5 58 AML MUD PBSCT 3/2/2006 yes 3/7/2006 6 59 AML MUD PBSCT 2/15/2006 yes 2/20/2006 7 60 ALL Related PBSCT 5/22/2006 yes 5/30/2006 8 61 AML Related PBSCT 1/9/2006 yes 1/17/2006 Abbreviations: AML, acute myeloid leukemia; ALL, acute lymphocytic leukemia; MUD, matched unrelated donor; PBSCT, peripheral blo od stem cell transplantation; aGVHD, acute graft-versus-host disease. Figure 3-2. Gene expression profil e of aGVHD. Shown is the uns upervised hierarchical cluster analysis of genes and samples for PBSCT r ecipients peripheral blood cells collected at engraftment: 1-4, aGVHD-free; 5-8, a GVHD. 1,658 probe sets were significant based on a 10% FDR and employing a pooled error estimate. 704 genes were upregulated and 954 genes were down-regulated.


34 Table 3-4. Immune-Related Transcrip tional Response in aGVHD at Engraftment Probe ID Gene Symbol Gene Title FC* 223629_at PCDHB5 protocadherin beta 5 5.81 237493_at IL22RA2 interleuki n 22 receptor, alpha 2 3.82 224399_at PDCD1LG2 programmed cell death 1 ligand 2 /// programmed cell death 1 ligand 2 3.75 207849_at IL2 interleukin 2 3.22 241090_at PKD1 Polycystic kidney disease 1 (autosomal dominant) 3.00 212013_at PXDN peroxidasin homolog (Drosophila) 2.81 216851_at IGL@ Immunoglobulin lambda locus 2.78 208090_s_at AIRE autoimmune regulator 2.71 206341_at IL2RA interleukin 2 receptor, alpha 2.51 207113_s_at TNF tumor necrosis factor (TNF superfamily, member 2) 2.35 202585_s_at NFX1 nuclear transcription factor, X-box binding 1 2.12 1559754_at LTB Lymphotoxin beta (TNF superfamily, member 3) 2.04 209906_at C3AR1 complement component 3a receptor 1 1.94 212580_at ARTS-1 Type 1 tumor necrosis factor receptor sheddi ng aminopeptidase regulator 1.89 214398_s_at IKBKE inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase epsilon 1.74 216852_x_at IGL@ Immunoglobulin lambda locus 1.63 225527_at CEBPG CCAAT/enhancer binding protein (C/EBP), gamma 1.52 222233_s_at DCLRE1C DNA cross-link repair 1C (PSO2 homolog, S. cerevisiae) 1.44 204212_at ACOT8 acyl-CoA thioesterase 8 1.39 214574_x_at LST1 leukocyte specific transcript 1 1.36 200759_x_at NFE2L1 nuclear factor (erythroid-derived 2)-like 1 1.35 208771_s_at LTA4H leukotriene A4 hydrolase 1.33 200758_s_at NFE2L1 nuclear factor (erythroid-derived 2)-like 1 1.30 204039_at CEBPA CCAAT/enhancer binding protein (C/EBP), alpha 1.28 212684_at ZNF3 zinc finger protein 3 1.27 200778_s_at SEPT2 septin 2 1.27 208097_s_at TXNDC thioredoxin domain containing /// thioredoxin domain containing 1.26 205603_s_at DIAPH2 diaphanous homolog 2 (Drosophila) 1.22 214459_x_at HLA-C major histocompatibility complex, class I, C 0.86 218520_at TBK1 TANK-binding kinase 1 0.82 213749_at MASP1 mannan-binding lectin serine peptidase 1 0.78 206584_at LY96 lymphocyte antigen 96 0.74 220813_at CYSLTR2 cysteinyl leukotriene receptor 2 0.74 228891_at C9orf164 chromosome 9 open reading frame 164 0.65 244811_at IRAK1BP1 Interle ukin-1 receptor-associated kina se 1 binding protein 1 0.64 239876_at NFKB1 Nuclear factor of kappa light polypeptide gene enhancer in B-cells 1 (p105) 0.60 242903_at IFNGR1 Interfer on gamma receptor 1 0.57 207037_at TNFRSF11A tumor necrosis factor recep tor superfamily, member 11a, NFKB activator 0.57 209079_x_at PCDHG protocadherin gamma subfamily A,B,C 0.54 1555086_at STAT5B signal transducer and activator of transcription 5B 0.54 208448_x_at IFNA16 interferon, alpha 16 0.52 210000_s_at SOCS1 suppressor of cytokine signaling 1 0.51 210081_at AGER advanced glycosylation end product-specific receptor 0.46 210081_at AGER advanced glycosylation end product-specific receptor 0.46 216846_at IGL@ Immunoglobulin lambda locus 0.42 211199_s_at ICOSLG inducible T-cell co-stimulator ligand 0.40 1552804_a_at TIRAP toll-interleukin 1 receptor (TIR) domain containing adaptor protein 0.38 234320_at CD244 CD244 molecule, natural killer cell receptor 2B4 0.38


35 Probe ID Gene Symbol Gene Title FC 234340_at PROCR Protein C receptor, endothelial (EPCR) 0.33 211129_x_at EDA ectodysplasin A 0.32 203954_x_at CLDN3 claudin 3 0.27 210487_at DNTT deoxynucleotidy ltransferase, terminal 0.26 1554091_a_at TIRAP toll-interleukin 1 receptor (TIR) domain containing adaptor protein 0.23 1560861_at SCAP1 Src family associated phosphoprotein 1 0.22 228895_s_at ASB1 Ankyrin repeat and SOCS box-containing 1 0.22 243706_at CDO1 Cysteine dioxygenase, type I 0.21 224499_s_at AICDA activation-in duced cytidine deaminase 0.17 228704_s_at CLDN23 Claudin 23 0.17 203854_at CFI complement factor I 0.16 207533_at CCL1 chemokine (C-C motif) ligand 1 0.15 222285_at IGHD immunoglobulin heavy constant delta 0.14 1552995_at IL27 interleukin 27 0.12 232099_at PCDHB16 protocadherin beta 16 0.11 Expressed as the log2 fold change (F C) in signal intensity, from preto post-G-CSF (FC>1, upregulated; FC<1, down-regulated) Table3 4.Continued


36 Table 3-5. Potential Transcrip tional Markers of aGVHD Progression Probe ID Gene Symbol Gene Title FC* 218800_at SRD5A2L steroid 5 alpha-reductase 2-like 10.86 1559653_at GRTP1 Growth hormone regulated TBC protein 1 10.77 243119_at NUFIP2 Nuclear fragile X mental retardation protein interacting protein 2 9.51 215591_at SATB2 SATB family member 2 9.48 220623_s_at TSGA10 testis specific, 10 9.19 238919_at PCDH9 Protocadherin 9 9.14 226360_at ZNRF3 zinc and ring finger 3 8.80 1570352_at ATM ataxia telangiectasia mutated 8.76 1564192_at SVIL Supervillin 8.03 220184_at NANOG Nanog homeobox 7.62 1570293_at TBL1X transducin (beta)-like 1X-linked 7.27 215254_at DSCR1 Down syndrome critical region gene 1 7.16 217780_at PTD008 PTD008 protein 7.09 1552572_a_at MIPOL1 mirror-image polydactyly 1 7.02 234039_at TANC1 Tetratricopeptide repeat, ankyrin repeat and coiled-coil containing 1 6.08 223629_at PCDHB5 protocadherin beta 5 5.81 220403_s_at P53AIP1 p53-regulated apoptosis-inducing protein 1 5.74 208245_at RAB9P1 RAB9, member RAS oncogene family, pseudogene 1 5.52 1554507_at NAALAD2 N-acetylated alpha -linked acidic dipeptidase 2 5.10 211601_at CATR1 CATR tumorigenicity conversion 1 0.32 1553636_at TMCO5 transmembrane and coiled-coil domains 5 0.29 213782_s_at MYOZ2 myozenin 2 0.24 1563522_at DDX10 DEAD (Asp-Glu-Ala-Asp) box polypeptide 10 0.22 1567697_at PHC3 Polyhomeotic like 3 (Drosophila) 0.19 204289_at ALDH6A1 aldehyde dehydrogenase 6 family, member A1 0.19 242604_at CLNS1A Chloride channe l, nucleotide-sensitive, 1A 0.18 233700_at PPP1R12B Protein phosphatase 1, regulatory (inhibitor) subunit 12B 0.18 213306_at MPDZ multiple PDZ domain protein 0.17 224499_s_at AICDA activation-in duced cytidine deaminase 0.17 228704_s_at CLDN23 Claudin 23 0.17 220120_s_at EPB41L4A erythrocyte membrane protein band 4.1 like 4A 0.16 240845_at EVI5 Ecotropic viral integration site 5 0.16 234535_at RAD50 RAD50 homolog (S. cerevisiae) 0.16 240564_x_at SNCA Synuclein, alpha (non A4 component of amyloid precursor) 0.16 1552661_at PCDHGB7 protocadherin gamma subfamily B, 7 0.15 1567687_at CECR9 cat eye syndrome chromosome region, candidate 9 0.15 220700_at WDR37 WD repeat domain 37 0.15 1569873_at LIPL1 Lipase-like, ab-hydrolase domain containing 1 0.14 1554328_at STXBP4 syntaxin binding protein 4 0.14 207299_s_at GRM1 glutamate receptor, meta botropic 1 0.14 1566339_at SNORD8 small nucleolar RNA, C/D box 8 0.14 204879_at PDPN podoplanin 0.14 202544_at GMFB glia maturation factor, beta 0.14 1566491_at WWOX WW domain containing oxidoreductase 0.14 207613_s_at CAMK2A calcium/calmodulin-dependent protein kinase (CaM kinase) II alpha 0.13 1561564_at TMCC1 Transmembrane and coiled-coil domain family 1 0.13 1556724_at ZNF606 Zinc finger protein 606 0.13


37 Probe ID Gene Symbol Gene Title FC 221304_at UGT1A10 UDP glucuronosyltransferase 1 family, polypeptide A10 0.13 1568190_at GOT1 Glutamic-oxaloacetic transaminase 1, soluble (aspartate aminotransferase 1) 0.13 238375_at EVI1 Ecotropic viral integration site 1 0.12 205413_at MPPED2 metallophosphoesterase domain containing 2 0.12 217067_s_at DMP1 dentin matrix acidic phosphoprotein 0.12 238292_at SFT2D1 SFT2 domain containing 1 0.11 232099_at PCDHB16 protocadherin beta 16 0.11 240903_at L3MBTL4 L(3)mbt-like 4 (Drosophila) 0.11 220145_at ASAP aster-associated protein 0.10 220835_s_at ZNF407 zinc finger protein 407 0.10 244891_x_at PVT1 Pvt1 oncogene homolog, MYC activator (mouse) 0.10 235556_at LOC153222 Adult retina protein 0.09 1560342_at ZFAND3 Zinc finger, AN1-type domain 3 0.09 Expressed as the log2 fold change (F C) in signal intensity, from preto post-G-CSF (FC>1, upregulated; FC<1, down-regulated) Table3 5.Continued


38 CHAPTER 4 DISCUSSION 4.1 Microarray and Flow Cytometric Analysis of the PBSCT Allograft Although G-CSF is known to have immune-a ltering characteristics, the precise mechanisms behind such immunomodulatory effects are a matter of debate. Like most growth factors and cytokines, G-CSF ex erts broad effects on immune cel l composition, cytokine profiles and immune cell responses (Tayebi et al. 2001). A possible way to el ucidate these mechanisms is through the use of high th roughput technology, such as DNA microarray, which can test multiple hypotheses in a single experiment. Ge ne expression profili ng of G-CSF-mobilized PBSC products for allogeneic tr ansplantation has been only mini mally investigated (Hernandez et al. 2005; Graf et al. 2001; Ng et al. 2004). Despite this, th e power of DNA microarray technology can afford new perspectives on the mechanisms of immunomodulation due to GCSF, which will benefit both stem cell transplanta tion and in treatment of diseases characterized by immune dysregulation. Therefore, in the current study we inve stigated the gene expression profile G-CSF-mobilized blood from five healt hy donors, and along with flow cytometry data, attempted to offer novel insi ghts into the toler ogenic effects mediated by G-CSF. The ability of G-CSF to consistently induce in creases in white cell counts confirms its role as a potent hematopoietic growth factor, particularly for cells within the granulocytic lineage (Demetri & Griffin, 1991). Although CD34+ cells were not measured in pre-G-CSF donor samples, these levels are very low or unde tectable in the peripheral blood under normal circumstances. After G-CSF, howev er, there were high numbers of CD34+ stem cells detected, indicating efficient mobilization. In enumerating cells of the lymphoid lineage, a nearly 7.5-fold increase in T cell counts occurred, which is characteristic of G-CSF mobilization (Bensinger et al. 1996). Thus, in addition to stimulating myel oid cell differentiation, G-CSF also appears to


39 simulate cells of the lymphoid lineageparticularly T cellswhich are known to express the GCSF receptor (Morikawa et al. 2002; Franzke et al. 2003). An increase in CD4+CD25+ Treg cells and monocytes was observed, suggesting a role for G-CSF in promoting tolerance th rough the generation of IL-10 producing re gulatory cells that inhibit the activation of alloreactive T cells (Morris et al. 2004; Rutella et al. 2002; Fraser et al. 2006; Mielcarak et al. 1998). The role of Tregs in facil itating tolerance in the HSCT and in autoimmune disease remains an ac tive area of investigation (Becker et al. 2006). Consistent increases in DC counts were noted after G-CSF mobilization, with a predominance of DC2 over DC1. DC2 are know n to polarize nave T cells toward a Th2 phenotype which is believed to pl ay a role in preventing increas ed risk of aGVHD after PBSCT, without compromising anti-malignancy effects (Arpinati et al. 2000; Rossi et al. 2002). Indeed, the potent activity of DC in terms of antigen pr esentation is central to graft-versus-leukemia (GVL) effects needed to kill off quiescent cancer cells that may have evaded prior conditioning regimens (Reddy et al. 2005). This likely accounts for GVL effects seen in our previous studies where patients with high DC counts had lower relapse and improved survival (Reddy et al. 2005). Several genes relevant to the regulation of stem cell mobilization and function showed up on the microarray analysis, notably genes of the Wnt and Notch signaling pathways. Upregulated genes such as CD164, CD44, PLAUR, KLH12, STAT5B, GSK1 and PSEN1 include genes involved in regulation of adherence to the marrow stroma and the proliferative and repopulating potential of HSCs (Watts et al. 2000; Dimitroff et al. 2001; Selleri et al. 2006; Angers et al. 2006; Bunting et al. 2002; Katoh et al. 2006).


40 Hematopoietic stem cells regulate their capaci ty to self-renew or differentiate based on activity of the canonical Wnt signaling pathway (Reya & Clevers, 2005). Two downstream effectors of Wnt signaling are the TCF/ LEF family of DNA-binding proteins, which transcriptionally activate Wnt target genes such as cmyc and cyclin D1. The products of these genes then act to regulate cell cycle status, differentiation and self-renewal. LEF1, TCF and the Wnt receptor ligand (WNT8B) were down -regulated in post-G-CSF donors. The serine/threonine kinase GSK3 (GSK3B) (which antagonizes Wnt signaling by targeting -catenin for proteosomal degradation) was up-regulated after G-CSF. Nega tive regulation of Wnt signaling in terms of gene expression observed here suggests that at five days post-G-CSF, HSCs are receiving signals to undergo differe ntiation as opposed to self-renewal. An analysis of immune-related gene expres sion changes occurring after administration of G-CSF revels that pro-inflammato ry gene expression is active, as evident by up-regulation of IL1 receptor signaling components including MyD88 (M YD88), toll interacting proteins (TOLLIP) and tumor necrosis factor (TNF) genes. High levels of TNF produced by donor cells can enhance both GVH and GVL effects af ter transplantation (Schmaltz et al. 2003) and a recent report has shown that donor-der ived TNF contributes to th e pathogenesis of early aGVHD (Ewing et al. 2007). In addition to up-re gulation of TNF, several ge nes implicated in neutrophil function, including matrix metalloproteinase s (MMP9, MMP25), elastase (ELA2) and celladhesion molecules (CD47, CD164) were up-regula ted post-G-CSF. Several of the neutrophil protease genes reported here are involved in se paration of stem cells from the marrow stroma, which is recognized as the mechanism by which G-CSF effects stem cell mobilization (Levesque et al. 2002; Winkler et al. 2006). This data confirms that inflammation and stem cell mobilization go together, and that inflammatory events triggered by increased neutrophil activity


41 are necessary for efficient stem cell releas e from the marrow microenvironment (Velders et al. 2004). In rare cases, excessive inflammation may cause vascular or pulmonary complications in otherwise healthy donors (Lindemann & Rumberger, 1993; Arimura et al. 2005). Additionally, our finding that G-CSF elicits pro-inflammato ry responses could explain the clinical manifestations of engraftment syndrome that arise in some patients in the immediate postgrafting period following autologous or allogeneic PBSCT (Madero et al. 2002; Spitzer, 2001). We propose that with G-CSF mobilization, a delicat e balance exists between inflammation that is conducive to stem cell mobilization or harm ful for the donor and recipient tissues. Genes and pathways implicated cell-media ted immune responses were overwhelmingly repressed after G-CSF. Of note, several MH C class II genes (HLA-DR, HLA-Q) and other components of the antigen processing and pr esentation pathway were down-regulated. In addition, the majority of genes designated as important for T ce ll activation were downregulated. Most significant are components of the T cell receptor complex such as CD3, accessory molecules (CD8A) and others involved in the DC:T cell dialogue, including costimulatory molecules, CD28 and CD86. Finally, al pha and beta subunits of the IL-2 receptor were down-regulated suggesting that overall, T cell activati on is suppressed after G-CSF administration. This is highly significant because from the flow cytometry data analysis, both DC and T cells were increased dr astically; however, in terms of gene expression, the ability of DC to activate T cells may be compro mised after G-CSF administration. The paradigm for propagation of DC-mediated T cell responses is: 1) engagement of the TCR:antigen:MHC complex; 2) co-stimulation th rough B7 molecules (CD86) on the DC and 3) production of IL-2 and up-regulation of the high affinity IL-2 receptor chain on the naive T cell. Our microarray data revealed reveals that these critical junctions in DC-mediated T cell


42 activation are suppressed thr ough down-regulation of TCR subun its, MHC class II genes, intracellular activator kinases (LCK, FYN), co-s timulatory molecules (CD86, CD28) and the IL2 receptor (IL-2RA). Interestingly, the anti-IL-2r monoclonal IgG1 drug daclizumab is used clinically as an immunosuppressive prophylax is for kidney transplantation (Vinceti et al. 1998); however, it has not shown promise as a treatment for aGVHD (Lee et al. 2004). In addition, cytotoxic T cell-related genes such as perfor in (PRF1), granzyme (GZMB) and Fas ligand (FASLG) were uniformly suppressed post-G-CSF. The general consensus is that such cellular effector responses are respons ible for both GVH and GVL reac tions imposed by alloreactive donor NK and CD8+ T cells (Schmaltz et al. 2001; Maeda et al. 2005). Further, the inverse relationship between dendritic ce ll counts and antigen pr esentation gene expre ssion suggests that DCs mobilized with G-CSF are functionally i mmature and may promote host tolerance after transplantation. With respect to cytokine gene expression, IL-32 was down-regulated after G-CSF. IL-32 is a newly discovered, pro-inflammatory cytokine that participates in Th1 responses (Kim et al. 2005). Since G-CSF administration is classically associated with a bias towards Th2 responses, the down-regulation of IL-32 may potentially play a role in re organization of the cytokine milieu. Up-regulation of the IL-4 and IL-10 receptors after G-CSF was noted, possibly in response to increased Th2 activity. An unexpect ed finding, however, was the down-regulation of GATA-3, a transcription factor involved in the ma turation of Th2 cells. Previous studies have reported that GATA-3 is up-regulated after G-CSF (Zhu et al. 2006; Hernandez et al. 2005). The difference between our findings and prior studies may be due to variations in G-CSF dosage, timing of administration, or donor characteristics


43 In summary, G-CSF induces innate, pro-in flammatory gene expression while adaptive immune responses are antagonized as evident by strong repression of antigen presentation and T cell activation gene expression. Therefore, administration of GCSFwhile initially promoting granulocyte counts and innate inflammati onactually down-regulates adaptive immune responses. Our findings suggest that G-CSF shou ld be used cautiously in those for which cell mediated immunity is needed. A microarray analysis of the peripheral blood stem cell products represents a novel way to describe the immunosuppressive properties of G-CSF by exploiting the power of multiple hypothesis testing. Future studie s in a similar regard may be undertaken to further elucidate ways in which G-CSF alters immune responses by performing microarray analysis on purified cell populations. The ultimate goal will be translati ng that understanding into treatments that can balance innate and adaptive immune responses in disease states. 4.2 Toward a Molecular Signature of aGVHD A transcriptional profile or molecular signature of aGVHD based on global gene expression analysis of peripheral blood leukocytes in HSCT patients has not been previously undertaken. Certain groups have attempted to measure changes in gene expression occurring in cutaneous or hepatic aGVHD using mouse models of BMT (Ichiba et al. 2003; Zhou et al. 2006; Sugerman et al. 2004). These studies have shown that early aGVHD gene expression changes are dominated by IFNresponsive genes and genes involve d in leukocyte cell adhesion and trafficking, such as interferon-indu cible inflammatory chemokines. Our microarray data descri be a pro-inflammatory gene expression profile in aGVHD patients. Most apparent is the up-regulation of IL-2 and its high-affinity receptor, IL-2r IL-2 is a critical growth factor fo r T cells and in response to IL -2, T cells will up-regulate IL-2 receptor, thus amplifying proliferation in an au tocrine fashion. A similar increase in AIRE


44 expression may reflect additional T cell activation, as AIRE has b een shown to be expressed in activated, IL-2 responsive CD4+ T cells (Nagafuchi et al. 2006). During early-mid stage aGVHD, donor T cells may be prolif erating in response to host-der ived alloantigens. However, their cytolytic capacities appear to not be fully de veloped at this stage as reflected in a decrease in cytolytic effector genes. TIA1a marker for effector CD8+ T cells in chronic intestinal GVHD (Patey-Mariaud de Serre et al. 2002) was down-regulated in the aGVHD patients. Upregulation of the complement component 3a rece ptor (C3AR1), which f acilitates cell migration to sites of inflammation, suggests activation of DC and lymphocytes, as both activated DC and T cells express high levels of this receptor (Gutzmer et al. 2004; Werfel et al. 2000). Upregulation of the critical proinflammatory cytokine TNF agr ees with the cytokine storm described during early GVHDa pr ocess which enhances the prol iferation and maturation of donor immune cells. Of note is the down regulati on of IL-27 in aGVHD patients. IL-27 is a novel IL-12 family cytokine with an apparently dichotomous role in immune responses. Some groups have reported that IL-27 exhibits pro-inflammatory properties by activating Th1 cells that may even enhance immunological responses to tumor cells (Shimuzu et al. 2006). Alternatively, it has been shown that IL-27 possesses anti-inflamm atory effects as well, mainly through the antagonism of IL-17-produc ing Th17 cells (Colgan & Rothman, 2006; Stumhofer et al. 2006). Th17 cells produce the pro-inflammatory cytokine IL-22 which is implicated in autoimmune disorders such as Crohns diseas e and dermal acanthosis (Brand et al. 2006; Zheng et al. 2006). Interestingly, the receptor for IL-22 (IL-22R 2) was consistently up-regulated in aGVHD patients. Thus, perhaps the most valuable knowledge gained from this stu dy is the importance of the IL-27/Th17/IL-22 axis in the development of aGVHD, which has not been previously


45 described. The recent finding that IL-27 plays a role in inhibiting angiogenesis makes this finding significant even beyond stud ies in aGVHD, as angiogenes is can promote tumor growth (Shimuzu et al. 2006). Thus, it would be a worthwhile future aim to measure IL-27 and IL-22 levels in transplant patients, and attempt to corr elate cytokine levels with risk of acute leukemia relapse, which may occur as a consequence of de novo vasculogenesis.


46 CHAPTER 5 CONCLUSIONS AND FUTURE AIMS The results contained herein both confirm what is already known about certain aspects of HSCT and introduce new findings into the field wh ich may pique further investigation. In postG-CSF-mobilized donors, there was a tremendous in crease in immune cell counts, yet an overall decrease in immune response genes, particularly those related to adap tive immunity and cellmediated responses. This agrees with prev ious reports characteri zing G-CSF as a potent immunomodulatory growth factor with its effects being predom inantly through increases in Th2 cytokines, such as IL-10 and IL-4, and suppr ession of Th1 cell responses. Such findings are clinically very relevant because aGVHD is driv en by Th1 cells and cytokines. Thus G-CSF administration in donors prior to tr ansplant can effect some degree of tolerance in the recipient, without compromising the GVL effects imposed agai nst quiescent cancer cells that have evaded prior conditioning regimens. A microarray analysis of the peri pheral blood stem cell products represents a new way to describe these m echanisms by exploiting the power of multiple hypothesis testing. Further studie s in this regard should be u ndertaken for two main reasons. First, to further elucidate the ways in which G-CSF alters immune responses with the ultimate goal of translating that understand ing into treatments for other Th1driven diseases, such as Type I diabetes; and second, to uncover other mechan isms for how G-CSF induces mobilization of stem cells from the marrow. Although the use of microarray analysis has permeated almost every aspect of medical research, it is rarely reported in the HSCT lite rature. To date, there is yet to be a single publication showing a molecular signature of GVHD in humans. Although our study was pioneering in that regard, there is much work to be done. Many more patient samples will need


47 to be analyzed before we can conclude what constitutes a molecular signature of GVHD, or what transcriptional markers can accurately predict di sease progression or response to treatment.


48 LIST OF REFERENCES Allison D.B., Cui X., Page G.P. & Sabripour M. ( 2006) Microarray data analysis: from disarray to consolidation and consensus. Nature Reviews Genetics 7 55-65. Angers S., Thorpe C.J., Biechele T.L., Goldenberg S.J., Zheng N., MacCoss M.J. & Moon RT. (2006) The KLHL12-Cullin-3 ubiquitin ligase ne gatively regulates the Wnt-beta-catenin pathway by targeting Dish evelled for degradation. Nature Cell Biology 8 348-357. Arimura K., Inoue H., Kukita T., Matsushita K., Akimot M., Kawamata N., Yamaguchi A., Kawada H., Ozak A., Arima N. & Te C. (2005) Acut e lung Injury in a healthy donor during mobilization of peripheral blood st em cells using granulocyte-co lony stimulating factor alone. Haematologica 90 (ECR10). Arpinati M., Green C.L., Heimfeld S., Heuser J.E. & Anasetti C. (2000) Granulocyte-colony stimulating factor mobilizes T he lper 2-inducing dendritic cells. Blood 95 2484-2490. Attal M., Harousseau J.L., Stoppa A.M., Sotto J.J., Fuzibet J.G., Rossi J.F., Casassus P., Maisonneuve H., Facon T., Ifrah N., Payen C. & Bataille R. (1996) A prospective, randomized trial of autologous bone marrow tr ansplantation and chemotherapy in multiple myeloma. Intergroupe Francais du Myelome. New England Journal of Medicine 335 91-97. Banovic T., MacDonald K.P., Morris E.S., Rowe V., Kuns R., Don A., Kelly J., Ledbetter S., Clouston A.D. & Hill G.R. (2005) TGF-beta in a llogeneic stem cell transplantation: friend or foe? Blood, 106 2206-2214. Becker C., Stoll S., Bopp T., Schmitt E. & Jonulei t H. (2006) Regulatory T cells: present facts and future hopes. Medical Microbiology and Immunology 195 113-124. Bensinger W.I., Clift R., Martin P., Appelbaum F.R ., Demirer T., Gooley T., Lilleby K., Rowley S., Sanders J., Storb R. & Buckner C.D. (1996) Allogeneic peripheral blood stem cell transplantation in patients with advanced hematologic malignancies: a retrospective comparison with marrow transplantation. Blood 88 2794-2800. Brand S., Beigel F., Olszak T., Zitzmann K., Eichhor st S.T., Otte J.M., Die polder H., Marquardt A., Jagla W., Popp A., Leclair S., Herrmann K ., Seiderer J., Ochsenkuhn T., Goke B., Auernhammer C.J. & Dambacher J. (2006) IL22 is increased in active Crohn's disease and promotes proinflammatory gene expression a nd intestinal epitheli al cell migration. American Journal of Physiology Gastrointestinal and Liver Physiology 290 827-838. Bunting K.D., Bradley H.L., Hawley T.S., Moriggl R., Sorrentino B.P. & Ihle J.N. (2002) Reduced lymphomyeloid repopulating activity from adult bone marrow and fetal liver of mice lacking expression of STAT5. Blood 99, 479-487. Cheng C., Pounds S.B., Boyett J.M., Pei D., Kuo M.L. & Roussel M.F. (2004) Statistical significance threshold criteria for analys is of microarray gene expression data. Statistical Applications in Geneti cs and Molecular Biology 3 (Article 36)


49 Cobb J.P., Mindrinos M.N., Miller-Graziano C., Calv ano S.E., Baker H.V., Xiao W., Laudanski K., Brownstein B.H., Elson C.M., Hayden D.L ., Herndon D.N., Lowry S.F., Maier R.V., Schoenfeld D.A., Moldawer L.L., Davis R.W ., Tompkins R.G. & Inflammation and Host Response to Injury Large-Scale Collaborativ e Research Program. (2005) Application of genome-wide expression analysis to human health and disease. Proceedings of the National Academy of Sciences 102 4801-4806 Colgan J. & Rothman P. (2006) All in the family: IL-27 suppression of T(H)-17 cells. Nature Immunology 7 899-901. Demetri G.D. & Griffin J.D. ( 1991) Granulocyte colony-stimulat ing factor and its receptor. Blood 78 2791-2808. Dimitroff C.J., Lee J.Y., Rafii S., Fuhlbrigge R.C. & Sackstein R. (2001) CD44 is a major Eselectin ligand on human hema topoietic progenitor cells. Journal of Cell Biology 53 12771286. Eisen M.B., Spellman P.T., Brown P.O. & Botste in D. (1998) Cluster analysis and display of genome-wide expression patterns. Proceedings of the National Academy of Sciences 95 14863-14868. Ewing P., Miklos S., Olkiewicz K.M., Muller G., Andreesen R., Holler E., Cooke K.R. & Hildebrandt G.C. (2007) Donor CD4+ T-cell production of tu mor necrosis factor alpha significantly contributes to the early proinflammatory events of graft-versus-host disease. Experimental Hematology 35 155-163. Franzke A. (2006) The role of G-CSF in adaptive immunity. Cytokine and Growth Factor Reviews 17 235-244 Franzke A., Piao W., Lauber J., Gatzlaff P., Konecke C., Hansen W., Schmitt-Thomsen A., Hertenstein B., Buer J. & Ganser A. (2003) G-CSF as immune regulator in T cells expressing the G-CSF receptor: implications for tr ansplantation and autoimmune diseases. Blood 102 734-739. Fraser A.R., Cook G., Franklin I.M., Templet on J.G., Campbell M., Holyoake T.L. & Campbell J.D. (2006) Immature monocytes from GCSF-mobilized peripheral blood stem cell collections carry surface-bound IL-10 and have the potential to modulate alloreactivity. Journal of Leukocyte Biology 80 862-869. Gianni A.M., Siena S., Bregni M., Tarella C ., Stern A.C., Pileri A. & Bonadonna G. (1989) Granulocyte-macrophage colony-st imulating factor to harvest ci rculating haemopoietic stem cells for autotransplantation. Lancet 2 580-585


50 Graf L., Heimfeld S. & Torok-Storb B. (2001) Co mparison of gene expression in CD34+ cells from bone marrow and G-CSF-mobilized peripheral blood by high-density oligonucleotide array analysis. Biology of Blood and Marrow Transplantion 7 486-494. Gutzmer R., Lisewski M., Zwirner J., Mommert S ., Diesel C., Wittmann M., Kapp A. & Werfel T. (2004) Human monocyte-derived dendritic cells are chemoattracted to C3a after up-regulation of the C3a receptor with interferons. Immunology 111 435-443. Hernandez J.M., Castilla C., Gutierrez N.C., Isidro I.M., Delgado M., de las Rivas J., Ferminan E., Garcia J.L., Ocio E.M., del Canizo M.C. & San Miguel JF. (2005) M obilisation with G-CSF in healthy donors promotes a high bu t temporal deregulation of genes. Leukemia 19 10881091. Hunter C.A. (2005) New IL-12-family members: IL-23 and IL-27, cytokines with divergent functions. Nature Reviews Immunology 5 521-531. Ichiba T., Teshima T., Kuick R., Misek D.E., Liu C., Takada Y., Maeda Y., Reddy P., Williams D.L., Hanash S.M. & Ferrara J.L. (2003) Early ch anges in gene expression profiles of hepatic GVHD uncovered by oligonucleotide microarrays. Blood, 102 763-71. Iwasaki T. (2004) Recent advances in the treatment of graft-versus-host disease. Clinical Medicine & Research 2 243-252. Kared H., Masson A., Adle-Biassette H., Bach J. F., Chatenoud L. & Zavala F. (2005) Treatment with granulocyte colony-stimulating factor pr events diabetes in NOD mice by recruiting plasmacytoid dendritic cells and func tional CD4(+)CD25(+) regulatory T-cells. Diabetes 54 78-84. Katoh M. & Katoh M. (2006) Cross-talk of WN T and FGF signaling pathways at GSK3beta to regulate beta-catenin and SNAIL signaling cascades. Cancer Biology and Therapy 5 10591064. Kim S.H., Han S.Y., Azam T., Yoon D.Y. & Dinare llo C.A. (2005) Interle ukin-32: a cytokine and inducer of TNFalpha. Immunity 22 131-142. Korbling M., Przepiorka D., Huh Y. O., Engel H., van Besien K.,Gir alt S., Andersson B., Kleine H.D., Seong D. & Deisseroth A.B. (1995) A llogeneic blood stem cell transplantation for refractory leukemia and lymphoma: potential advantage of blood over marrow allografts. Blood, 85 1659-1665. Korzenik J.R. & Dieckgraefe B.K. (2005) An openlabeled study of granulocyte colony-stimulating factor in the treatment of active Crohn's disease. Ailment Pharmacology & Therapeutics 21 391-400. Lane T.A., Law P., Maruyama M., Young D., Burg ess J., Mullen M., Mealiffe M., Terstappen L.W., Hardwick A. & Moubayed M. (1995) Ha rvesting and enrichment of hematopoietic


51 progenitor cells mobilized into the peri pheral blood of normal donors by granulocytemacrophage colony-stimulating factor (GM-CSF) or G-CSF: Potential role in allogeneic marrow transplantation. Blood 85 275-282 Lawlor K.E., Campbell I.K., Metcalf D., O'D onnell K., van Nieuwenhuijze A., Roberts A.W. & Wicks I.P. (2004) Critical role for granulocyt e colony-stimulating factor in inflammatory arthritis. Proceedings of the Nati onal Academy of Sciences 101 11398-11403. Lee S.J., Zahrieh D., Agura E., MacMillan M.L., Maziarz R.T., McCarthy P.L. Jr, Ho V.T., Cutler C., Alyea E.P., Antin J.H. & Soiffer R.J. (2004) Effect of up-front daclizumab when combined with steroids for the treatment of acute graft-versus-host disease: results of a randomized trial. Blood, 104, 1559 Levesque J.P., Hendy J., Takamatsu Y., Williams B., Winkler I.G. & Simmons P.J. (2002) Mobilization by either cyclophosphamide or granul ocyte colony-stimulating factor transforms the bone marrow into a highly proteolytic environment. Experimental Hematology 30 440449. Lindemann A. & Rumberger B. (1993) Vascular complications in pati ents treated with granulocyte colony-stimulating factor (G-CSF). European Journal of Cancer, 29 2338-2339. Madero L., Vicent M.G., Sevilla J., Prudencio M ., Rodriguez F. & Diaz M.A. (2002) Engraftment syndrome in children undergoing autologous peri pheral blood progenitor cell transplantation. Bone Marrow Transplantation 30 355-358. Maeda Y., Levy R.B., Reddy P., Liu C., Clouthier S. G., Teshima T. & Ferrara J.L. (2005) Both perforin and Fas ligand are required for the regulation of alloreactive CD8+ T cells during acute graft-versus-host disease. Blood 105 2023-2027. Mielcarek M., Graf L., Johnson G. & Torok-St orb B. (1998) Production of interleukin-10 by granulocyte colony-stimulating factor-mobili zed blood products: a mechanism for monocytemediated suppression of T-cell proliferation. Blood 92 215-222. Morikawa K., Morikawa S., Nakamura M. & Miyawa ki T. (2002) Characte rization of granulocyte colony-stimulating factor receptor expressed on human lymphocytes. British Journal of Hematology 118 296-304. Morris E.S., MacDonald K.P., Rowe V., Johnson D.H., Banovic T., Clouston A.D. & Hill GR. (2004) Donor treatment with pegylated G-CSF augments the generation of IL-10-producing regulatory T cells and promotes transplantation tolerance. Blood 103 3573-2581. Nagafuchi S., Katsuta H., Koyanagi-Katsuta R., Yamasaki S., Inoue Y., Shimoda K., Ikeda Y., Shindo M., Yoshida E., Matsuo T., Ohno Y., K ogawa K., Anzai K., Kurisaki H., Kudoh J., Harada M. & Shimizu N. (2006) Autoimmune Regulator (AIRE) Gene Is Expressed in Human Activated CD4(+) T-Cells and Regulat ed by Mitogen-Activat ed Protein Kinase Pathway. Microbiology and Immunology 50 979-987.


52 Ng Y.Y., van Kessel B., Lokhorst H.M., Baert M.R ., van den Burg C.M., Bloem A.C. & Staal F.J. (2004) Gene-expression profiling of CD34+ cells from various hematopoietic stem-cell sources reveals functional diffe rences in stem-cell activity. Journal of Leukocyte Biology 75 314-323. Patey-Mariaud de Serre N., Reija sse D., Verkarre V., Canioni D ., Colomb V., Haddad E & Brousse N. (2002) Chronic intestinal graft-versus -host disease: clinic al, histological and immunohistochemical analysis of 17 children. Bone Marrow Transplantation 29 223-30. Philip T., Guglielmi C., Hagenbeek A, Somers R., van der Lelie H., Bron D., Sonneveld P., Gisselbrecht C., Cahn J.Y., Harousseau J.L., Co iffer B., Biron P., Mandelli F. & Chauvin F. (1995) Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sens itive non-Hodgkin's lymphoma. New England Journal of Medicine 333 1540-1545. Reddy P. & Ferrara J.L. (2003) Immunobiology of acute graft-versus-host disease. Blood Reviews 17 187-194. Reddy V., Hill G.R., Pan L., Gerbitz A., Teshima T., Brinson Y. & Ferrara J.L. (2000) G-CSF modulates cytokine profile of dendritic cells and decreases acute graf t-versus-host disease through effects on the donor ra ther than the recipient. Transplantation 69 691-693. Reddy V., Iturraspe J.A., Tzolas A.C., Meier-Krie sche H.U., Schold J. & Wingard J.R. (2004) Low dendritic cell count after alloge neic hematopoietic stem cell tran splantation predicts relapse, death, and acute graftversus-host disease. Blood, 103 4330-4335. Reddy P., Maeda Y., Liu C., Krijanovski O.I., Korngol d R. & Ferrara J.L. (2005) A crucial role for antigen-presenting cells and a lloantigen expression in graft-versus-leukemia responses. Nature Medicine 11 1244-1249. Reiner A., Yekutieli D. & Benjamini Y. (2003) Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 19 368-375. Reya T., Morrison S.J., Clarke M.F. & Weissman I.L. (2001) Stem cells, cancer, and cancer stem cells. Nature 414 105-111. Rossi M., Arpinati M., Rondelli D. & Anasetti C. (2002) Plasmacytoid dendritic cells: do they have a role in immune responses after hematopoietic cell transplantation? Human Immunology 63 1194-200. Rutella S., Pierelli L., Bonanno G., Sica S., Am eglio F., Capoluongo E., Mariotti A., Scambia G., d'Onofrio G. & Leone G. (2002) Role for gr anulocyte colony-stimulating factor in the generation of human T regulatory type 1 cells. Blood, 100 2562-2571. Schmaltz C., Alpdogan O., Horndasch K.J., Muriglan S.J., Kappel B.J., Teshima T., Ferrara J.L., Burakoff S.J. & van den Brink M.R. (2001) Di fferential use of Fas ligand and perforin


53 cytotoxic pathways by donor T cells in graft-ve rsus-host disease and graft-versus-leukemia effect. Blood, 97 2886-2895. Schmaltz C., Alpdogan O., Muriglan S.J., Kappel B.J ., Rotolo J.A., Ricchetti E.T., Greenberg A.S., Willis L.M., Murphy G.F., Crawford J.M. & va n den Brink MR. (2003) Donor T cell-derived TNF is required for graft-versus-host diseas e and graft-versus-tum or activity after bone marrow transplantation. Blood, 101 2440-2445. Selleri C., Montuori N., Ricci P ., Visconte V., Baiano A., Carriero M.V., Rotoli B., Rossi G. & Ragno P. (2006) In vivo activity of the cleaved form of soluble urokinase receptor: a new hematopoietic stem/pr ogenitor cell mobilizer. Cancer Research 66 10885-10890. Shimizu M., Shimamura M., Owaki T., Asakawa M ., Fujita K., Kudo M., Iwakura Y., Takeda Y., Luster A.D., Mizuguchi J. & Yoshimoto T. ( 2006) Antiangiogenic and an titumor activities of IL-27. Journal of Immunology 176 7317-7324. Spitzer T.R. (2001) Engraftment syndrome follo wing hematopoietic stem cell transplantation. Bone Marrow Transplantation 27 893-898. Storek J., Dawson M.A., Storer B., Stevens-Ayer s T., Maloney D.G., Marr K.A., Witherspoon R.P., Bensinger W., Flowers M.E., Martin P., St orb R., Appelbaum F.R. & Boeckh M. (2001) Immune reconstitution after allogeneic marrow tr ansplantation compared with blood stem cell transplantation. Blood, 97 3380-3389. Stricker R.B. & Goldberg B. (1996) G-C SF and exacerbation of rheumatoid arthritis. American Journal of Medicine 100 665-666. Stumhofer J.S., Laurence A., Wilson E.H., Huan g E., Tato C.M., Johnson L.M., Villarino A.V., Huang Q., Yoshimura A., Sehy D., Saris C.J., O'Shea J.J., Hennighausen L., Ernst M. & Hunter C.A. (2006) Interleukin 27 negatively regulates the development of interleukin 17producing T helper cells duri ng chronic inflammation of th e central nervous system. Nature Immunology 7 937-945. Sugerman P.B., Faber S.B., Willis L.M., Petrovic A., Murphy G.F., Pappo J., Silberstein D. & van den Brink M.R. (2004) Kinetics of gene expres sion in murine cutaneous graft-versus-host disease. American Journal of Pathology 164 2189-2202. Tayebi H., Kuttler F., Saas P., Lienard A., Petracca B., Lapierre V., Ferrand C., Fest T., Cahn J., Blaise D., Kuentz M., Herve P., Tiberghien P. & Robinet E. (2001) E ffect of granulocyte colony-stimulating factor mobilization on phe notypical and functional properties of immune cells. Experimental Hematology 29 458-70. Thomas E.D., Lochte H.L. Jr, Lu W.C. & Ferre bee J.W. (1957) Intravenous infusion of bone marrow in patients receiving radiation and chemotherapy. New England Journal of Medicine., 257 491-496.


54 van den Brink M.R. & Burakoff S.J. (2002) Cyto lytic pathways in haematopoietic stem-cell transplantation. Nature Reviews Immunology 2 273-81. Vinceti F., Kirkman R., Light S ., Bumgardner G., Pescovitz M., Ha lloran P., Neylan J., Wilkinson A., Ekberh H., Gaston R., Backman L. & Burdic k J. (1998) Interleukin-2-receptor blockade with daclizumab to prevent acute rejection in renal transplantation. New England Journal of Medicine 338 161-165. Volpi I., Perruccio K., Tosti A., Capanni M., Ruggeri L., Posati S., Aversa F., Tabilio A., Romani L., Martelli M.F. & Velardi A. (2001) Postgr afting administration of granulocyte colonystimulating factor impairs functional immune recovery in recipients of human leukocyte antigen haplotype-mismatched hematopoietic transplants. Blood, 97 2514-2521. Watt S.M. & Chan J.Y. (200) CD164--a novel sialomucin on CD34+ cells. (2000) Leukemia & Lymphoma 37 1-25. Werfel T., Kirchhoff K., Wittmann M., Begemann G., Kapp A., Heidenreich F., Gotze O., & Zwirner J. (2000) Activated human T lymphoc ytes express a functional C3a receptor. Journal of Immunology 165 6599-6605. Winkler I.G. & Levesque J.P. (2006) Mechanisms of hematopoietic stem cell mobilization: when innate immunity assails the cells that make blood and bone. Experimental Hematology 34 996-1009. Zheng Y., Danilenko D.M., Valdez P., Kasman I ., Eastham-Anderson J., Wu J. & Ouyang W. (2007) Interleukin-22, a T(H)17 cy tokine, mediates IL-23-induced dermal inflammation and acanthosis. Nature 445 648-651. Zhou L., Askew D., Wu C. & Gilli am A.C. (2007) Cutaneous Gene Expression by DNA Microarray in Murine Sclerodermatous Graft-Versus-Hos t Disease, a Model for Human Scleroderma. Journal of Investigative Dermatology 127 281-292. Zhu J., Yamane H., Cote-Sierra J., Guo L. & Pa ul W.E. (2006) GATA-3 promotes Th2 responses through three different mechanisms: induction of Th2 cytokine production, selective growth of Th2 cells and inhibition of Th1 cell-specific factors. Cell Research 16 3-10.


55 BIOGRAPHICAL SKETCH Matthew Peter Buzzeo was born in New York, NY, in 1981 and grew up in Tampa, FL. He graduated summa cum laude with a Bachelor of Science from the University of South Florida in 2003. Here he was inducted into the College of Arts and Sciences Honor Society and named a Provosts Scholar. Subsequently, Mr. Buzzeo join ed the H. Lee Moffitt Cancer Center as a Research Associate in Department of Interdis ciplinary Oncology where he investigated the evolution of adaptive immune receptors and st udied novel cell cycle regulatory elements as potential targets for cancer therapy. In 2005, he began graduate studies into the molecular aspects of hematopoietic stem cell transplantat ion at the University of Florida College of Medicine.