UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Novel pathway of thymus-dependent NK cell development Veinotte, Linnea Lora 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2007-268179.pdf [ 15.11MB ]
Metadata
JSON: 831-1.0100554.json
JSON-LD: 831-1.0100554-ld.json
RDF/XML (Pretty): 831-1.0100554-rdf.xml
RDF/JSON: 831-1.0100554-rdf.json
Turtle: 831-1.0100554-turtle.txt
N-Triples: 831-1.0100554-rdf-ntriples.txt
Original Record: 831-1.0100554-source.json
Full Text
831-1.0100554-fulltext.txt
Citation
831-1.0100554.ris

Full Text

N O V E L P A T H W A Y OF T H Y M U S - D E P E N D E N T N K C E L L D E V E L O P M E N T  by LINNEA LORA VEINOTTE  B . S c . H , Acadia University, 2001  A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF  D O C T O R OF  PHILOSOPHY  in  T H E F A C U L T Y OF G R A D U A T E STUDIES  (Genetics)  T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A December 2006 © Linnea Lora Veinotte, 2006  ABSTRACT Natural killer ( N K ) cells are the major lymphocytes o f the innate immune system, but their developmental pathway is not fully defined. It is commonly assumed that all N K cells develop in the bone marrow. In this thesis, I describe a novel thymus-dependent pathway o f N K cell development that is specific to those in the thymus and lymph nodes. Microarray analysis revealed T C R y m R N A expression in N K cells. Genomic and R T - P C R showed that some N K cells have rearranged T C R y genes while T C R p and T C R 8 genes are i n germline order. N K cells with rearranged T C R y (Tcry N K cells) were absent in nude mice indicating that they are thymus-dependent. Approximately half o f thymus N K cells have rearranged T C R y genes and in vitro cultures o f immature thymocytes (double negative (DN)1 and D N 2 progenitors) produced Tcry N K cells, strongly suggesting that these are the thymus-dependent N K cellprogenitors in vivo. Thymus N K cells are C D 9 4 / N K G 2  h i  Ly49  l0  Mac-1  10  IL-7Ra  h i  and they  have normal cytotoxicity levels but reduced IFNy production. B y using T C R y gene rearrangement as a marker o f thymus-dependent N K cells, we showed that they are also present in lymph nodes ( L N s ) but in no other tissues tested. N K progenitors similar to immature thymocytes were found in L N s and L N Tcry N K cells and L N progenitors were also absent in nude mice. In vitro cultures and preliminary in vivo studies suggest that the N K progenitors in the L N give rise to mature N K cells. The results suggest that immature thymocytes migrate to L N s and differentiate into N K cells. It is likely that the thymusdependent N K cells play a special role in the immune response since their phenotype is unique. Finally, this study suggests that multiple pathways o f N K cell commitment exist in multiple tissues. The differences i n tissue environment may influence the phenotype and function o f N K cells, resulting in multiple subsets o f N K cells throughout the body.  ii  TABLE OF CONTENTS Abstract  ii  Table of contents  iii  List of tables  vi  List of figures  vii  List of abbreviations  ix  Acknowledgements  xi  Chapter 1. 1.1. 1.2. 1.3. 1.4.  1 1 5 7 10  1.5.  1.6. 1.7. 1.8. 1.9.  Introduction Introduction to N K cells Hematopoietic lineage commitment Commitment to the N K cell lineage N K cell development 1.4.1. Stages o f N K cell Development: from N K P to the mature N K cell 1.4.2. Factors involved in N K P production: Cytokines... 1.4.3. Factors involved in N K P production: Transcription factors N K cell maturation 1.5.1. M H C Class I Receptor acquisition 1.5.2. Factors involved in N K cell maturation T cells and development T C R rearrangement: V ( D ) J recombination T C R y locus and rearrangement patterns Thesis objectives and hypotheses  Chapter 2. 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.7. 2.8. 2.9.  Materials and methods Mice Antibodies Microarray Sample Preparation and Analysis Measuring D N A Genomic P C R Southern Blot RT-PCR Sequencing o f P C R products Tissue culture 2.9.1. L-cells 2.9.2. O P 9 cells 2.9.3. L A K cells 2.9.4. Thymus or L N D N progenitor culture and B M N K P progenitor culture 2.10. C e l l preparation iii  11 13 14 16 16 18 24 27 29 33 35 35 35 37 37 37 39 40 42 42 42 '42 42 42 43  2.11. 2.12. 2.13. 2.14.  Staining and F A C S sorting or analysis o f cells IFNy production assay Cytotoxicity assay L N D N cell transplantation 2.14.1. Intraperitoneal injection 2.14.2. Intravenous inj ection 2.15. Statistics  44 44 45 45 45 45 46  Chapter 3: Identification of a novel pathway of N K cell development that is thymus-dependent and includes T C R gene rearrangement 47 3.1. Introduction 3.2. Results 3.2.1. Microarray analysis reveals expression o f T C R y gene i n N K cells 3.2.2. T C R y genes are rearranged and expressed in N K cells 3.2.3. Specificity o f rearrangement 3.2.4. T C R y gene rearrangements in N K cells represent unselected, random recombination 3.2.5. Tcry + N K cells represent a small population o f total splenic N K cells 3.2.6. N K cells have a germline TCR(3 locus and may have initiated T C R 8 rearrangement 3.2.7. T C R y gene rearrangement in N K cells is not due to R A G expression in B M C L P s 3.2.8. The thymus is required for the development o f 7 c r y N K cells 3.3. Discussion +  47 48 48 53 54 55 56 59 60 61 63  Chapter 4. The thymus-dependent developmental pathway generates unique subsets of N K cells in thymus and lymph node 71 4.1. Introduction 4.2. Results 4.2.1. D N 1 and D N 2 thymocytes differentiate into N K cells 4.2.2. DN2-derived N K cells are TCRy 4.2.3. Thymus N K cells are phenotypically different from N K cells o f other tissues 4.2.4. T C R y gene rearrangement in L N N K cells suggests a link with DN-derived thymus N K cells 4.2.5. D N 1 and pre-DN2 cells in L N s give rise to N K cells in culture 4.2.6. The D N cells and Tcry N K cells in the L N are thymus-dependent  iv  71 72 72 75 75 80 81 84  4.2.7. Preliminary results suggest that L N D N progenitors give rise to N K cells in vivo 4.2.8. Thymus N K and L N N K cells produce lower levels o f IFNy upon stimulation 4.2.9. Thymus N K cells have normal levels o f cytotoxicity 4.3. Discussion  General discussion Thymus-dependent N K developmental pathway Mouse N K cell subsets Human N K cell subsets and L N pathway o f development Thymus N K cells L N N K cells T cell vs. N K cell lineage commitment 5.6.1. Notch signals 5.6.2. E proteins vs. Id2 5.7. Medical relevance 5.8. Final conclusions  84 86 87 88  Chapter 5. 5.1. 5.2. 5.3. 5.4. 5.5. 5.6.  94 96 98 99 100 102 103 104 105 106 107  Bibliography  109  LIST OF T A B L E S Table 1.1. Genetic mutations affecting the maturation o f N K cells  23  Table 2.1. List o f antibodies used in studies  36  Table 2.2. List o f primers used in studies  41  Table 4.1. Global F A C S analysis o f N K cells from the thymus, L N , B M , spleen, liver, and lung  77  Table 4.2. L N D N l / p r e - D N 2 cells produce small numbers o f N K cells following one i.p. and one i.v. transplantation  86  vi  LIST OF FIGURES Figure 1.1. N K cell strategies for detecting and killing target cells  2  Figure 1.2. Progenitors with N K cell lineage potential  9  Figure 1.3. Bipotent T / N K progenitors  10  Figure 1.4. Stages o f N K cell maturation  12  Figure 1.5. Stages o f T cell development  26  Figure 1.6. The process o f V ( D ) J recombination  29  Figure 2.1. A . The murine T C R y locus  39  Figure 3.1. Microarray data analysis o f differentially expressed genes between adult N K and neonatal N K cell samples  49  Figure 3.2. Detection o f T C R y gene expression in N K cells by microarray analysis  52  Figure 3.3. T C R y gene rearrangement and expression in N K cells  54  Figure 3.4. Sequences o f productive and non-productive T C R y gene rearrangements  56  Figure 3.5. L o w frequency o f N K cells with rearranged T C R y genes  58  Figure 3.6. Tcry N K cells are not due to contamination of T cells  59  Figure 3.7. TCR(3 and T C R 5 gene rearrangements in N K cells.  60  Figure 3.8. Lack o f T C R y gene rearrangement in B 6 mouse B cells  61  Figure 3.9. Lack of T C R y gene rearrangement in nude mouse N K cells and high T C R y gene rearrangement in thymus N K cells  62  Figure 3.10. Thymus N K cells have high levels of T C R y gene rearrangement  63  Figure 4.1. Scheme of thymus (or L N ) D N progenitor culture  72  Figure 4.2. Thymus D N 1 and D N 2 progenitors have the potential to give rise to N K cells during in vitro cultures  74  Figure 4.3. Thymus D N 2 derived N K cells have rearranged T C R y genes  75  vii  Figure 4.4. Thymus N K cells appear 'immature' ( M a c - l ' ° I L - 7 R a ) compared to other tissue N K cells ,  78  Figure 4.5. Thymus N K cells havethe highest levels of N K G 2 A / C / E and C D 9 4 expression and average 2B4 expression  79  Figure 4.6. Thymus N K cells have the lowest levels o f L y 4 9 A , G , D , and C/I expression  79  Figure 4.7. L N N K cells have the highest percentage of Tcry N K cells other than thymus N K cells  81  Figure 4.8. D N 1 and pre-DN2 progenitors are present in the lymph node and they possess N K cell potential in vitro  82  Figure 4.9. D N progenitors from IL-15" " mouse L N s still show N K cell potential in vitro  83  Figure 4.10. Nude mouse L N s have lower levels o f D N 1 and pre-DN2 progenitors than w i l d type L N s and nude L N N K cells do not have T C R y gene rearrangement  84  hl  7  Figure 4.11. L N N K cells and thymus N K cells produce lower levels o f IFNy after IL-12 and IL-18 stimulation than spleen N K cells  87  Figure 4.12. Spleen and thymus N K cells have similar levels o f cytotoxicity  88  Figure 5.1. Revised model o f N K cell development in the mouse  95  viii  LIST OF ABBREVIATIONS: ADCC: antibody-dependent cell-mediated cytotoxicity BM: bone marrow CLP: common lymphoid progenitor CMP: common myeloid progenitor DC: dendritic cell DEC: dendritic epidermal T cell  DMEM: Dulbecco's modified eagle's medium DN: double negative  dNTP: Deoxyribonucleotide triphosphate DP: double positive ELP: early lymphoid precursor ETP: early thymus progenitor FBS: fetal bovine serum FL: fetal liver F l t 3 L : Flt-3 ligand  FTOC: fetal thymic organ culture HEV: high endothelial venules HSC: hematopoietic stem cells i.p.: intraperitoneally i . v . : intravenously Id: inhibitors o f D N A binding  IFNy: interferon y KIR: killer inhibitory receptor LAK: lymphokine activated killer cell LN: lymph node L T P R : lymphotoxin beta receptor  MCMV: mouse cytomegalovirus 2 M E : 2-mercaptoethanol MEM: M i n i m u m essential medium eagle MHC: major histocompatibility complex mNK: mature N K cell MPP: multipotent progenitor NIK: N F - K B inducing kinase NK: natural killer NKP: N K precursor PCR: polymerase chain reaction P/S: penicillin streptomycin R A E T 1 : retinoic acid early transcripts RT-PCR: reverse transcription polymerase chain reaction RSS: recombination signal sequence ix  SIPi: sphingosine 1-phosphate type 1 receptor S A G E : serial analysis o f gene expression SCF: stem cell factor SP: single positive T C R : T cell receptor Tcry N K : N K cell with T C R y gene rearrangement T d T : terminal deoxyribonucleotide transferase T E C : thymic epithelial cell T - I E L : T intraepithelial lymphocyte TNF: tumour necrosis factor T N K P : bipotent T / N K cell precursor T R A F 6 : tumor necrosis factor receptor associated factor 6 T R A I L : TNF-related apoptosis-inducing ligand V D U P 1 : vitamin D3 upregulated protein 1 y c : common y chain  ACKNOWLEDGEMENTS Firstly, thank you to my supervisor Dr. Fumio Takei. M y studies were greatly enhanced by having you as a supervisor. I feel very lucky to have had you as my supervisor. Y o u have been extremely helpful and always interested and involved in my research. Y o u have put so much time into working with me on various projects and have helped so much with writing the thesis. Y o u have always been enthusiastic and encouraging and I have learned so much from you. Thank you! Thank you to my graduate committee members, Drs. D i x i e Mager, Rob K a y , and K e l l y McNagny, for their advice and guidance throughout my studies. Thank you to all o f my lab members, past and present, for their friendship and help. Special thanks to M o t o i Maeda for teaching me many laboratory techniques when I first started working in the lab. To N o o s h i n Tabatabaei, we have gone through classes, research, travelling to conferences together, etc. at the same time. It was nice to have you around to talk with. Thanks to Evette Haddad, E m i l y Mace, Erica Wilson, E v a Backstrom, Valeria A l c o n , L i s a Dreolini, Matt M a c L e o d , Reza Marwali, Carmine Carpenito, for teaching me how to do certain techniques, for answering questions, helping me through problems, and for making the lab such a great place to be! Y o u are what made my grad studies so much fun! Thank you to Nastaran Mohammadi and Christine Parachoniak for doing the cloning and sequencing o f the V y - C y R T - P C R products. Thanks to Chelsea Greenwood for helping me work on single cell R T - P C R . A n d finally thanks to T i m H a l i m for working on the L N D N progenitor in vitro cultures and for working together with me on the transplantation studies, especially for getting all o f the peripheral L N s and learning how to do intravenous injections. A very big thank you to the F A C S sorting staff, Lindsey, Gayle, Jaime, C a m , and Rick. Y o u were always a big help. Thank you to K a i Lucke for help with i.v. injections and supplying mice. Thank you to the Marcel B a l l y lab for nude mice and to the J A F and A R C staff. I would like to acknowledge my funding from the Michael Smith Foundation for Health Research for both junior and senior trainee awards and from U B C U G F awards. Thank you to my family, especially my parents, for their support and encouragement. A n d finally, thank you most o f all to my husband, Matt, to whom I owe very much. Thank you for giving up so much to come to Vancouver for me to do my P h D . Thank you for your constant support and understanding throughout the five years. I really needed your encouragement to keep working hard and to strive to always do my best. Y o u always make me want to reach higher. Y o u mean the world to me. Thank you!  xi  1 INTRODUCTION  1.1. Introduction to N K cells  Natural killer ( N K ) cells are lymphocytes that belong to the innate immune system. They have a large granular morphology and can mediate cellular cytotoxicity as well as release chemokines and inflammatory cytokines. N K cells become activated by cytokines or upon encountering target cells that express ligands for N K cell receptors. N K cells also play a role in activating adaptive immunity and participate in cross talk with dendritic cells (DCs). Upon activation, N K cells can directly lyse target cells by exocytosis o f perforin and granzymes . 1  They also secrete cytokines such as interferon y (IFNy) and tumor necrosis factor (TNF) and 2 3  chemokines ' . Elimination o f cells v i a FasL and TNF-related apoptosis-inducing ligand ( T R A I L ) pathways also occurs in N K cells ' . These ligands are mediators o f caspase4  5  dependent apoptosis in target cells expressing the corresponding receptors Fas and T R A I L - R . 6  N K cells also express C D 1 6 (FcyRIII) which binds I g G antibody coated cells and mediates antibody-dependent cell-mediated cytotoxicity ( A D C C ) .  N K cells have a recognition system that is encoded by non-rearranged genes and it involves multiple types o f receptors rather than one dominant receptor. These receptors can trigger N K cells individually or in combination, depending on each target cell it encounters. N K cells use various strategies to recognize target cells including recognition o f pathogen-encoded molecules, induced self recognition, and the most classically described missing-self recognition (Fig. 1.1).  1  Inhibitory receptor M H C class I molecule  Activating receptor Stimulatory ligand  Transformation or infection Induced-self recognition or pathogen molecule recognition  Missing-self recognition  Figure 1.1. N K cell strategies for detecting and killing target cells. When an N K cell encounters a healthy " s e l f cell (top), the inhibitory signals outweigh any activating signals and the N K cell does not k i l l . If an N K cell encounters a cell that has downregulated its M H C class I expression (left), inhibitory signals are absent while activating signals still occur. Therefore, the activating signals lead to target cell killing. If an N K cell encounters a target cell that is expressing induced stress molecules or pathogen molecules (right), inhibitory signals w i l l still occur but more activating signals w i l l be received upon recognition o f the stress or pathogens molecules and the net signal w i l l be activating, resulting i n N K cell killing. Revised from Raulet . 1  A s part o f the well characterized missing self recognition strategy, N K cells express receptors that are specific for major histocompatibility complex ( M H C ) class I molecules (reviewed in 8  ) . M H C class I molecules are expressed on the surface o f all healthy nucleated cells but are  often downregulated as a consequence o f infection, mutation, or transformation. Most o f the M H C class I-specific receptors transduce an inhibitory signal. Ligation o f these receptors delivers a dominant-negative signal to N K cells that prevents natural killing o f self cells. It is 2  the balance o f inhibitory and stimulatory signaling that determines the outcome o f NK-target cell interactions such that i f a N K cell encounters a healthy cell, its inhibitory signals received w i l l override activating signals. If, on the other hand, an N K cell encounters an unhealthy cell ( M H C class I "), an inhibitory signal is not received and the activating signals w i l l result in lo/  cytolytic function. Adult mice express two families o f M H C class I-specific receptors, Ly49 and C D 9 4 / N K G 2 , whereas fetal and neonatal N K cells express only the latter. Human N K cells express C D 9 4 / N K G 2 receptors and the killer inhibitory receptor ( K I R ) family.  Another recognition strategy is detection o f pathogen-encoded molecules. N K cells express the L y 4 9 H receptor which is stimulatory and binds to an M H C class I-like protein, m l 57, which is a mouse cytomegalovirus ( M C M V ) encoded protein . L y 4 9 H enables N K cells to 9  undergo considerable proliferation and to limit early stage M C M V infection . Other N K cell 10  receptors that are specific for pathogens are N K p 4 6 and N K p 4 4 which bind the influenza virus hemagglutinin . 11  N K cell receptors can also recognize induced self signals, (stress signals) that are upregulated during infection or on tumor cells. N K p 4 6 , N K p 4 4 , and N K p 3 0 receptors likely recognize ligands on tumor cells but the ligand identities are not yet k n o w n . The best characterized 12  receptor is N K G 2 D , which is expressed by almost all N K cells and plays a key role in immune responses, especially as an activating receptor that triggers N K cells to respond against tumors (reviewed by ) . N K G 2 D ligands are stress signals encoded by the host genome. These 6  include a diverse family o f ligands, retinoic acid early transcripts ( R A E T 1 ) that are shared between mice and humans. Family members include Rae-1, H-60, M u l t l and U L B P . Although the expression patterns o f N K G 2 D ligands are diverse and complex, the general  3  pattern is that they are expressed poorly on healthy cells but are upregulated on infected or tumor cells . 6  N K cells express multiple activating and inhibitory receptors that recognize M H C class I as well as many n o n - M H C class I ligands such as pathogen encoded molecules and self-induced ligands. It is ultimately the balance o f signals that determines the functional outcome o f the N K celhtarget cell interaction. N K cells exhibit complex repertoires o f receptors because there is random coexpression o f many possible combinations o f N K cell receptors. This process o f variable receptor gene expression results in a repertoire o f N K cells that can detect multiple changes in cells. This somewhat resembles T cell and B cell lymphocytes but N K cells differ from T and B cells since they can k i l l target cells without prior sensitization, their N K cell receptor loci do not require gene rearrangement and their repertoire is much smaller.  Although many o f the activating/inhibitory receptors and their ligands have been discovered, the biological relevance o f these molecules in host defense and the interactions between N K cells and other immune system cells is less well characterized. Through multiple pathways (cytokine secretion, direct killing, interaction with other cells), N K cells have been shown to be involved i n combating viral infections (cytomegalovirus, sendai virus, influenza virus, H I V , and ebola virus), parasites (P. falciparum  which causes malaria, P. berghei, and T. cruzi), and  bacteria (Shingella flexneri, M. tuberculosis).  These roles are reviewed by Lodoen, et al. .  N K cells also interact with D C s . They can reciprocally activate one-another by unknown receptor-ligand pairs and by cytokines (reviewed i n ) . D C s secrete IL-2, IL-12, IL-18, and 14  I F N a / p which induces cytotoxicity, proliferation, C D 6 9 expression and IFNy production in N K cells. N K cells, on the other hand, activate or induce maturation o f D C s through cell-  4  contact and T N F c c . One example o f N K cell-DC cross talk is during M C M V infection. D C s 14  induce N K cell expansion by release o f IL-12 and IL-18 and through L y 4 9 H . A l s o , the 1 5  presence o f L y 4 9 H N K cells results i n the maintenance o f D C s i n the spleen during acute +  M C M V infection . 15  Since N K cells play such an important role in the immune system, it is very important to understand all aspects o f N K cell biology, including their lineage commitment and maturation pathways as well as factors that affect their final phenotype and function. In this study, we have identified a new pathway o f N K cell lineage commitment that is thymus-dependent and gives rise to N K cells that reside i n the thymus and L N s . This thymus-dependent pathway influences the N K cell phenotype and function. Therefore, not all N K cells develop via the B M pathway that is currently accepted in the literature as the sole location o f N K cell development. We w i l l therefore review the current understanding o f N K cell precursors and their development i n the B M as well as their maturation steps and factors involved in this process. T cell development and V ( D ) J recombination o f T cell receptors (TCRs) w i l l also be reviewed since this step o f development occurs i n the thymus-dependent pathway of N K cell development.  1.2. Hematopoietic lineage commitment  A l l blood cells develop from hematopoietic stem cells (HSCs). These cells are multipotent and are capable o f self r e n e w a l ' . Fetal hematopoiesis occurs i n the liver and it switches to the 16  11  B M in the adult. The first step in hematopoiesis is typically described as the division into common myeloid progenitors ( C M P s )  1 8  and common lymphoid progenitors ( C L P s ) . This 19  5  model o f hematopoiesis assumes that cells consistently reach branch points i n the same order and that the timing o f the developmental choices is fixed such that when a cell reaches a branching point, it cannot progress further without choosing one branch or the other. It also assumes that the choice is binary so that i f a cell makes a positive choice for the one branch, it automatically makes a negative choice for the other branch. If developmental choices are considered in terms o f gene regulatory mechanisms though there is no longer a need for cells to encounter binary choices but rather cells pass through overlapping 'windows o f opportunity' to 20  give rise to certain cell types  *  *  . Lineage opportunities remain open as long as the cells express  essential enabling factors. B y turning on or off other subsets o f transcription factors, a precursor cell can switch its cell type to more than one other choice. The exact timing and order o f the choices may not always be the same and one cell fate can be blocked without commitment to another. Since there is diversity in gene expression o f progenitors within the same stage, the lineage decision o f each cell can occur somewhat randomly due to the fluctuations i n gene expression. The end point could be reached v i a more than one pathway . 20  Oligonucleotide array analysis o f H S C s , M P P s (multipotent progenitors), C L P s , and C M P s revealed that H S C s express non-hematopoietic genes as well as many hematopoietic genes in a 21  lineage promiscuous manner  . M P P s express both myeloid and lymphoid genes and C L P s  express genes for T, B , and N K cells before they're committed to a lineage. Interestingly, C L P s express the germline transcripts o f many T C R and B C R s , including germline T C R y and T C R p \ C L P s also express R A G - 1 . Lineage commitment is a result o f inactivation of nonlineage specific genes and the retention or activation o f enabling factors for the lineage. Therefore cells can express genes specific for a certain lineage before they are committed to that lineage and cells can express genes o f other lineages even though they w i l l differentiate  6  into another lineage. This stresses the separation o f the terms "specification" and "commitment" . 20  1.3. Commitment to the N K cell lineage  It is the common assumption that all N K cells develop in the B M o f adult mice. N K cells and B cells are both dependent on the B M , because B M ablation by 8 9 S R  22  or oestradiol results in 23  N K cell deficiencies. This defect is also seen i n osteopetrotic animals which have a defective and much smaller B M compartment due to an osteoclast deficiency . Commitment to the N K 24  cell lineage requires progressive restriction of lineage potential from multipotent H S C s to oligopotent precursors to unipotent committed N K cell precursors which no longer have the ability to become any other cell type. Several oligopotent precursors have been described that retain N K cell potential (Fig. 1.2). The two main populations i n the adult B M are lymphocyte precursors which posses B cell, T cell, and N K cell potential. Early lymphoid precursors (ELPs) are Lin'c-kit F l t 3 L and common lymphoid progenitors ( C L P s ) are Lin"c-kit'°IL-7Ra . +  It is unclear whether these are obligate intermediates for N K cells to arise. In c-kit-deficient mice, that lack all C L P s , N K P and mature N K cell numbers are normal '  .  O n the other hand, several studies suggest that N K cells do arise from C L P s or E L P s . Two studies using reporter genes for V ( D ) J recombinase and R A G 1 27  28  show that N K cells arise  from progenitors which express R A G 1 or V ( D ) J recombinase. One study showed that V ( D ) J recombinase is active at the C L P stage and its activity or evidence o f its activity is detectable in B , T, N K , and D C lineages . This does not necessarily mean that N K cells or others arise 27  only from C L P s because V ( D ) J recombinase could be active i n other progenitors as well. The  7  recombinase expression at the C L P stage was shown to be controlled by the B cell specific enhancer, Erag and approximately 5% o f N K cells do have I g H rearrangements . Therefore, 27  perhaps this I g H portion does arise from C L P s in which the B cell specific Erag enhancer +  controls recombination. In the other study, R A G 1 expression was detectable in a progenitor upstream o f the C L P stage, the E L P progenitors . R A G expression (or evidence o f previous 28  expression) was also detectable in mature N K cells i n this study.  This earlier activity in E L P s  suggests that N K cells can arise from earlier progenitors as well as C L P s . Kouro, et al. examined potentials in B M precursors. N K cells can develop from early Lin"c-kit cells (like hl  ELPs) but also from more mature Lin"c-kit'°(Flt3L ) cells (which include C L P s ) (Fig. 1.2). It +  seems that a significant proportion o f BM-derived N K cells do arise from a Lin"c-kit'° subset, because when mice are treated with estrogen, the precursors that produce N K and B cells are diminished from the Lin~c-kit'° fraction. When the remaining Lin"c-kit'° cells are cultured, the B and N K cell numbers are reduced. Most o f the progenitors for B and N K cells within the Lin"c-kit'° subset are i n separate populations though, because upon single cell assays most wells produced either B cells (1 o f 6-10) or N K cells (1 o f 19-25) but a bipotent B / N K cell precursor was rarer (1 o f 64-96). The Lin"c-kit'° population is heterogeneous with N K cell precursors and B cell precursors having substantial differences. For example, all B cell precursors are I L - 7 R a while only half o f N K cell precursors are. A l s o , N K cell precursors are +  more resistant to (5-estradiol hormone treatment than B c e l l s . A committed N K cell precursor 29  ( N K P ) was isolated in the adult B M . These cells are L i n " C D 1 2 2 N K l . l " D X 5 " . CD122 is II3 0  +  2 / I L - l 5RP and this is used by two cytokines, I L - 2 and IL-15 , both dependent on signaling through a third subunit, known as common y chain (yc) . These cells lack mature N K cell markers and are not functionally developed. N K P s produce mature N K cells v i a an in vitro culture system and they lack all other cell potential.  8  l _ j  B o n em a r r o w  n  Sca-1 c-kit  hi  Lin-  Lin-  hl  Figure 1.2. Progenitors with N K cell lineage potential. Multiple progenitors in the bone marrow and thymus have been shown to possess N K cell lineage potential.  NK cells may share a closer developmental relationship with T cells in fetal mice and B cells in adult mice. Numerous studies have shown that fetal T cell precursors possess N K lineage potential ( F i g . l .3). A bipotent T/NK cell precursor ( T N K P ) is present in the fetal liver, spleen, and blood. The earliest T N K P is B 2 2 0 c - k i t C D 1 9 - and is found in the fetal liver ( F L ) . They lo  +  33  reconstitute the T and N K cell compartments upon transplantation into mice lacking T, B and NK cells. These cells have been shown to be truly bipotent, because a single cell gives rise to both T and N K cells in a fetal thymic organ culture ( F T O C ) . This bipotent population represents 70% o f the cells that seed the thymus . 33  The earliest prethymic T cell progenitors  in the fetal b l o o d and the first cells to colonize the fetal thymus 34  35  both have T, NK, and  dendritic cell potential. O f 40 cells examined from the fetal thymic anlage, 7 cells gave rise to  9  T cells and all o f these 7 cells also gave rise to N K cells. Four o f these cells produced dendritic cells as w e l l . Subpopulations o f immature CD4"CD8" ( D N ) thymocytes in adult 3 5  mice also possess N K cell potential. N K cell potential o f D N 1 thymocytes was also shown by Balciunaite et al? . 6  Adult D N 1 c-kit cells, as well as D N 2 cells, upon culturing on O P 9 cells +  developed into functional N K cells. Limiting dilution analysis showed that 1 in 15 D N 1 cells and 1 in 7 D N 2 cells developed into N K c e l l s . Taken together, these studies suggest a link 36  between T cells and N K cells. This notion has further been supported by a transgenic mouse model o f human C D 3 e  3 7  which displays a block in development o f N K and T cells but not B  cells. M i c e transgenic for FceRIy exhibit a similar phenotype  Fetal blood  Fetal liver  38  .  Fetal  thymus  (TNKP) B22QI0 c-kit CD19+  Line-kit*  IL-7Fr  CD44 CD25FcR-  CD44* CD25-  +  IL-2RP*  Figure 1.3. Bipotent T / N K cell progenitors. T N K P cells are present i n the fetal liver, blood, and thymus. Figure is revised from that designed by Tabatabaei .  1.4. N K cell development  M u c h about the N K cell commitment pathway was first determined from i n vitro cultures. Nearly a decade ago, an i n vitro culture system was used to demonstrate that there are two major stages i n N K cell development. The first stage involves acquisition o f C D 122, which marks commitment to the N K cell lineage and allows the cells to become responsive to IL-15 which is crucial for the second stage o f N K cell maturation. W h e n Williams et al.  40  cultured  multipotent progenitors with IL-15 alone there was no cell expansion whereas when the cells  10  were cultured in IL-6, IL-7, S C F , and flt3L substantial expansion occurred but little to no N K cells were produced. But i f these cells were first cultured i n the cytokine cocktail (where they acquired C D 122 and became IL-15-responsive) and then switched to a culture with IL-15 alone, cells expanded well and almost all became N K 1 . 1 and had lytic capabilities. There is +  also a critical requirement for stromal cells in N K cell development. Cells cannot express Ly49 receptors without stroma cells i n the culture . 41  1.4.1. Stages of N K cell Development: from N K P to the mature N K cell  Recently K i m et al.  10  defined a developmental model of N K cell maturation that divides their  in vivo development i n the B M into five stages (Fig. 1.4). These stages are defined by the cells' functional characteristics and expression o f molecules which may not provide specific functional significance but work for defining specific stages o f maturity in N K cells. Developmental stage I represent the N K P s which are defined as CDlZZ^TSfKl.rDXS". Vossenrich et al.  42  have further characterized this population and have shown that a subset of  N K P s express C D 127 (IL-7Ra), IL-21R, CD117(c-Kit), and C D 135 (Flk2), but are CD25 (IL2Ra) negative. These markers are characteristic o f early hematopoietic precursors but C D 122 is not. In addition, they found that N K P s express N K G 2 D , while earlier studies have seen this at stage II. Stage II o f development involves acquisition o f the p a n - N K cell markers; high N K 1 . 1 expression and l o w D X 5 expression. The cells also express their first functional receptors, C D 9 4 / N K G 2 and, as reported by K i m et al. , N K G 2 D . This stage o f development 10  resembles mature fetal/neonatal N K cells which do not express L y 4 9 receptors. Stage II cells also express the integrin a v and are M a c - l C D 4 3 c - k i t " . The profile o f Stage III N K cells is +  lo  lo  similar to that o f the previous stage except that they are now c-kit and they begin to acquire +  11  the Ly49 molecules. D X 5 is upregulated in stage I V and the N K cells undergo a substantial proliferation in the bone marrow. A t stage V , N K cells complete their maturation by highly expressing Mac-1 and C D 4 3 , and proliferation is slowed until N K cells are stimulated by pathogens . These mature N K cells have full functional capabilities as well. The N K cell 10  maturation status varies in the peripheral tissues. Mature N K cells in the periphery express DX5 and CD 1 l b and lack CD51 and CD117. Expression o f K L R G 1 , a killer cell lectin-like receptor, is restricted to these most mature cells. Therefore, D X 5 C D 1 l b K L R G l C D 5 1 " +  +  +  CD 117" NK cell population is the most mature phenotype i n the periphery .  Stage  1  Stage  Stage III II MHC class I receptor acquisition  Stage  N K P  CD94/ NKG2  C D 1 2 2 + C D 1 2 2 NK1.1* NK1.1DX5'° DX5C D 9 4 / N K G 2 N K G 2 D Mac-1 a CD43' h i  v  0  CD94/ NKG2  Ly49  Ly49  (vC#)  IFNy  IFNy  C D 1 2 2 NK1.1* DX5'° C D 9 4 / N K G 2 Mac-1 a CD43 c-kit Ly49  +  1 0  h i  v  1  0  +  +  C D 1 2 2 NK1.1* DX5 C D 9 4 / N K G 2 Mac-1 a '° CD43 c-kit Ly49 Cytotoxicity IFNy +  +  +  i 0  V  Functionally mature  Expansion CD94/  CD94/ NKG2  ® &  Stage  IV  h  +  +  +  i  h  +  1 0  +  h  0  l  + A  +  +  +  1 0  i  h l  v  1  C D 1 2 2 N K 1 . 1 DX5 C D 9 4 / N K G 2 Mac-1 CD43 c-kit Ly49 Cytotoxicity I F N y ' h i  1 0  h  i  Figure 1.4. Stages of N K cell maturation. N K cells proceed through multiple stages following commitment to the N K cell lineage until they are fully mature. Each stage is marked by changes in receptor expression, proliferation and functional capabilities. Figure is from Veinotte et a l . 8  Mature N K cells express a full repertoire o f receptors and have the functional capacity to k i l l target cells and produce cytokines. N K cells are functionally mature when they can distinguish between s e l f - M H C class I and self class I" (expressing allogeneic class I or no class I) cells, +  rather than just when the cells acquire cytotoxicity potential. Mature N K cells are relatively static in terms o f self renewal and proliferative capacity, except after pathogen infection.  1.4.2. Factors involved in N K P production: Cytokines  To delineate the crucial cytokines for N K cell commitment, Williams et al.  41  cultured  multipotent progenitors with various combinations o f IL-7, S C F , and F l t 3 L . TL-7 and Flt3L combination yielded the highest proportion o f N K P s . Although addition o f S C F did not change N K P numbers, it did increase N K cell yield. Therefore, I L - 7 and F l t 3 L expand or induce N K P s , and S C F has an additive effect to the overall N K cell numbers. The I L - 7 R a chain is expressed by some N K P s  3 0  and splenic N K cells are reduced 3-fold in I L - 7 " m i c e , 7  44  whereas normal numbers o f N K cells are produced in IL-7Ra" m i c e . Since IL-7 only A  minimally affects N K cell development, Williams et al.  41  45  further concentrated on Flt3L and its  role in N K cell development. F l t 3 L progenitors produce more N K cells than Flt3L" +  progenitors. It seems that Flt3L either induces N K P s or more likely expands the N K P population, because N K P cell numbers are only slightly reduced in Flt3L" m i c e . F l t 3 L " A  46  v  mice have a 5.3-fold reduction o f mature N K cells and a loss o f cytotoxicity against Y A C - 1 . 4 7  A s for a role for S C F , when c-kit" (receptor for S C F ) progenitors were transplanted into /_  RAG2/y " " mice, they produced N K cells at lower levels; absolute cell numbers were 40% o f /  c  AO  that o f wild type progenitors. Their lytic capacity was greatly reduced . N K P cell numbers are not affected in c-kit " mice though so it appears that it is important for survival and _/  13  proliferation o f committed N K c e l l s ' . Individually, these pathways are not essential for the 25  26  first step o f N K cell lineage commitment but there may be redundancy i n the system and analysis o f the double-mutant mice would be informative.  Even though early studies suggested that signaling via the common y chain (yc) is essential for N K cell commitment since defects i n the gene coding for yc (U2rg) cause an X-linked severe combined immune deficiency characterized by several immune defects, including an absence of mature N K cells and IL15 " mice have reduced N K cell numbers " , further investigations _/  4 9  5 2  showed that the commitment to the N K lineage does not occur through yc-dependent cytokine stimulation , indicating that factors other than early^acting cytokines or yc-cytokines lead to 42  N K cell lineage commitment. Recent studies indicates the expression N K G 2 D at the N K P cell stage but it has yet to be clarified whether it plays a role in restriction o f the progenitors to 42  the N K cell lineage.  1.4.3. Factors involved in N K P production: Transcription factors  There are some factors that indirectly affect the production o f N K P cells probably by affecting earlier upstream progenitors. One example is Ikaros, a zinc-finger D N A binding protein that is present in H S C s and is up-regulated during lymphocyte differentiation. A n Ikaros null 53  mutation in mice causes a severe defect in the lymphoid compartment including N K cells . Analyses o f B M hematopoietic progenitors from Ikaros mutant mice revealed very low expression o f Flt3 and c-Kit receptors in these progenitors , which suggested that Ikaros might 54  regulate E L P / C L P homeostasis through growth factor receptors, thereby affecting the development o f all lymphoid lineages including N K P s . Another factor is P U . l , an Ets family 14  transcription factor, which is highly expressed in most hematopoietic cells and regulates the development o f myeloid and lymphoid lineages . P U . l and G A T A 1 have a mutually 55  antagonistic relationship because they bind each other i n the cytoplasm. This causes them to lose their translocation and binding ability respectively. The levels o f these two transcription factors mediate the decision between lymphoid and myeloid lineages. H i g h P U . l induces myeloid genes and high G A T A - 1 promotes erythroid and megakaryocyte genes. A t low P U . l levels, lymphocyte lineages are permitted. This is likely because only at l o w levels can P U . 1 induce IL-7Roc expression which promotes survival and proliferation o f T and B cells . P U . l 56  knockouts do show defects i n N K cells, but they are less severe than T and B cell defects, and this is likely due to N K cells' reduced dependency on I L - 7 . When transferred to Rag2" yc' " 57  A  7  mice, P U . 1-deficient fetal liver cells generate normal numbers o f H S C s , but produce reduced numbers o f N K P s and N K c e l l s . Although the N K P cell numbers were affected, the 57  phenotype o f mature N K cells was not substantially different but the deficiency does affect cell expansion and homeostasis.  58  E t s - l , a winged helix-turn-helix transcription factor is essential for development of N K cells . In Ets-l" " mice, T and B cells develop normally but the number o f D X 5 C D 3 " splenic N K cells 7  +  and their cytolytic activity is remarkably reduced. The defects i n Ets-l" " N K cells are intrinsic 7  to N K precursors. One mechanism by which E t s - l controls N K cell development might be through survival, as the increased apoptosis has been reported i n B - and T-cell populations from Ets-1" " m i c e . /  59  A class o f b H L H proteins known as inhibitors o f D N A binding (Id) proteins are key molecules for N K cell development. They negatively regulate E proteins that are crucial for B and T cell  15  lineages. Overexpression o f Id3 in human C D 3 4 H S C s has been shown to block T cell +  development and promotes N K cell development in fetal thymic organ cultures . Development 60  o f N K cells is also accelerated when both Id2 and Id3 are constitutively expressed i n C D 3 4  +  H S C s , whereas the generation o f T, B , and lymphoid-DCs is completely abrogated . These 61  observations were further complemented by the finding that targeted disruption o f Id2 gene results in a selective block i n N K cell development , accompanied by a lack o f N K P s . These 62  6 3  data present a model in which Id proteins promote commitment o f common lymphoid precursors to N K P s while inhibiting their option to develop to committed T- and B progenitors.  1.5. N K cell maturation  1.5.1. M H C Class I Receptor acquisition  The M H C - c l a s s I receptor acquisition on N K cells is complex. Although the numbers o f L y 4 9 and C D 9 4 / N K G 2 receptors is quite l o w in comparison to T and B cell receptor diversity, the repertoire o f the N K cell population, as a whole, is very diverse. Each L y 4 9 or C D 9 4 / N K G 2 receptor family member is expressed on a subset o f N K c e l l s . Therefore, their expression 64  partially overlaps with the expression o f the other receptors. N K cells express at least one M H C I-specific receptor and individual N K cells can coexpress up to five different Ly49 receptors in addition to C D 9 4 / N K G 2 receptors . Expression o f the receptors is not co64  regulated and it appears that they are expressed with a great degree o f independence from each other . 8  16  The receptor expression on developing N K cells differs between the fetal and neonatal mouse and the adult mouse. M o s t fetal and neonatal N K cells express C D 9 4 / N K G 2 but not L y 4 9 receptors, with the exception o f L y 4 9 E , which is expressed i n the fetal cells but not the adult cells . A s the mouse matures, N K cells begin to acquire L y 4 9 receptors, while C D 9 4 / N K G 2 65  +  N K cells decline. The frequency o f N K cells expressing high levels o f C D 9 4 / N K G 2 decreases from - 9 0 % on neonatal N K cells to - 5 0 % on adult N K c e l l s . This decease i n C D 9 4 / N K G 2 66  expression coincides with the increase in the frequencies o f N K cells expressing Ly49. Ly49 receptor expression begins very gradually, starting at approximately 1 week after birth and reaching adult levels at 6 to 8 weeks o f l i f e . The decease o f C D 9 4 / N K G 2 is not due to a 67  downregulation i n cells but rather due to a generation o f C D 9 4 / N K G 2 X y 4 9 N K cells in +  young m i c e . 67  N K cells appear to undergo a process o f ensuring self tolerance during development. Since many normal self cells express ligands for N K cells, it is important that N K cells express self M H C class I inhibitory receptors to ensure that autoreactivity does not occur. Evidence shows that N K cells that lack inhibitory receptors for self M H C class I are hyporesponsive. Two main theories exist on how this self tolerance is acquired by N K cells during development (reviewed in ) . The first is the arming model. In this model, N K cells receive positive signals 7  i f their L y 4 9 receptors bind self M H C class I. This positive signal 'arms' the N K cells and allows them to mature into functional N K cells while those that do not express a self L y 4 9 remain unarmed and unresponsive. The second model is the disarming model. In this model N K cells express a variety o f stimulatory and inhibitory signals and interact with self cells. If the N K cell binds both activating ligands and inhibitory ligands (the self M H C class I) no net stimulation occurs and the responsiveness o f the N K cell is maintained. Conversely, i f an N K  17  cell without a self L y 4 9 binds a self cell, it w i l l only receive activating signals and hyporesponsiveness w i l l be induced i n this cell. In other words, the cell is 'disarmed'.  1.5.2. Factors involved in N K cell maturation  N K cell maturation requires cell-to-cell interactions and soluble factors derived from B M l  stromal cells. In vitro culture systems that are designed to give rise to mature N K cells, as described earlier, are not complete without stromal cells in the culture . Multipotent H S C s can 41  be cultured with certain cytokines to produce C D 1 2 2 N K P s but without addition o f IL-15 and +  stromal cells in a second stage of N K cell culture, the cells w i l l not fully mature. If IL-15 is added without stromal cells, the cells become functionally mature and express C D 9 4 / N K G 2 but not L y - 4 9 . This shows the importance o f cell-cell interactions between developing N K 41  cells and stromal cells.  IL-15 is the critical cytokine, or the "essential fuel" for N K cell development. IL-15 works by binding the IL-15R complex, which is made up of IL-2/IL-15RP ( C D 122), IL-15Ra, and the y c •  68  but it appears that there may be other methods in which IL-15 signaling occurs  . There is an  N K cell deficiency, as well as deficiencies in N K T cells, intestinal epithelial cells, and memory C D 8 T cells, i n IL-15 and IL-15Ra-deficient m i c e ' . Disrupting y c , the third subunit o f IL+  4 9  5 2  15R, generates a similar defect in N K cell production, suggesting an essential role for IL-15, overall, in N K cell development ' . In an effort to identify the developmental stage at which 69  10  IL-15 and other y cytokines are important for N K cell development, Vossehenrich et al.  42  c  produced various y c deficient mice (IL-2, IL-4, IL-7, IL-15) including multiple y deficiencies c  per mouse (i.e. I L - 2 l L - 7 " " , IL-2" "IL-4" "IL-7" ", and IL-4" "IL-7" "IL-15" ") on a Rag2-deficient v  /  /  /  /  /  18  /  /  background. They were able to show that N K P s do not require yc cytokines because they are present in normal numbers and IL-2, IL-4, and IL-7 have no role in the generation o f mature N K cells. Conversely, there is an essential singular role for IL-15 in the generation of immature and mature N K cells in the B M and mature N K cells in the spleen. The N K cells that do remain in IL-15-Rag2 deficient mice have an immature phenotype o f Mac-1 CD43". The lo  cells were also C D 9 4 , N K G 2 A / C / E and L y 4 9 D b u t L y 4 9 G 2 and L y 4 9 C / I N K cell numbers +  were reduced. The cells could k i l l Y A C - 1 and produce IFNy upon stimulation but at reduced levels . One clue to IL-15 signaling function is the antiapoptotic factor B c l - 2 . Transgenic 42  71  overexpression o f B c l - 2 in IL-2RP-deficient mice leads to normal N K cell numbers . It also allows peripheral N K cells to persist in IL-15 deficient hosts after adoptive transfer '  . A  recent study has isolated a downstream target o f IL-15 in human N K cells and therefore provides another new clue to its function . Because IL-2 and IL-15 signal through common receptor subunits, it is likely that they regulate a shared set o f downstream target genes. They showed that IL-2 and IL-15 stimulation results i n the post-transcriptional increase i n Ets-1 protein. The exact role for Ets-1 i n N K cell function and the particular target genes controlled by Ets-1 are still unknown . These studies still have not revealed the exact role o f IL-15 in N K cell development and many possibilities exist. To summarize, IL-15 may promote N K cell development by indirectly controlling the responsiveness o f maturing N K cells to other growth and survival factors. It may also act as a survival factor for N K cells at certain developmental stages or it may act to support proliferation o f developing N K cells. A l s o , IL-15 may have an indirect effect on N K cells by regulating their susceptibility to the IL-21 maturation factor . 42  Factors that regulate the IL-15 signaling pathway and its effects on N K cells have been identified. These include lymphotoxin and IRF-1. The lymphotoxin L T a ^ i s expressed on N K  19  cells and it binds the L T p receptor on non-lymphoid cells. A reduction o f N K cell numbers is seen in mice lacking the L T a on N K cells or L T p R on stromal c e l l s ' . The impaired N K cell 74  75  development is due to the absence o f LT|3R triggered signals in the stromal cells. F L cells from w i l d type mice cannot develop into N K cells when grown in vitro on L T p R deficient stromal cells. If the reverse culture is performed, LTa-deficient F L cells can differentiate into N K cells when grown on w i l d type stroma . Lian et al? recently suggested that this impaired 76  1  N K cell development is a result o f impaired IL-15 production i n the stromal cells. The signals derived from L T and L T p R interaction result in the up-regulation o f IL-15 production by the B M stroma . IRF-1 deficient mice also have severely reduced numbers o f N K cells due to a 77  78 79  defect in the production o f IL-15 by the bone marrow microenvironment '  .  A very recent study has made a breakthrough in determining factors required for L y 4 9 receptor expression and N K cell maturation . Tyro3 receptors' ( A x l , Tyro3, Mertk) expression on 80  committed N K cell precursors is crucial for both the acquisition o f nearly all inhibitory and activating receptors on N K precursors i n the B M and for the functional maturation o f these cells in the spleen. N K cells express all three receptors and B M stroma expresses their two ligands (Gas6 and protein S ) . They seem to be required for progression from the immature 8 0  N K cell stage because N K cells in Tyro3-deficient mice have high N K G 2 and C D 9 4 expression and l o w L y 4 9 , D X 5 , N K G 2 D , M a c - 1 , CD43 expression. Their cytotoxicity is substantially decreased and the cells cannot produce IFNy either. The Tyro3 receptors seem to act through an uncharacterized pathway because they do not alter expression o f known factors SO  that play roles i n N K cell development .  20  Other factors, besides those that play a role in the B M stroma signals, are important for N K cell maturation and/or effector function. For example, V D U P 1 (Vitamin D 3 upregulated protein 1)" mice show a profound reduction in the number o f mature N K cells i n the spleen A  and B M  . Expression o f C D 122 (the marker o f N K P s ) was reduced and V D U P 1 expression is  known to be induced from the N K P stage. Therefore V D U P 1 may be important for the differentiation o f N K P s or for their maturation because upregulated V D U P 1 may induce cell cycle arrest, which is necessary for the onset o f differentiation . The G A T A - 3 transcription 81  factor may also play an important role in the transition from the immature M a c - l ' ° C D 4 3 to the M a c - l C D 4 3 h l  h l  7  l0  NF-KB  stage  stage. Peripheral N K cells that develop i n G A T A - 3 " " mice resemble  immature M a c - l ' ° C D 4 3 N K c e l l s . A third factor that controls N K cell maturation is The  l0  82  NF-KB.  family members normally are kept in an inactive state by inhibitors but i f these  inhibitors are degraded the  NF-KB  transcription factors become a c t i v e  83,84  . When the  NF-KB  transcription factor family members are hyperactivated, N K cell maturation is arrested and the residual N K cells i n the B M , spleen, and liver are immature (Mac-1 C D 4 3 ) . N K cell lo  l0  cytotoxicity is not affected but there is reduced capacity for IFN-y production. The defect in maturation is likely a result o f N K cells being unable to provide adequate proliferative signals. The cells express normal levels o f IL-2RP and y but that they could not proliferate in the c  presence o f IL-2 or IL-15. The molecular mechanism for this defect is unclear .  T-bet is a transcription factor that regulates N K cell maturation and effector functions (cytokine production and to a lesser extent, cytotoxicity). In T-bet deficient mice there is a reduction in N K cell number, specifically in cells with late maturation markers. The perforin and granzyme B genes are T-bet target genes even though T-bet only plays a minor role in controlling cytotoxicity . There may be a compensatory relationship with T-bet. E O M E S , 8 6  21  which is also a T-box family transcription factor, is highly expressed i n N K cells (even in Tbet" " mice) and it regulates IFNy, perforin, and granzyme expression . It w i l l be informative to 7  87  study these transcription factors' roles in EOMES^T-bet" " mice. G A T A 3 participates with T7  bet in N K cell function as well. There is an intrinsic defect in IFN-y production in G A T A - 3 " _/  go  N K c e l l s . In these mice, T-bet expression was 4-fold reduced while H l x (another 00  transcription factor that controls IFNy production) expression was 10-fold reduced. T-bet establishes an active chromatin configuration of the IFN-y locus and it cooperates with H l x , which is broadly expressed in multiple hematopoietic cell lineages, including N K cells, to upregulate IFN-y e x p r e s s i o n ' . The combined deficiency in T-bet and H l x expression may 89  90  explain the poor inducible IFN-y production in G A T A 3 " " mice .  A few other factors that are critical for N K cell effector function rather than maturation are M E F - 1 , C / E B P y , M I T F and N E M O . A member o f the ETS-family, M E F - 1 , is required for N K and N K T cell development as a profound reduction o f these cells is observed in MEF-deficient mice. The few N K cells found in these mice are functionally impaired because M E F regulates transcription o f the perforin gene . MITF-deficient mice have a cytotoxicity defect due to 82  decreased perforin expression  91  while NEMO-deficient N K cells have a defect in their  NF-KB  signaling pathway which is important in the regulation o f perforin and other cytotoxic factors . 92  The basic leucine zipper transcription factor, C / E B P y , is also important for proper  N K cell cytotoxicity and IFNy production but not for earlier N K cell development. C / E B P y expression is ubiquitous and constitutive and while it does not have a transcription activating domain, it can interact with other transcription factors and augment their D N A binding. In C/EBPy"" mice, N K cell cytotoxic activity and IFNy production are impaired .  22  A cytokine that plays a late role in N K maturation is IL-21. IL-21 induces further differentiation o f activated, mature N K cells and is an initiator o f IFN-y production as well. The terminal differentiation o f mouse N K cells leads to an increase i n cell size and granularity, loss o f N K 1 . 1 expression, and upregulation o f the N K G 2 - C D 9 4 complex. It also leads to a significant increase i n cytotoxicity and a massive induction o f cytokine secretion. This late function o f IL-21 is important for activated N K cells because it enhances IFN-y production more than IL-15 alone . 94  In conclusion, the precise pathway o f N K cell lineage commitment and subsequent maturation is still unclear. The stages o f N K cell development are not yet fully characterized but we are reaching a point now where an accepted model o f N K cell development is present and being built upon. Future studies w i l l contribute to this model by characterizing more cell surface markers and molecular pathways which contribute to either the commitment or the maturation of N K cells.  Table 1.1. Genetic mutations affecting the maturation o f N K cells. Revised from DiSanto . Transcription factor PU.1  Phenotype  Repertoire  Stage  Cytotoxicity  Cytokines  Ref.  Normal  Decreased Ly49  NKP  Nomal  Normal  bl  Ets-1  ? ?  ? ?  NKP  Reduced  Reduced  NKP  Reduced  Reduced  Decreased Ly49 Normal  Immature NK Mature NK Late mature NK  Normal  Reduced  88  Normal  Normal  Bb  Reduced  Normal  82, 91, 93  Id2 Gata-3  Mac-1 , CD43 Mac-1 , CD43 Normal 10  10  T-bet  10  10  MEF, MITF, C/EBPy  Normal  23  ba  1.6. T cells a n d development  The T cell population is able to mount a response to virtually any foreign antigen because each T cell expresses a unique variant o f heterodimeric receptors (a.p or y 8 ) . These T cell 95  receptors (TCRs) are formed from rearrangement and co-expression o f either a and P genes or y and 8 genes. aP T cells are part o f the adaptive immune system. They localize primarily in secondary lymphoid organs and respond to infection by facilitating the production of antibodies and by lysing target cells. aP T cells recognize peptide ligands presented by class I and II M H C molecules. y8 T cells participate in the early immune response, similar to the innate immune system . They comprise only a minor population i n the blood (1-5%) but in 96  epithelial tissues, they represent 50% o f the T cells. They recognize a much wider variety o f antigens such as nonclassical M H C molecules, heat shock proteins, and l i p i d s . The structure 96  and signaling potential o f the y 8 T C R complex differs slightly from the ocpTCR complex. Both complexes include invariant accessory chains such as the C D 3 proteins. While in the a P T C R complex there is one heterodimer o f CD3e and 8 and another heterodimer o f CD3e and y, the y S T C R complex lacks C D 3 8 and has two heterodimers o f C D 3 s and y . The signal transduction by the y 8 T C R complex is superior to that o f the a P T C R complex .  T cell development occurs in the thymus and T cell precursors migrate from the fetal liver and the adult B M to continuously seed the thymus. The first half o f T cell development is independent o f the T C R while the second half is T C R dependent (Fig. 1.5). T cell development can be defined by a series o f stages (reviewed by Janeway ). The initial stages 95  are traditionally termed double negative since they do not express C D 4 or C D 8 but they are also referred to as triple negative to include C D 3 as well. The initial precursor stage is termed 24  double negative 1 ( D N 1 ) and the cells are C D 4 4 C D 2 5 " . The D N 2 stage is marked by gain o f +  C D 2 5 and the initiation o f T C R gene rearrangement. +  This rearrangement continues in CD44"  C D 2 5 cells at the D N 3 stage. T cell development then becomes T C R dependent and cells die +  i f they fail to generate a productive in-frame T C R P chain or a pair o f T C R y and T C R 8 chains. Because the joining events o f V ( D ) J recombination are imprecise, two out o f three attempts are non-productive and fail to maintain the translational reading frame o f the T C R subunit. Cells that do successfully rearrange and express a productive T C R chain(s) on their surface w i l l then continue differentiation to either the y8 lineage or the otp lineage. H o w the T cell progenitors decide to follow the y8 T cell lineage versus the aP T cell lineage is not yet known. One OR  model suggests that it depends on the strength o f the signal . If a weak signal is received from either a pre-TCR (the T C R p chain is expressed on the surface at this stage with an invariant p T a chain) or a y8 T C R that has not encountered ligand, the cells w i l l become ocP T cells. Conversely, i f a stronger signal is received from a yS T C R that has encountered a ligand o f intermediate affinity, cells w i l l become y8 T cells. Finally, i f a very strong signal is received OR  from a y8 T C R and ligand, the cells w i l l die . aP T cells continue their development by undergoing proliferation and differentiate into C D 4 C D 8 double positive (DP) aP T cells . +  +  95  These od3 T cells then rearrange their T C R a genes and undergo positive and negative selection for self/nonself discrimination. T C R s must be able to recognize M H C class I or II on cortical thymic epithelial cells to pass positive selection and become C D 8 SP cells or C D 4 SP cell respectively. If SP cells recognize a self antigen presented by B M D C s or macrophages in the thymus, they are killed by negative selection to ensure that self-reactive cells do not leave the thymus. After selection and maturation, T cells finally exit the thymus to take up their roles in the immune system . 95  25  Natural killer T ( N K T ) cells are a conserved T cell sublineage with unique properties (reviewed by Kronenberg"). The main subpopulation o f N K T cells are invariant N K T cells, which likely arise from D P T cells during development in the thymus (Fig. 1.5). These cells express invariant T C R s and they recognize a synthetic glycolipid presented by an M H C class Ilike molecule, C D Id. U p o n stimulation, N K T cells rapidly produce many cytokines and can influence diverse immune responses. Another subset o f N K T cells is C D l d independent and their developmental pathway is not defined.  Figure 1 . 5 . Stages of T cell development. The initial stages are T C R independent and involve V ( D ) J recombination o f T C R y , 5, and p genes as well as loss o f alternate cell potential and final commitment to the T cell lineage at the D N 3 stage. The remainder o f the stages are T C R dependent and result in mature single positive C D 4 and C D 8 T cells.  Two groups characterized the kinetics and timing o f T C R recombination during T cell development ' . T C R y , 5, and P recombination occurs before T C R a . T C R rearrangements 100  101  26  at the D N 1 stage are negligible and rearrangements that were detected are likely contamination. A t the D N 2 stage, both T C R y and T C R 5 rearrangements are present while T C R p rearrangements are almost absent. Specifically for T C R y , V y 2 - J y l rearrangements were more abundant than V y 5 - J y l and by the D N 3 stage, the level o f V y 2 - J y l rearrangements were within range o f mature y8 T cells while V y 5 - J y l levels were still lower. For T C R 5 , Capone et al.  100  detected V55-J51 and V54-J81 rearrangements at D N 2 stage but V55 rearrangements are  more prominent at this stage. Livak et al.  101  saw that D N 2 cells have substantial amounts o f  partial D51 and D82 to J81 rearrangements but not V 8 - D J 8 and that by D N 3 75-100% have completed V - D J 8 rearrangements where only a minority o f cells have full  V-DJP  rearrangements. A l l major y and 8 genes are rearranged at D N 3 or D N 4 at levels similar to total thymocytes. Therefore, the recombination o f the majority o f T C R y and T C R 8 loci is completed by D N 3 but completion of  V-DJP rearrangement to maximal  levels is not seen until  more advanced stages. A l s o , rearrangement of the T C R y and T C R 8 loci occurs in most a p T cells  102  "  107  and productive T C R p rearrangements can be found in y8 T c e l l s ' 104  1 0 8  " . m  1.7. TCR rearrangement: V ( D ) J recombination  The ability o f T cells to respond to a vast array of antigens is dependent on the generation o f unique surface receptors with diverse binding specificities. The T C R s are assembled during lymphocyte development from germline variable (V), diversity (D), and joining (J) gene 117  segments by the process o f V ( D ) J recombination (Fig. 1. 6) (reviewed by  ). The segments  are flanked by recombination signal sequences (RSSs). These conserved noncoding sequences guide the D N A rearrangements. The R S S includes a heptamer sequence which is always 27  contiguous with the coding sequence. Next is a nonconserved spacer region that is either 12 or 23 bases long, which is followed by a second conserved nonamer sequence. Therefore the heptamer-spacer-nonamer sequence motif makes up the R S S and it is always directly adjacent to the coding sequence o f the V , D , or J gene segment. Normally, a 12-RSS is joined with a 23-RSS. The complex o f enzymes that carries out recombination is collectively called V ( D ) J recombinase. This contains lymphocyte specific enzymes, R A G - 1 and R A G - 2 , and ubiquitously expressed D N A - m o d i f y i n g proteins. T w o R A G complexes recognize and bind two R S S sequences. The R A G complexes then bind each other and therefore bring together the two gene segments to be joined. The R A G complexes each introduce a single strand nick precisely between the R S S and the coding sequence. This leaves a free 3' O H group which attacks the phosphodiester bond on the other strand, creating a hairpin at the end o f the gene coding segment. This process simultaneously creates a double-stranded break at the ends o f the heptamer sequences. The two non-coding R S S s are joined i n a precise head to head linkage to form a signal joint. The coding joint formation involves extra steps. The R A G complex remains bound to the hairpin structures and proteins i n the complex cleave the hairpins at random. The D N A repair enzymes i n the complex may remove some nucleotides while the terminal deoxyribonuclease transferase (TdT) enzyme adds nucleotides randomly. Finally, D N A ligase I V joins the ends together and repair enzymes trim off non-matching bases and synthesize complementary bases to fill in the remaining single stranded D N A . Since the number o f nucleotides added by T d T is random, the added nucleotides often disrupt the reading frame o f the coding sequence. This produces a nonproductive rearrangement, which occurs in 2 / 3  rds  o f rearrangements, as mentioned above.  28  3 _J_J  A  J xJ)  I  Coding joint  Proteins  Signal joint  Figure 1.6. The process of V(D)J recombination. 1. T w o R A G complexes recognize and bind two R S S sequences, one on the V segment and one on the J segment. 2. The R A G complexes then bind each other and therefore bring together the two gene segments to be joined. 3. The R A G complexes cleave the D N A to create hairpin ends on the V and J segments. Other proteins bind the hairpins and the cleaved R S S ends o f the gene coding segment. 4. The D N A hairpins are cleaved at random. Additional bases are added (TdT) or subtracted (exonuclease) to create imprecise ends. 5. D N A ligase I V joins the ends o f the segments to form the coding joint. It also joins the RSSs to form a signal joint.  The process o f V ( D ) J recombination is regulated by the availability o f the recombination machinery and also by accessibility o f the target gene segments  . Multiple epigenetic  mechanisms are involved in regulating rearrangements including: histone acetylation, D N A methylation, allelic exclusion, nuclear location, and cis- and trans- acting factors . 113  Specifically for T C R y recombination, IL-7 regulates chromatin accessibility for germline expression and R A G enzymes via histone acetylation which opens the chromatin structure " 114  116  1 . 8 . T C R y locus and rearrangement patterns  The T C R y locus in mice features four clusters o f V y , Jy, and Cy regions containing 7 V y segments, 4 Jy segments and Cy regions (Fig. 2 . 1 ) 29  103  '  117  . Each cluster contains one C region,  one J segment and one to four V segments. The V segments rearrange to the J segment in the same cluster. Cluster 3 has been deleted in most T C R y haplotypes and is believed to be nonfunctional in some s t r a i n s  1 0 2  ' . The C y l cluster contains four closely linked but distantly m  related V region genes.  The utilization o f V y genes is developmentally regulated such that Vy3 and Vy4 gene rearrangements are prominent i n the fetal thymus but are very rare i n the adult while V y 2 and Vy5 rearrangements have the opposite p a t t e r n  103  '  119  '  120  . This rearrangement pattern is  reflected in the y5 T cells present i n the mouse. V y 3 y8 T cells appear first at approximately +  embryonic day 13 and disappear from the thymus by embryonic day 1 8  121  . V y 4 y8 T cells are +  also seen i n the fetal m o u s e . Interestingly, the cells expressing different V y rearrangements 122  have specific functions as well. V y 3 y8 T cells home to epidermal epithelial tissues and are +  called dendritic epidermal T cells ( D E C s )  123  . These cells are unique i n that they secrete  keratinocyte growth factor . V y 4 yS T cells, on the other hand, concentrate i n the female 124  +  reproductive tract and tongue. These cells also have a fixed T C R 8 chain and therefore, the population only has one T C R specificity. These cells likely recognize common stress-induced self antigens and eliminate damaged cells as well as promote epithelial growth and differentiation . In the adult environment, y8 T cells rearrange V y 2 and Vy5 chains but they 123  are not as limited with T C R specificity since they can pair to different T C R 8 chains, to some extent, and they also have V ( D ) J junctional diversity, which is lacking i n the fetal cells. Therefore, these cells likely recognize foreign antigens  30  125  .  It appears that developmentally regulated V y gene recombination is an intrinsic genetically programmed process. In models where the T C R y locus contains frame shift mutations that prevent functional expression, the V y genes are still rearranged at the appropriate times even though they cannot influence the fate o f the cells  . For the fetal T C R genes (Vy3 and Vy4),  the location o f the V segment influences the rearrangement pattern. In a transgenic model where V y 2 and Vy3 are switched, fetal cells had higher V y 2 rearrangement levels than Vy3, the 12  opposite o f what normally occurs . V segment location does not affect the rearrangement pattern in adult thymocytes though and they are mostly influenced by promoters. The sequences upstream o f V y 2 , 3, and 4 are quite divergent, suggesting that each promoter could be regulated by distinct trans-acting factors. Baker et al.  126  swapped the promoter regions o f  Vy2 and Vy3 and the pattern was reversed in adult with Vy3 being rearranged. Since germline transcription is present for both V y 2 and Vy3 in the fetal stage, it seems that both genes are accessible in the early stages, with an advantage to the more proximal gene, followed by a stage when Vy3 is repressed and/or V y 2 is activated, leading to a strong preference for V y 2 rearrangement. Therefore, the "developmental switch" (i.e. Vy3,4 to Vy2,5) i n V y gene usage is imposed by two mechanisms, one sensitive to gene location and the other dependent on differential V-promoter activity.  During development, each precursor can try multiple rearrangements at the T C R y locus  . For  example when y5 T cells were sorted for V y 2 , Vy5 or V y l . l y8 T C R s on their cell surface +  (these three represent 90% o f the y8 thymocytes in adult B 6 mice), all clones had multiple rearrangements at the y locus. Most cells had 2 to 6 rearrangements and V y 2 cells normally +  31  had a maximum o f 4 rearrangements. Rearrangements involving J y l were found in all cells and about half had J y l rearrangements in both chromosomes.  Since multiple V - J rearrangements occur in a single cell there are several factors that ensure that y8 T cells w i l l not express more than one T C R specificity on the cell surface. These include: 1. the frequencies at which each V y and V 8 gene segments participate in recombination, 2. the frequencies at which the rearranged products produce a functional chain, and 3. whether the functionally rearranged T C R y and T C R 8 chains can pair to form a y8TCR  1 2 7  . The end result is the formation o f a pool o f y8 T cells expressing a diverse  repertoire o f V y and V 8 chains i n which the majority o f the cells bear a single T C R specificity at the cell surface. This works in these cells without specific checkpoints to test and control for the functionality o f each o f the rearranged chains, unlike a.p T cells. This is only possible because the three mechanisms work differently on each V gene segment. For example, V y 2 and V y l . 2 rearrange at similar frequencies but they are approximately -10-12 times more frequent than V y l . l rearrangements and 16-20 times more common than Vy5 rearrangements. V S segments also rearrange at different frequencies  .  The frequencies at which different V y and V 8 segments participate in recombination are inversely correlated with the apparent ability o f the resulting T C R y or T C R 8 chain to participate in the formation o f a functional y S T C R . For example, V y l . 2 is one o f the most frequent rearrangements but it shows the highest level o f restriction for pairing with a T C R 5 chain. In contrast, V y l . 1 and V y 5 , which are rearranged at l o w frequencies, lack restriction in pairing with V 8 chains.  32  Also, there is a stop codon at the 3'end o f the very frequently rearranged germline Vy2 gene 100 128  segment  '  . This substantially lowers the frequency at which rearrangement w i l l produce a  functional chain. This therefore lowers the frequency o f a functional V y 2 being expressed on the cell surface and therefore increases the chances that y8 T cells can express T C R y chains on their surface that are from l o w frequency rearrangements (i.e. Vy5).  1.9. Thesis objectives and hypotheses  The  original objective of my thesis was to determine i f adult and neonatal N K cells follow  separate pathways o f lineage commitment. Since adult and neonatal N K cells differ in their phenotype and function, I hypothesized that the differences are due to their developmental pathways. In Chapter 3, this hypothesis was tested by examining differences i n gene expression patterns between the two types o f N K cells. This study resulted i n the detection o f T C R y gene rearrangement in a subset o f N K cells. Based on this interesting and unexpected finding, my subsequent thesis research was redirected to further characterize N K cells that have rearranged T C R y genes. The  objectives of my subsequent study were to determine the  developmental pathway responsible for the generation o f N K cells with T C R y gene rearrangement and to further characterize these N K cells. M y hypotheses  were: 1) there are at  least two separate developmental pathways for N K cells, one i n the bone marrow and the other in the thymus, 2) a subset o f N K cells arises from immature T cell progenitors that have begun T C R y gene rearrangement but still retain N K cell potential, and 3) N K cells with rearranged T C R y genes are different from bone marrow-derived conventional N K cells i n phenotype and function.  To test these hypotheses, the following objectives were set: 1. to determine i f the  T C R y gene-rearranged N K cells are thymus-dependent, 2. to identify the progenitors that give 33  rise to the N K cells with T C R y gene rearrangement, 3. to determine i f N K cells generated by the developmental pathway involving T C R y gene rearrangement differ from conventional B M derived N K cells i n phenotype, function, and tissue distribution.  34  2 MATERIALS AND METHODS  2.1. M i c e . C 5 7 B L / 6 (B6) mice were bred in our animal facility. Adult mice used in this study were 6 to 10 weeks old and neonatal mice were 1 to 3 days old. R A G 2 " " H Y T C R transgenic /  mice were also bred from breeders purchased from Taconic Farms (Germantown, N Y ) . Adult TCRp7TCR6-double knockout (B6.\29P2-Tcrb"" Tcra " /J) IMom  imlMc  n  mice were purchased from  The Jackson Laboratories (Bar Harbour, M E ) . Heterozygous nude (B6.Cg-Foxnlnu) mice were purchased from The Jackson Laboratories and were mated, and athymic nude neonatal mice (3 days old) were used. IL-15" " mice ( C 5 7 B L / 6 N T a c - / I i 5 " " 7  Taconic Farms. ~NOD.Cg-Prkdc Il2rg"" /SzS scicl  7/TOC  ) were purchased from  mice were from The Jackson Laboratories.  IWjl  2.2. Antibodies. Antibodies used in this thesis are listed in Table 2.1.  35  Table 2.1. L i s t o f antibodies: Animal mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse  Isotype  PM homemade homemade PM  biotin FITC FITC biotin biotin biotin biotin biotin biotin biotin FITC FITC biotin  PM PM PM homemade Biolegend eBioscience PM homemade PM PM Boehringer PM PM  biotin FITC biotin biotin FITC FITC biotin  PM PM PM PM PM homemade homemade  Rat I g G , K  biotin biotin  PM homemade  Rat lgG , K  FITC  PM  biotin FITC biotin biotin PE PE FITC PE APC  homemade PM PM PM PM PM PM PM PM  Clone PK136 PK136 PK136 PK136 145-2C11  Ly-49 C and Ly-49 I Ly-49 D Ly-49 G2 Ly49 G2 NKG2D CD127 KLRG1 Mac-1  5E6 4E5 4D11 4D11 C7 A7R34 2F1  CD8a (LY-2) CD8b.2 (Ly-3.2) CD8 CD3e CD3e  53-6.7 53-5.8  mouse mouse mouse mouse mouse mouse mouse mouse mouse mouse  145-2C11 145-2C11  mouse mouse  Ar Ham I g G l , K  GL3 GL3 H57-597 PK136 PK136  mouse mouse mouse mouse mouse  Ar H a m I g G 2 , K  Ar Ham I g G 2 , XI  TER-119 Gr-1  Ter-119  mouse  B220  RA3-6B2  mouse  gamma-delta TCR gamma-delta TCR TCR beta chain NK1.1 NK1.1 Mac-1 Mac-1  B220 CD19 CD19 CD44 CD25 CD122 Pan-NK cells Pan-NK cells IFN-gamma  145-2C11 145-2C11 145-2C11 20d5 20d5 18d3 YE/148 1F8 5E6  1D3 1D3 1M7 3C7  TM-31 DX5 DX5 XMG1.2  Ms I g G , K 2 a  Ms I g G , K 2 a  Ms I g G , K 2 a  Ms I g G , K 2 a  Ar Ham I g G l , K Ar Ham I g G l , K Ar H a m I g G l , K Ar Ham I g G l , K • Rat I g G , K 2 a  Rat I g G , K 2 a  Rat I g G , K 2 a  Ms I g G , K 2 a  Ms I g G , K 2 a  Rat I g G , K 2 a  Rat I g G , K 2 a  Ar Ham IqG Rat I g G , K 2 a  Syr Ham I g G 2 , K  Rat I g G , K 2 a  Rat I g G K w  Ar Ham I g G l , K  Ar H a m I g G 2 , K  Ms I g G , K 2 a  Ms I g G , K 2 a  2 b  2a  mouse mouse mouse mouse mouse mouse mouse mouse  Rat I g G , K 2 a  Rat I g G , K 2 a  Rat I g G , K 2 b  Rat I g G , K 2 b  Rat I g G b , K 2  Rat I g M , K  36  Company PM PM PM PM PM PM PM PM PM PM  Conjugate APC PE FITC biotin FITC biotin PE PerCP-Cy5.5 biotin FITC biotin FITC FITC FITC  Antibody NK1.1 NK1.1 NK1.1 NK1.1 CD3e CD3e CD3e CD3e NKG2A/C/E NKG2A/C/E CD94 Ly-49A Ly49H Ly-49 C and Ly-49 I  Rat I g M ,  K  Rat I g d ,  K  2.3. Microarray Sample Preparation and Analysis. Total R N A was isolated by using the RNeasy M i n i K i t ( Q I A G E N Inc., Mississauga, O N ) . Double stranded c D N A was synthesized from total R N A with the Superscript double stranded c D N A kit (Invitrogen, Carlsbad, C A ) . The Enzo B i o A r r a y high yield R N A transcript labeling kit (Affymetrix Inc., Santa Clara, C A ) produced biotin labeled c R N A which was fragmented and hybridized to Affymetrix GeneChip Mouse Genome U 7 4 A v 2 arrays. The first two microarray experiments were performed at the D N A Array Laboratory, Wine Research Centre, University o f British Columbia and the third experiment was performed at the Affymetrix GeneChip Facility at the Michael Smith Genome Sciences Centre, British Columbia Cancer Agency. A l l data analysis was performed with Genespring version 7 (Silicon Genetics, Redwood City, C A ) . Expression values were background corrected, normalized, and summarized by using the default settings o f the program package.  2.4. Measuring D N A . For determining the percentage o f N K cells that had T C R y gene rearrangements, D N A templates were first quantitated using Pico G r e e n ® d s D N A quantitation kit (Molecular Probe, Invitrogen, Carlsbad, C A ) , and the fluorescence was measured with a CytoFluor™ 2300 Fluorescence Measurement System (Millipore, Billerica, M A ) .  2.5. Genomic P C R . To isolate genomic D N A , cells were lysed with 50 ul o f d F ^ O and vigorous pipetting, placed at 98°C for 10 minutes, and then 5 ul o f 1 mg/ml proteinase K was added and incubated at 55°C for 2 hours followed by incubation at 98°C for 10 minutes. D N A thus isolated was used as template for P C R . Primers for Vy-Jy and N K G 2 A genomic P C R are listed in Table 2.2. The genomic P C R scheme is shown in Fig. 2.1. The reaction volume for these P C R s was 50 ul, containing 5 ul o f 10* P C R buffer, 1.5 ul o f 50 m M M g C l , 1 ul o f 10 2  37  m M dNTPs, 1.25 ul each o f 10 u M primers, and 0.5 ul o f 5 U / u l Taq D N A polymerase. Thermocycling conditions were as follows: 3 minutes at 94°C followed by 30 cycles o f 45 seconds at 94°C, 2 minutes at 55°C, 1 minute at 72°C and finally 7 minutes at 72°C. 10 ul o f P C R products mixed with 1 ul 10* loading buffer were analyzed on a 1% agarose gel. T C R P and T C R 8 P C R primers are listed in Table 2.2. P C R for T C R P gene rearrangement was as described by Ikawa, et al . 63  The reaction volume was 20 ul, containing 1.5 ul o f 10x P C R  buffer (with M g C l ) , 0.16 ^1 o f 25 m M dNTPs, 0.4 ul each o f 10 u M primers, and 0.2 ul o f 2  5U/ul Taq D N A polymerase. Thermocycling conditions were as follows: 5 minutes at 94°C followed by 35 cycles o f 1 minute at 94°C, 1 minute at 60°C, 2 minutes at 72°C and finally 10 minutes at 72°C. Primers for T C R 5 P C R were described by Capone et al.  1 0  ° . Thermocycling  conditions were as follows: 5 minutes at 94°C followed by 32 cycles o f 1 minute at 94°C, 30 seconds at 55°C, 2 minutes at 72°C and finally 7 minutes at 72°C.  38  a y2  Cluster y1  5  2  43  f-Hf  Jy1  i  Cy2  cpJy3  Cy1  Vy  y1E  I  i n  n  G e n o m i c P C R  Vy1.1 JY  y2E ^ Jy2  Vy1.3 ^ cpCy3 y3E  rnn  r~i  +  JY1  4  jj II I  ni  s c h e m e Cy1  Cy1  Vy5  y4  \  y1E  Vy5 Jy1  _L  OR  \  y1E  _o_  Germline order  Rearranged  NO PCR product  PCR product  Figure 2.1. A . The murine T C R y locus. There are 4 clusters in the T C R y locus. Each cluster contains one J segment and one C segment and variable numbers o f V segments. The nomenclature is that o f Garman et a l . . B. Schematic of genomic P C R design. The genomic P C R design was based on that by Itohara et a l . . If the locus is in germline order, the primers specific to the V and J segment w i l l be too far apart to produce a P C R product. O n the other hand, i f the V and J segment have been rearranged, the primers w i l l be close enough together and a P C R product w i l l result. 103  1 2 9  2.6. Southern blot. 10 ul o f P C R products mixed with 1 ul 10* loading buffer were analyzed on a 1% agarose gel. The gel was alkaline blotted to B i o R a d ' s (Hercules, C A ) Zeta-Probe® membrane. The southern blot was probed with a biotin-labeled oligonucleotide and visualized by Pierce's (Rockford, IL) N o r t h 2 S o u t h ® Chemiluminescent Nucleic A c i d Hybridization and Detection kit. The oligonucleotide probes were labeled with 3' end labeling D N A with biotin14-dATP Invitrogen protocol.  39  2 . 7 . R T - P C R . R N A was isolated from cells with Q I A G E N ' s R N e a s y ® M i n i K i t and reversed transcribed into c D N A with Q I A G E N ' s Omniscript® Reverse Transcription kit. The c D N A samples for R T - P C R templates were equal to lOOng o f R N A . Forward primers for V y 2 , 3, 4, and 5 were paired with a reverse primer for a constant region sequence that is shared by all T C R y gene clusters ( C y l , 2, and 4). The primer sequences are listed in Table 2.2. The reaction volume was 50 p i , containing 5ul o f 10x P C R buffer, 1.5 u.1 o f 50 m M M g C l , 1 pi o f 2  10 m M dNTPs, 1.25 ui each o f 10 u M forward and reverse primers, and 0.5 pi o f 5U/ul Taq D N A polymerase. Thermocycling conditions were as follows: 5 minutes at 96°C followed by 32 cycles o f 15 seconds at 96°C, 40 seconds at 50°C, 1 minutes at 72°C and finally 10 minutes at 72°C. 1 pi o f P C R product was analyzed on a 1% agarose gel.  40  Table 2.2. List of primers  Genomic P C R : Vy2: T G G A C A T G G G A A G T T G G A G Vy3: G A T C A G C T C T C C T T T A C C C Vy4: C T G G G G T C A T A T G T C A T C A A Vy5: G C T A A C C T A C C A T T C T C T G T Jyl: C A G A G G G A A T T A C T A T G A G C V y l . l , Vyl.2: C T T C C A T A T T T C T C C A A C A C A G C Jy4: (PAIRS W I T H V y l . l ) : A C T A C G A G C T T T G T C C C T T T G G Jy2: (PAIRS W I T H V y l . 2 ) : A C T A T G A G C T T T G T T C C T T C T G C A A V54: C C G C T T C T C T G T G A A C T T C C V55: C A G A T C C T T C C A G T T C A T C C J51: C A G T C A C T T G G G T T C C T T G T C C V84: C C G C T T C T C T G T G A A C T T C C Dp2: G C A C C T G T G G G G A A G A A A C T Jp2.6: T G A G A G C T G T C T C C T A C T A T C G A T T NKG2A(5'): C C T T C T C A G G A G C A T C C C T G G A T NKG2A(3'): G A C A A A A C A G A T G A G G C C C A G G G Oligonucleotide probes: Jy2: C A A A T A C C T T G T G A A A G C C C G A G C Jy4: C A A A T A T C T T G A C C C A T G A T G T G C Jyl: T G C A A A T A C C T T G T G A A A A C C T G A G J81:  GTTCCTTGTCCAAAGACGAGTT  RT-PCR: Vy2: C C T T G G A G G A A G A A G A C G A Vy3: C A T C G G A T G A A G C C A C G T A Vy4: A G T G A C A G A A G A G G A C A C G Vy5: C G A T T C T G C T C T G T A C T A C T Vyl.l: AACTTCTACCTCAACCTTGA Vyl.2: A A G T T C T A C C T C A A C C T T G G C-region: C T T A T G G A G A T T T G T T T C A G C  41  2.8. Sequencing of P C R products. R T - P C R products from thymocytes, IL-2-activated adult N K and newborn N K cells were purified using Wizard P C R preps D N A purification from Promega (Madison, WI). The P C R products were ligated into the p G E M - T easy vector (Promega). The plasmid clones were sequenced at the N A P S Sequencing Service (University of British Columbia, Vancouver, Canada).  2.9. Tissue culture: 2.9.1. L cells. The murine fibroblast L-cells were cultured in Dulbecco's modified eagle's medium (with 4500 mg D-glucose/L) ( D M E M ) plus 10% F B S , L-glutamine, penicillin, streptomycin, and 5x10" M 2-mercaptoethanol. 5  2.9.2. OP9 cells. O P 9 stroma cells were cultured in M i n i m u m essential medium eagle, alpha modification with nucleosides ( M E M ) with 10% F B S , penicillin, and streptomycin. 2.9.3. L A K cells. Single cell suspensions o f bulk cells from spleen, lymph node, B M , thymus, liver or lung were cultured with 1000 U / m l IL-2 (PeproTech, R o c k y H i l l , N J ) to expand the population. Cells were incubated i n tissue culture dishes for 3 hours at 37°C and the nonadherent cells were cultured for 7-10 days i n R P M I 1 6 4 0 media containing 10% F B S , L glutamine, penicillin, streptomycin, and 5x10" M 2-mercaptoethanol. M e d i a was changed 5  during the middle o f the culture. 2.9.4. Thymus or L N D N progenitor culture and B M N K P progenitor culture. Thymocytes, L N cells, or B M cells were blocked with 2.4G2 hybridoma culture supernantant and then stained with lineage marker m A b s ( C D 3 , C D 8 , T C R p \ T C R y 5 , C D 1 9 , B220, Mac-1, G R - 1 , N K 1 . 1 (and possibly N K G 2 A / C / E , L y 4 9 G , and L y 4 9 D ) , and T e r l 19). Lineage marker positive cells were removed from the sample with EasySep F I T C Positive Selection kit (StemCell Technologies). Thymus and L N cells were then stained with C D 4 4 and CD25  42  mAbs and D N 1 (Lin"CD44 CD25") and D N 2 ( L i n " C D 4 4 C D 2 5 ) or pre-DN2 (Lin" +  +  +  C D 4 4 C D 2 5 ) cells were sorted. For B M N K P cells, cells were stained with N K 1 . 1 and +  l0  C D 122 and N K P cells were sorted ( L i n " N K l . l " C D 1 2 2 ) . Cells were then seeded onto 0 P 9 +  stroma at 20,000-40,000 cells per 500 ul well o f a 24-well plate. If less cells were sorted, all cells were seeded into one well. O P 9 stroma was grown 2 days in advance in M i n i m u m essential medium eagle, alpha modification with nucleosides ( M E M ) with 10% F B S and P/S to ensure that the stroma was confluent before D N or N K P progenitors were added. The OP9 media was removed and the progenitor cells were grown in M E M with 10% F B S and P/S as well as 150 u M monothioglycerol, 30 ng/ml stem cell factor (SCF), 100 ng/ml recombinant human Flt-3 ligand (Flt3L), 1 ng/ml IL-7, and 25 ng/ml IL-15. H a l f o f the media was replaced on day 4 and i f necessary, cells were transferred to new O P 9 with new media at a later stage. Cultures were grown for 10-12 days.  2.10. C e l l preparation. C e l l suspensions were prepared from spleen, thymus, or L N tissue and passed through a 70 um filter. Cells were washed, red blood cells were lysed with ammonium chloride and washed again. To prepare single cell suspensions o f B M , muscle was removed from femur and tibia and B M was plunged from the bone with a syringe and 28 gauge needle. B M was then made to single cell suspension by passing through a 21 gauge needle repeatedly. Cells were washed, red blood cells were lysed and cells were washed again. For lung and liver, the tissues were perfused with 2% Phosphate buffered saline (PBS). The tissue was cut into small sections and digested with DNase and collagenase. Liver was digested in R P M I with 5% F B S , P/S, 2 M E . Lung was digested in D M E M with 5% F B S , P/S. Liver tissue was digested with 25 U / m l Dnase I and 250 U / m l collagenase I V while rotating at room temperature for 45 minutes. Lung tissue was digested with 50 U / m l Dnase and 250U/ml  43  collagenase I V while rotating at 37°C for 1 hour. Tissues were then passed through a 70 urn filter and washed. The liver pellet was resuspended i n 40% percoll (dilutions made with P B S ) and layered on top o f 70% percoll. For the lung, the percoll was first diluted to 90% percoll with 1 OX P B S . The lung pellet was resuspended i n 44% percoll and layered on 67% percoll (dilutions made with D M E M ) . Gradients were spun at 2100 rpm for 20 minutes. The interface was collected, cells were washed, and red blood cells were lysed.  To isolate B cells, bulk splenocytes from adult and newborn mice were stained with antiCD19-biotin plus streptavidin-PE and anti-CD3-FITC, and C D 1 9 C D 3 " cells were purified by +  cell sorting.  2.11. Staining and F A C S sorting or analysis of cells. Cells were washed in 2 % P B S , counted and pellets were incubated on ice for 15 minutes in 50 u.1 2.4G2 hybridoma supernantant per 4x 10 cells to block Fc receptors. m A b was then added at appropriate 6  concentration for 30 minutes at 4°C. Finally, propidium iodide was added to 5 u,g/ml. Cells were purified by cell sorting by a F A C S Caliber® ( B D , Mountain V i e w , C A ) . Sorted cells were checked for purity. For analysis o f F A C S data CellQuestPro and W i n M D I were used.  2.12. IFNy production assay. B u l k cells (2 x 10 ) from L N , thymus or spleen were 6  resuspended in 2 m l o f R P M I media with 10% F B S , P/S, and 2 - M E with 1 ng o f IL-12 and 0.5 ng o f IL-18. These cells were incubated at 37°C for 24 hours. Approximately 6 hours before the end o f the culture, 1 ul o f G o l g i Plug ( B D Biosciences) was added to each well. This inhibits the secretion o f the IFNy produced and makes it accumulate i n the cells. Intracellular  44  staining o f IFNy was performed with the B D Biosciences B D Cytofix/Cytoperm™ Plus kit. Cells were then analyzed by F A C S .  2.13. Cytotoxicity assay. R M A - S cells were used as target cells. These cells were fluorescence labeled using Invitrogen Vybrant C F D A S E C e l l Tracer kit. A stock C F D A S E solution was made at 100 u M and a working dilution o f 1 u M was used to stain the R M A - S cells. The effector cells were either splenic or thymus N K cells that were cultured with IL-2 in a L A K culture. Thymus N K cells were prepared from B 6 thymocytes depleted o f C D 3 cells or +  from TCRp" "8" " mice. 10,000 C F D A S E labeled R M A - S cells were mixed with either thymus 7  7  N K or splenic N K cells at ratios o f 1:1, 1:2, 1:5, 1:10, and 1:20 in 500 ul o f R P M I in 24-well plates. Following a four-hour incubation, the cells were collected, washed and resuspended i n PI buffer. The cells were then analyzed by F A C S and the percentage o f C F D A S E cells that +  were positive for PI buffer was recorded.  2.14. L N D N cell transplantation. 2.14.1. Intraperitoneal injection. D N 1 and pre-DN2 cells were sorted from IL-15" " mice as 7  above. 20,000 cells were injected intraperitoneally in 500 ul P B S into three N o d Scid IL-2Ry" " 7  mice. Three weeks later, spleens, thymuses, B M , and L N were removed and examined by F A C S analysis for N K cells.  2.14.2. Intravenous injection. D N 1 and pre-DN2 cells were sorted from Pep3b mice. 20,000 cells were injected intravenously i n 500 ul P B S into two N o d Scid IL-2Ry" " mice. Four weeks 7  later, spleens and B M were removed and examined by F A C S analysis for N K cells.  45  2.15. Statistics. Data was analyzed statistically using the Student's T-test (Microsoft Excel). Differences o f p<.05 were considered statistically significant.  46  3 IDENTIFICATION O F A N O V E L P A T H W A Y OF N K C E L L D E V E L O P M E N T T H A T IS T H Y M U S - D E P E N D E N T A N D I N C L U D E S T C R G E N E REARRANGEMENT  1  3.1. Introduction Currently, the relationship between BM-derived N K cells and the N K cell potential demonstrated by T cell progenitors is unclear. Although adult D N thymocytes have N K cell potential, it is believed that this pathway is not realized in steady state N K cell development in normal mice since N K cell numbers are normal in nude mice, which lack a thymus. While it is well demonstrated that N K cells arise from bipotent T / N K cell progenitors found in the fetal liver, blood, spleen and thymus, it is currently thought that this pathway is restricted to the fetal environment. A s hematopoiesis switches from fetal liver to adult B M , it is assumed that the N K cell developmental pathway switches exclusively to the B M as well. In this regard, it is o f interest that N K cells i n fetal and neonatal mice are different from those i n adult mice. The former express C D 9 4 / N K G 2 , which recognizes non- classical M H C class I Q a - l , and L y 4 9 E , b  but not other L y 4 9 receptors that recognize classical M H C class I, whereas the latter express a full repertoire o f N K cell receptors. What causes the differences between fetal/neonatal and adult N K cells is still unknown, especially the control o f L y 4 9 expression.  We began this study by comparing gene expression patterns between adult and neonatal N K cells to look for differences that may suggest different pathways o f N K lineage commitment in neonatal and adult mice. Unexpectedly, we found that a subpopulation o f N K cells expresses  A version of this chapter has been published. Veinotte L L , Greenwood CP, Mohammadi N , Parachoniak C A , Takei F. Expression of rearranged TCRy genes in natural killer cells suggests a minor thymus-dependent pathway of lineage commitment. Blood, 107, 2673-2679 (2006). 1  47  T C R y genes and that expression is higher in neonatal N K cells than adult N K cells. These studies suggest that a subset o f N K cells develop in the thymus from T cell precursors that have rearranged T C R y genes. Therefore, since both adult and neonatal N K cells express T C R y genes, the bipotent T / N K cell pathway still contributes in the adult environment.  3.2. Results 3.2.1. Microarray analysis reveals expression of T C R y gene in N K cells To examine whether adult and neonatal N K cells follow different pathways o f lineage commitment, global gene expression o f IL-2 activated N K cells, which were purified by two rounds o f cell sorting (over 99% N K 1 . 1 CD3"), were compared by performing triplicate +  microarray experiments on Affymetrix M G - U 7 4 A v 2 chips. The raw expression values o f the microarray experiments were submitted to an online microarray database and they are available, at E B I ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession # E - M E X P - 3 5 4 . The genes were normalized and then filtered on expression, confidence, and fold change. A parametric student's t-test with a p-value cut off o f 0.05 and a Benjamini and Hochberg false discovery rate multiple testing correction was applied. Out o f 12,488 genes on the chip, 12 had statistically significant differences in expression between the two samples (Fig. 3.1). A s expected, 7 o f the 12 differences in gene expression were from L y 4 9 genes with high expression in adult N K cells and l o w or absent in neonatal N K cells.  48  a Affymetrix p r o b e ID  p-value:  100325_at  0.0499  glycoprotein48 B  102744_at  0.0475  M.muscutus T cell receptor V gamma 2 arid T cell receptor J gamma 2 m R H A  102424_at  .0.C0777  :  6 e n e name  c h e m o k i n e ( C - C motJt)ligand3  93430/at;  0.00771 .  chemokine ( C - C m o t i f ) l i g a n d 5  94146_at  0.00267: :  M.muscufus MIP-1b gene tot macrophage inflammatory protein 1b.;  10038?_f_at  0.00287  Ml us musculus natural killer cell receptor (L/491) gene, exon 7, partial cds L y 4 9 B  100326_f_at  0.00267  killer cell ledri-like receptor, subfamily A , m e m b e r s  Ly49H  94779_f_at  0.00287  killer cell lectin-like receptor, subfamily A . m e m b e r 3  Ly49C  93893_f_at  0.000633  killer cell lectm-like receptor subfamily A . member 12  Ly49C  93894_f_at '  0.000333  killer'cell lectm-like receptor, subfamily A . m e m b e r 3  Ly49C  97781J_at I  0.000443  killer cell lectin-like receptor, subfamily A , m e m b e r ?  Ly49G2  97782_f_at  0.000381  killer cell lectsn-like receptor, subfamily A , m e m b e r ?  Ly49G2  b  II ;  :  Adult  •  -••-<• ..  .  NK  •••  ^Neonatal NK  Figure 3 . 1 . Microarray data analysis of differentially expressed genes between adult N K and neonatal N K cell samples. Genes with statistically significant differences when grouped by 'Sample Source' (adult vs. neonatal samples); parametric test, variances assumed equal (Student's t-test). p-value cutoff 0.05, multiple testing correction: Benjamini and Hochberg False Discovery Rate. Genes from adult and neonatal were normalized and were filtered on expression, confidence, and fold change. These remaining genes were then tested by 1 - w a y A N O V A . (a) List o f the genes with statistically significant differences after 1 - w a y A N O V A test, (b) Graph representation o f the same group o f genes as i n (a).  49  Sample  Sampte  source  source  naonatal  adult  P-valUft  Gp49  O.MM  T cell receptor gamma chain  0.0475  0.00777  Ccl3  Ccl5  Ly49H  Ly49C  0.00771  0.00267  0.00267  O IS  Ccl4  0.00287  Ly49B  0.00287  Ly49C  0.000633  Ly49C  0.000833  S3  BS  Ly49G2  Ly49G2  0000443  0.000381  Figure 3 . 1 . Microarray data analysis of differentially expressed genes between adult N K and neonatal N K cell samples. (c) Graph representation o f the same group o f genes as in (a).  50  The most striking and unexpected result was the expression o f T C R y genes i n both adult and neonatal N K cells. A n analysis o f various T C R gene expression in N K cells showed that only T C R y genes were consistently detected in both neonatal and adult N K cells (Fig. 3.2a). T C R 8 gene expression was detectable at a lower level and the probe identification describes it as germline T C R 5 expression. T C R P gene expression was undetectable. Only one o f three T C R a gene probes detected positive expression. The other T C R a and the p r e - T C R a probes were negative and T C R a rearrangement only begins after full T cell lineage commitment and therefore, it is likely that this probe detected non-specific gene expression. N o t only was T C R y gene m R N A expression i n N K cells detected, its expression was also shown to be significantly higher in neonatal N K cells than in adult N K cells with a student's t-test p-value o f 0.0475 (Fig. 3.2b). To confirm that T C R y is not expressed on the N K cell surface, T cells and N K cells were stained with T C R y 8 m A b and F A C S analysis showed no expression on N K cells (Fig. 3.2c).  51  Figure 3.2. Detection of T C R y gene expression in N K cells by microarray analysis. (a) Gene expression patterns o f purified IL-2 activated adult and neonatal N K cells were analyzed in triplicate using Affymetrix GeneChip Mouse Genome U 7 4 A v 2 arrays. Expression of a, P, y, and 8 T C R genes ± s.d. are shown. The black bars represent neonatal N K cells and the white bars represent adult N K cells. The expression values (0-700 on the graph) are based on raw values after default normalization o f the 6 chips, with the 3 adult samples grouped together and the 3 neonatal samples grouped together. For Affymetrix gene chips, each gene is represented by a probe set o f 10-25 oligonucleotide pairs, each pair consisting o f a perfectly matching probe and a probe with one nucleotide mismatch in the middle o f the sequence. The detection call o f whether a gene is present (expressed) or absent (not expressed) is based on binding to the perfect match and mismatch pairs. Therefore, an expression value can be assigned but it does not necessarily mean the gene w i l l be called present. Those with a * show that they are present and the expression values are valid. The affymetrix probe numbers for these genes are: p i : 101311, P 2 : 94202, P3: 99798, a l : 101823, a 2 : 97944, a3: 97945, pa: (preTa): 98354, 8 : 92328, y l : 102745, y2: 102685, y3: 102744. (b) Significant difference in expression levels o f T C R y gene (probe 102744) in IL-2 activated adult N K cells (white bar) and neonatal N K cells (black bar) determined by a 1-way A N O V A (p-value: 0.0475). The values represent the log ratio which is the intensity ratio (adult N K cell sample gene divided by the neonatal N K cell sample gene) log transformed (log2). (c) Surface staining for TCRSy on neonatal N K cells and T cells from a splenic L A K culture show that N K cells do not express T C R on their cell surface.  52  3.2.2. T C R y genes are rearranged and expressed in N K cells The microarray data revealed T C R y gene expression in N K cells. However, the probes for the T C R y genes on the microarrays were specific for the 3' end o f the transcripts and the results did not reveal whether the microarray data were detecting germline expression or expression o f rearranged T C R y gene segments that resulted from V J recombination. In addition, i f it was detecting rearrangement, the microarray data did not show the extent o f possible rearrangement combinations that were expressed. There are four clusters in the murine T C R y locus, each containing variable (V), joining (J), and constant (C) regions. Cluster 1, which is the most commonly studied, consists o f four V segments (Vy2, 3, 4, and 5), one J segment (Jyl) and one constant region ( C y l ) . T o further characterize the T C R y gene expression i n N K cells, genomic P C R was performed to determine whether T C R y genes are rearranged i n N K cells. Forward primers specific to the V segments in the locus (Vy2, 3, 4, 5) and the reverse primers specific to their respective J segment (Jyl) were used. The genomic P C R was designed so that i f the locus was in germline configuration, the V and J primers would bind to segments too far apart from each other to produce a P C R product. O n the other hand i f the V and J segments were rearranged, the primers would be close enough to each other to produce a P C R product o f about 350 bp. IL-2-activated N K cells from adult and neonatal mouse spleen were purified as above and analyzed by genomic P C R . Southern blot analysis o f the P C R products hybridized to Jy-specific oligonucleotide probes determined that neonatal N K cells exhibited rearrangement o f all possible V - J combinations that were examined while adult N K cells had some o f the possible rearrangements (Fig. 3.3a, left panel). The identity o f the larger band seen for the adult N K cell V y 3 - J y l P C R is unknown. These results showed that T C R y gene rearrangement with multiple V - J combinations does occur i n N K cells. To determine whether the rearranged T C R y genes are expressed in N K cells, R T - P C R was performed using the  53  primers specific to the V-segments and their corresponding C-region. Consistent with the genomic P C R results, neonatal N K cells expressed all of the possible combinations while adult N K cells expressed only some (Fig. 3.3a, right panel). The genomic and R T - P C R experiments were also performed with freshly isolated N K cells without culturing with IL-2, and the results were the same to those with IL-2 activated N K cells (data not shown).  o 2j z E c  o c  s 3  r-  U_  <  as  c a Z  I  i Z 1 l e g s £ A t) 0) r-  U_  Z  p  V  e 2 < i? ^ DZ I- u_ J  . -  Vy3  -  Vy4 Vy5  .  o  <  Control  Vy2  JS  Vy2Jy1  Control  4m Mil!  Genomic PCR  RT-PCR  Figure 3.3. T C R y gene rearrangement and expression i n N K cells. (a) Southern hybridization with Jyl-specific oligonucleotide probe o f genomic P C R (left panel) or R T - P C R (right panel) o f IL-2 activated adult and neonatal N K cells to test for rearrangement o f T C R y locus and expression of rearranged T C R y genes. Thymocytes are the positive control and fibroblasts ( L cells) are the negative control. N K G 2 A P C R and Glyseraldehyde-3-phosphate dehydrogenase ( G A P D H ) R T - P C R were used as control for genomic and R T - P C R , respectively, (b) N K cell D N A from adult Rag2" was tested by genomic P C R for V y 2 - J y l rearrangement as in (a). A  3.2.3. Specificity of rearrangement To confirm that the detected T C R y gene rearrangement occurs as a result o f normal V ( D ) J recombination that requires R A G enzymes, N K cells from adult R A G 2 " mice were sorted and 7  tested by genomic P C R . A s expected, no T C R y gene rearrangement was detected in these N K cells (Fig. 3.3b). The specificity o f the R T - P C R was also confirmed by cloning and  54  sequencing the P C R products from thymocytes and from purified adult and neonatal N K cells. The sequences showed that P C R cross-amplified non-specific T C R y genes. However, Southern hybridization to Jy-specific oligonucleotide probes detected only the specific sequences (data not shown).  3 . 2 . 4 . T C R y gene rearrangements in N K cells represent unselected, random recombination Since V y 2 - J y l recombination was most prominent among N K cells, R T - P C R sequencing results for this rearrangement were further analyzed. V y 2 has an in-frame stop codon at the 3' end, which can be removed during the V J gene recombination process. Out of the V y 2 - J y l rearrangements that were sequenced, 4 out o f 9 (44%) were in-frame, productive rearrangements in adult N K cells and 3 out o f 8 (44%) were in-frame, productive rearrangements in neonatal N K cells (Fig. 3.4). These are similar to the expected frequencies (33%) o f random unselected rearrangements.  55  Germline  > o T3 O i_ CL  t  )duc tive  zo  i—  o_  Vy2  Jy1  TCC TAC G G C TAA  T A G C T C A GGT  TCC TAC G G C TAA  TAG CTC AGG T  T C C TAC G G C TAA  CTG  T C C TAC G G C T  TT  TAG C T C AGG T  T C C TAC G G C TAA  AGCT  TA GCT CAG GT  TCC TAC G G C  ATT A  TA GCT CAG GT  TCC TAC GGC TA TCC TAC GGC TCC TAC G G C TA  AGC TAT ATA  T AGC TCA GGT CA  T AGC TCA GGT T AGC TCA GGT  Figure 3.4. Sequences of productive and non-productive T C R y gene rearrangements. Sequences o f neonatal N K cell R T - P C R products for V y 2 - C y l transcripts. Only the sequences at the junction o f V y 2 - J y l are shown. The in-frame stop codon in V y 2 is underlined.  3.2.5. Tcrf N K cells represent a small population of total splenic N K cells To determine the percentage o f splenic N K cells with T C R y gene rearrangement (termed Tcry N K cells hereafter), fresh and IL-2 activated N K cells from adult and neonatal mice were purified by two rounds o f cell sorting. The levels o f purity o f the N K cell samples used in this experiment were always over 99% with 0-0.05% C D 3 or T C R y 8 (Fig. 3.5a). D N A was +  +  isolated from the purified N K cells and subjected to genomic P C R analysis for T C R Vy2-Jyl gene rearrangement. To determine the frequency o f V y 2 - J y l rearrangements among N K cells, D N A was also isolated from purified y8T cells and mixed with fibroblast (L-cell) D N A at various ratios, and genomic P C R was performed in the same way. The intensity o f the bands for adult and neonatal N K cells was compared to the various control percentages. To ensure that the starting amount o f D N A was identical for each sample, the D N A was first measured with P I C O green staining. Results consistently showed that about 5% o f neonatal splenic N K  56  cells and about 1% o f adult splenic N K cells had V y 2 - J y l gene rearrangements (Fig. 3.5b, c). The frequency was the same with freshly isolated N K cells and IL-2-activated N K cells, as the P C R bands were o f similar intensity (Fig. 3.5b, d). Since these percentages are low, it was important to rule out the possibility of T cell contamination i n the N K cell samples. The same experiment was also performed with N K cells from TCRP"'^" " mice which have no T cells in 7  the spleen. T C R y gene rearrangement was still observed i n these N K cell samples, thus ruling out the possibility o f T cell contamination (Fig. 3.6).  57  a  A  Neonatal  _ J . .11  Adult  t  99.3  99.3 m  •  id'  ' 'w  CWfflC  •  CD3e  * * 5 %yoTcell DNA  Z  Z  Z *i  = ±i  (0 C  <  <  D  20% 10%  % y5T cell DNA 10%  5%  1% 0%  5% 2.5%  CM  —  d  1  -g <  § z  1% 0%  D  O  (0 C  O  z  % v5T cell DNA 10%  d  5% 1% 0%  z  7  t  ~5  T3 <  1 c  o  0) z  Figure 3 . 5 . L o w frequency of N K cells w i t h rearranged T C R y genes. (a) Purity o f IL-2 activated N K cell samples from adult and neonatal mice after two rounds o f cell sorting. The numbers show the percentages o f N K cells ( N K 1 . 1 C D 3 " ) . (b) Genomic P C R (Vy2-Jyl) performed with 8yT cell D N A and fibroblast D N A mixed at various ratios and with fresh and IL-2 activated adult and neonatal N K cells. The P C R products were analyzed by agarose gel electrophoresis and stained with ethidium bromide, (c) Southern hybridization to J y l probe o f the genomic P C R products generated from IL-2-activated N K cells from adult and neonatal mice i n (b). The top panel shows a short exposure o f the southern blot whereas the bottom panel shows a long exposure to visualize rearranged V y 2 - J y l in adult N K cells, (d) Southern blot o f genomic P C R , as in (c), but with freshly isolated adult and neonatal cell D N A . Gels divided by lines are groupings o f images from different parts o f the same gel. +  58  % y8T cell DNA  <n. <i> or « * 10% 5% 2.5% 1% 0% o I— z  Figure 3.6. 7c#y N K cells are not due to contamination of T cells. The frequency o f IL-2 activated N K cells from TCRp'^TCRS" " mice with V y 2 - J y l rearrangement was estimated by genomic P C R and ethidium bromide staining o f agarose gel. Genomic P C R (Vy2-Jyl) was performed with 8yT cell D N A and fibroblast D N A mixed at various ratios and with N K cell D N A . +  7  3.2.6. N K cells have a germline T C R P locus and may have initiated T C R 8 rearrangement Rearrangement o f T C R p and T C R 8 genes in N K cells was also tested by genomic P C R . Rearrangement o f T C R p genes was not detectable in neonatal and adult N K cells with the genomic P C R scheme used (Fig. 3.7a). O n the other hand, a very small fraction (less than 1%) of neonatal N K cells had V84-J81 rearrangement (Fig. 3.7b).  59  a  % yST cell DNA  W ±;  10% 2.5% 1% 0% Germline TCRp  -a <  a)1  S c  z  V54  H  -J51  Rearranged TCRp  V55 -J51  Figure 3.7. T C R p and T C R 8 gene rearrangements in N K cells. D N A from purified I L - 2 activated adult and neonatal N K cells was tested by genomic P C R for T C R P (a) or T C R 8 (b) rearrangement, (a) Genomic P C R using primers specific to D p 2 and JB2.6 genes was analyzed by agarose gel electrophoresis and stained with ethidium bromide. The largest band represents non-rearranged germline T C R p locus whereas multiple smaller bands represent T C R p gene rearrangements. Thymocyte D N A was used as positive control and fibroblast D N A was used as negative control, (b) Genomic P C R (V84-J81 or V85-J81) performed with 8yT cell D N A and fibroblast D N A mixed at various ratios and with fresh and IL-2 activated adult and neonatal N K cells. The P C R products were analyzed by agarose gel electrophoresis, blotted and hybridized to J81 -specific oligonucleotide probe. Gels divided by lines are groupings o f images from different parts o f the same gel.  3.2.7 T C R y gene rearrangement found in N K cells does not occur in CLPs which generate mature B cells 77  R A G genes have been shown to be activated in C L P s in the B M  78  '  , and about 5% o f adult  N K cells have been shown to have rearranged immunoglobulin heavy chain gene . Therefore, 27  whether T C R y genes are also rearranged in B cells in neonatal and adult mice was tested. Genomic P C R analysis o f purified B cells detected no V y 2 - J y l rearrangement in B cells (Fig. 3.8). These results suggest that a subpopulation o f N K cells develop from thymic T / N K bipotential progenitors that have rearranged T C R y genes and lost B cell potential.  60  0)  o CO  S1  IT  t—  5  u-  Vy2-Jy1  Neonatal  control  Vy2-Jy1  Adult  control  Figure 3.8. L a c k of T C R y gene rearrangement i n B 6 mouse B cells. Southern blot with J y l probe o f genomic P C R (Vy2-Jyl) o f B 6 adult and neonatal splenic B (CD19 CD3"). Genomic P C R for a part of N K G 2 A gene confirms the presence o f B cell D N A . +  3.2.8. T h e thymus is r e q u i r e d for the development of Tcry N K cells +  The T C R gene rearrangement in N K cell subsets suggested that they develop in the thymus, since this is the location where the majority o f T cells undergo T C R V ( D ) J recombination. To examine whether the Tcry N K cells develop in the thymus, N K cells were isolated from nude +  mice, which lack a proper thymic environment and lack conventional T cells. Only extrathymic T cells accumulate in the spleen o f older nude mice (up to 5.4% o f splenocytes) (Fig. 3.9a). It was known that N K cells were present at normal or elevated levels in adult nude mice, but the N K cell status o f neonatal nude mice was not known. Normal numbers o f N K cells were found in the spleen o f neonatal (3 day old) nude mice. Genomic P C R analysis o f freshly isolated N K cells and IL-2-activated N K cells from nude mice showed no V y 2 - J y l rearrangement i n N K cells from nude mice (Fig. 3.9b). It should be noted that extrathymic T cells, which accumulate in the spleen o f old nude mice, had rearranged T C R y genes, but N K cells isolated from the same mice did not (Fig. 3.9b) demonstrating that the T C R rearrangement can still occur in these mice and the absence o f it in N K cells is due to the lack o f the thymus.  61  Furthermore, genomic P C R analysis o f highly purified fresh and IL-2 activated N K cells from the B M and thymus showed that T C R y gene rearrangement is present in at least half o f thymic N K cells while very low ( - 5 % or less) rearrangement was detected in B M N K cells (Fig. 3.10). Therefore, the thymus is required for the development o f Tcry N K cells. +  Adult nude  Neonatal nude 0.05  2.3  m  0'  CD3  0.15  w  w  CD3  b  Figure 3 . 9 . Lack of T C R y gene rearrangement in nude mouse N K cells and high T C R y gene rearrangement in thymus N K cells. ( a ) Percentages o f N K ( N K 1 . 1 C D 3 " ) cells and T ( N K 1 . 1 " C D 3 ) cells i n the spleen of 1 year old (left) and three day old (right) nude mice, ( b ) Agarose gel electrophoresis and ethidium bromide staining o f genomic P C R (Vy2-Jyl) o f IL-2 activated and fresh N K cells from adult and neonatal nude mice. T cells that accumulate i n aged nude mice were also isolated from spleen o f the same adult mouse by cell sorting. Thymocytes were used as positive control and fibroblasts (L cells) were used as negative control. Genomic P C R for a part o f N K G 2 A gene confirms that comparative amounts o f template D N A was used for all the genomic P C R . +  +  62  <D O  %y6T cell DNA  =o  *  *  |  100% 50% 25% 10% 5% 0% CO  •a z ! I *i .Q  control  * E  1I  ?5  Figure 3 . 1 0 . Thymus N K cells have high levels of T C R y gene rearrangement. Southern blot with J y l probe o f genomic P C R products generated from I L - 2 activated N K cells from adult thymuses and B M and 8y T cell and fibroblast D N A mixed at various ratios. The middle panel is a longer exposure o f the same membrane. The bottom panel shows control N K G 2 A P C R confirming that comparable amounts o f D N A were used for the analysis.  3 . 3 . Discussion  The microarray analysis suggested that adult and neonatal N K cells have very similar gene expression profiles. This argues against our initial hypothesis that neonatal and adult N K cells follow different pathways o f development. Although the microarray data was not further analyzed with respect to this original objective, it was extremely useful because it suggested that N K cells may follow more than one pathway o f development within the body. The discovery o f T C R y m R N A i n N K cells led to the discovery and characterization o f a thymusdependent N K cell developmental pathway.  63  The microarray analysis revealed that adult and neonatal N K cells do not differ much in gene expression other than the L y 4 9 receptors, which has been well characterized. Fetal and neonatal N K cells do not express L y 4 9 receptors, except L y 4 9 E . Cells begin to acquire L y 4 9 6 6  receptors after birth and only reach full expression levels at 6 weeks o f age. This is what we observe in the microarray expression results as well. The other four receptors that were significantly higher in adult N K cells than neonatal N K cells were N K cell functional factors: C C L 3 , C C L 4 , C C L 5 and gp49B.  The most striking result was that N K cells express a T cell specific gene, TCRy. This is unexpected because it is commonly assumed that T C R rearrangement marks final commitment to the T cell lineage. Both adult and neonatal N K cells express the T C R y and the difference in expression between the two is significant as well. More recently, another study has also detected T C R y expression (but not other T C R expression) in N K cells. K a n g et al.  130  performed serial analysis o f gene expression ( S A G E ) o f H S C s , N K precursors ( N K P s ) , and mature N K cells ( m N K s ) and they detected T C R y expression in m N K cells but did not expand upon this discovery.  It is important to note that the T C R y m R N A expression detected i n the N K cells does not indicate this protein is expressed on the cell surface. Firstly, the N K cells were sorted as N K 1 . 1 C D 3 " cells and D N thymocytes cannot assemble a functional T C R complex on the cell +  131 132  surface in the absence o f C D 3 s  '  . T C R y cannot be expressed on the cell surface without  T C R 8 either . This was confirmed by flow cytometry and no N K cells express T C R y on their 129  surface. It is only known that T C R y m R N A is expressed i n N K cells since a western or intracellular staining for the T C R y chain was not performed to check i f the protein was 64  expressed. Since the T C R is not expressed on the cell surface, it likely has no functional role in the N K cells. Rather, it serves as a marker o f N K cells that have followed this developmental pathway.  There are several examples o f cells that are not T cells or B cells, respectively, with T C R or Ig 77  gene rearrangements. For example, 5% o f N K cells have I g H gene rearrangements  , 30-50%  of peripheral T cells have I g H gene rearrangement , and 16% o f adult B cells had V - D T C R 5 133  rearrangements . 134  There are multiple cases o f gene expression promiscuity in hematopoietic  progenitors, as elaborated upon in the Introduction.  This novel pathway o f N K cell lineage commitment includes T C R rearrangement and the thymus. The higher expression o f T C R y m R N A in neonatal N K cells suggests that the pathway is more common in the fetal/neonatal mouse as would be expected due to the lack o f a B M environment during fetal development and the multiple studies that reveal bipotent T / N K progenitors i n multiple tissues. The adult pathway is important for the production o f N K cells for most o f the lifetime and it is o f more interest that the thymus dependent pathway is maintained in the adult. While adult N K cells do develop i n the B M , we now see that this is not the case for all adult N K cells. This is significant because it shows that N K cells follow at least two separate pathways o f lineage commitment and it is important to know that the B M pathway is not representative o f all N K cells i n steady-state N K cell development.  The presence o f T C R y gene rearrangement in N K cells does not necessarily prove that a subset of N K cells develop i n the thymus since R A G enzymes are also expressed in C L P s in the B M . However, B cells, which arise from C L P s , do not have T C R y gene rearrangement.  65  This  suggests the T C R rearrangement in N K cells does not occur due to R A G expression in the B M progenitors. W e have also directly shown that Tcry N K cells are thymus dependent since they are absent in nude mice. It also rules out the possibility o f the Tcry N K cells arising from an extrathymic pathway that undergoes V ( D ) J recombination outside o f the thymus because extrathymic T cells that accumulate in old nude mice have T C R y gene rearrangement, while any N K cells still present do not. In addition, to confirm that the absence o f N K cells in nude mice is not due to the defective FoxNl  gene instead o f the lack o f a thymus, we show that half  of thymus N K cells have T C R y gene rearrangement while B M N K cells do not. It is likely that the small percent o f B M N K cells that are Tcry are circulating N K cells. The thymus N K cells that do not have V y 2 - J y l rearrangement may have other V y rearrangements or they may be N K cells that arise from D N 1 progenitors before V ( D ) J recombination begins.  Recently a second functional thymus that is the size o f a small lymph node in the neck has been described . It is unlikely that Tcry N K cells arise exclusively from this second thymus, 135  since they are absent in nude mice, which lack the classic thymus. The only way that they may arise from the second thymus is i f this thymus is also defective in nude mice, which is not yet known or the number o f N K cells produced from this thymus is so small that they are undetectable i n nude mice.  It is most likely that Tcry ~NK cells arise from D N 2 T cell precursors since this is the stage that +  T C R y gene rearrangement begins . T C R y and T C R 8 gene rearrangements occur at least one 100  full stage ahead o f T C R P gene rearrangement. Specifically, V y 2 - J y l rearrangements are clearly seen at the D N 2 stage and maximal rearrangement is reached by the D N 3 and D N 4 stages. This is similar for T C R 8 . While a few complete T C R y and T C R 8 gene rearrangements  66  can be detected at the D N 2 stage, the T C R P locus is mainly in germline f o r m  100  . The  possibility that Tcry N K cells arise from D N 2 thymocytes is consistent with what has been +  described about T N K P development in the fetal thymus as w e l l . Fetal D N 1 cells include 63  both T N K P s and N K progenitors while D N 2 cells on the other hand, have a large number o f T progenitors as well as T N K P and N K progenitors. B y the D N 3 stage, there is little progenitor activity left except for a limited number o f T progenitors . In our study, T C R P gene rearrangements were undetectable in N K cells, suggesting that there may be a narrow window of opportunity for T N K P s in the thymus to rearrange T C R y genes and still retain N K potential. It is likely that once thymocytes rearrange both T C R y and T C R 8 genes or T C R P genes, they become committed to the T lineage and lose N K potential. Our results suggest that T C R y gene rearrangements may initiate earlier than T C R 5 since multiple Vy-Jy recombinations were detected whereas T C R 8 gene rearrangement is very rare (although microarray results do report germline T C R 8 expression). Although only V84-J81 and V85-J81 recombinations were examined i n our study, the adult T C R 8 repertoire is dominated by V 8 4 , V 8 5 , and V 8 7 representing  8 0 - 9 0 % of the  cells  136  . Studies on the kinetics o f T C R rearrangement state that  T C R y and T C R 8 gene rearrangements begin at the same stage. So, another possibility is that N K cells diverge from the T cell lineage at the beginning o f the D N 2 stage with germline T C R y expression. The N K cells may continue T C R y gene rearrangement following N K cell lineage commitment i f they retain the enabling factors for T C R y gene rearrangement but lack those for T C R 8 . Some N K P s express I L - 7 R a and this signal does induce T C R y gene rearrangement specifically "' . Our results also dispute the common assumption that T C R 114  16  gene rearrangement marks the final irreversible commitment o f cells to the T cell lineage as N K cells can still arise after T C R gene rearrangement has begun. It would be o f interest to test  67  T C R y gene rearrangement in D C s since their potential is seen at the D N 1 and D N 2 stage as well.  Our studies concentrate on cluster 1 rearrangement. Primers were originally used for V y l . l and V y l . 2 but following sequence analysis to test R T - P C R specificity, it was determined that the P C R was not specific and we therefore ended the analysis o f these clusters. Another study experienced technical problems with V y l . l as well and their P C R was not specific and amplified other V y genes . Once it was determined that rearrangements involving all V y 100  segments o f cluster 1 were seen i n N K cells, we primarily focused on V y 2 - J y l recombination as it was most prominent among several possible Vy-Jy recombinations. V y 2 and V y 1.2 rearrangements are about -10-12 times more frequent than V y l . l and 16-20 times more than Vy5 rearrangements . Although V y 2 - J y l is one o f the most common rearrangements in adult 127  y8 T cells, the percentage o f N K cells with T C R y gene rearrangements is likely an underestimate since we are basing the value on only V y 2 - J y l rearrangements and other cells may have rearrangements other than this combination. A l s o , additional N K cells likely arise from D N 1 cells but since they are not expected to have rearranged T C R y genes, we cannot make any conclusions about whether a larger proportion o f N K cells develop in the thymus 63  before the initiation o f T C R y gene rearrangement. Ikawa et al.  examined T C R gene  rearrangement in N K cells generated from fetal thymic progenitors by F T O C and did not detect rearrangement o f T C R P genes or T C R V y 3 - J y l or V y 4 - J y l gene rearrangement. This is inconsistent with our current results, but the discrepancy may be explained by the fact that 63  V y 2 - J y l is the most prevalent rearrangement detected in our study whereas Ikawa et al.  did  not examine this rearrangement. It is also possible that their genomic P C R may not have been 68  sensitive enough to detect the small percentage o f rearrangements. Older studies examined T C R p \ 8, and y gene rearrangement in N K cells v i a Northern blot and did not detect rearrangement. These studies used splenic N K cells, which have l o w levels o f rearrangement so it was likely that their method was not sensitive enough  137 139  "  .  The method used to calculate the percentage o f cells with T C R y gene rearrangement can only serve as a rough estimate. Originally, single cell R T - P C R was attempted but the method was unsuccessful. Multiple attempts to optimize the sensitivity o f the P C R to detect gene expression o f housekeeping genes at low cell numbers did not work. Therefore, the alternative approach was to use y8 T cell D N A mixed with fibroblast D N A at various percentages. The majority, but not all, o f adult y8 T cells have rearranged V y 2 - J y l and therefore 100% o f yS T cells does not completely equal 100% V y 2 - J y l rearranged loci for the genomic P C R . When comparing the intensity o f the y5 T cell/fibroblast percentage bands with the N K cell bands, it is not a quantitative comparison o f intensity, rather it is just by visual comparison that conclusions are made. Southern blot analysis o f the P C R is useful because at different exposures, only certain bands appear. Therefore i f two bands appear at the same time, they are similar in intensity. This method serves only as a rough measurement o f T C R y gene rearrangement and provides a general idea o f the percentage o f rearrangements but is not an accurate quantitative tool.  To conclude, our study provides compelling evidence that a population o f N K cells does, in fact, develop i n the thymus during steady state N K cell production. The N K cell potential demonstrated by others i n the bipotent T / N K progenitors i n the fetal environment and in the  69  adult DN1 and DN2 thymocytes does represent a normal pathway o f N K cell development and reveals a close developmental relationship between T cells and N K cells.  70  4 THE THYMUS-DEPENDENT DEVELOPMENTAL PATHWAY GENERATES U N I Q U E S U B S E T S O F N K C E L L S IN T H Y M U S A N D L Y M P H N O D E  •4.1. Introduction In the previous chapter we determined that Tcry N K cells are produced v i a a thymusdependent pathway. N o w that the pathway has been identified, it is important to identify the precursors that give rise to the thymus-dependent N K cells. W e hypothesize that the thymus progenitors that produce Tcry N K cells are the D N 2 progenitors. D N 3 cells have irreversibly committed to the T cell lineage while D N 1 and D N 2 cells retain N K cell potential and D N 2 cells undergo T C R rearrangement. It has been shown that 1 in 9 D N 1 cells and 1 i n 14 D N 2 cells become N K cells i n v i t r o . Although N K cell potential has previously been 140  demonstrated in these early T cell progenitors, the N K cells produced in culture have not been further characterized beyond N K 1 . 1 expression. A l s o , it was assumed that this pathway is not +  realized in steady state N K cell development in vivo. The knowledge o f T C R y gene rearrangement i n thymus-dependent N K cells can be used as a marker to test whether D N progenitors are the precursors to thymus N K cells. It is also important to determine i f the thymus-dependent pathway plays a unique role i n the generation o f N K cells. For example, do the N K cells have different functional capabilities or a phenotype that would suggest a unique role for the cells? Does this pathway produce cells that localize in other tissues? While we have determined that the thymus-dependent pathway does not significantly contribute to N K cell populations in the spleen or B M since their percentage o f N K cells with T C R y gene rearrangement is very low, it is unknown i f this pathway is important i n other tissues besides the thymus. The significance o f the thymus-dependent pathway is addressed in this chapter.  71  4.2. Results  4.2.1. D N l and DN2 thymocytes differentiate into N K cells A n N K cell differentiation culture system was used to generate N K cells from D N l and D N 2 thymocytes (Figure 4.1).  DN2  DNl  v A 7 LirvCD44 CD25+  (DN1) Q  cfiP  Thymus  USA  LirrCD44 CD25 (DN2) cells +  -,$f  to*  to*  CD3FITC  W  •  lf|'  Grow on OP9 cells with IL-7.SCF, Flt3L, IL15  DNS  +  io  •#  4  /° \ R  io' IO IO o « ore 2  3  <O  4  Sort NK cells  Analyze differentiated cells by FACs  Isolate DNA  Analyze for TCRy rearrangement  Figure 4.1. Scheme of thymus (or LN) D N progenitor culture. D N progenitors are sorted and then cultured with a cytokine cocktail to induce N K cell development. The N K cells produced are then analyzed by F A C S or sorted and D N A is isolated for genomic P C R analysis.  In previous studies, N K cell potential was reported in T cell differentiation conditions using OP9-DL1 stroma, which promotes T cell development. B u l k thymocytes were first depleted o f lineage marker ( C D 3 , C D 8 , T C R p , TCRy8, C D 1 9 , B220, M a c - 1 , G R - 1 , N K 1 . 1 , and T e r l 19) +  72  cells and then D N 1 (Lin"CD44 CD25") and D N 2 ( L i n " C D 4 4 C D 2 5 ) cells were sorted. To +  +  +  compare the N K cell differentiation o f these cells with those o f defined N K cell progenitors, we also sorted and cultured B M N K P ( L i n " N K l . l " C D 1 2 2 ) cells. Cells were grown on OP9 +  stroma with a cytokine cocktail o f IL-15, IL-7, S C F , and flt3L. After 10-12 days cells were analyzed by F A C S . A t the end o f all cultures, the majority (over 90% in all cases) o f cells were N K 1 . 1 (Fig. 4.2). N K G 2 A and C D 9 4 expression was very high while L y 4 9 expression +  was variable but was always lower than normal in vivo N K cell levels (Fig. 4.2b). This does not appear to be a result o f the progenitor tissue origin as the same was evident in both the thymus D N cultures and the B M N K P cultures. Rather, it appears to be a result o f the culture conditions. Although the L y 4 9 positive population was small, it was always a clearly visible, separate population, as seen i n Fig.4.2a. Cells also expressed low levels o f IL-7Rct, Mac-1 and N K G 2 D . 2 B 4 expression was only examined in one culture o f D N 1 and D N 2 cells but not N K P cells and its expression was very high (Fig. 4.2b). A l l other values are taken from 5 to 9 replicate cultures.  73  a  Bone marrow NKP  10  89  '•  10  10  2  10  3  ' *  "r^ssir-K:;-:—  10"  Thymus DN1  10°  95!  n O  *  10  10°  10'  *  3  10  4  0.2  2  o o> 6  2  CD Z  T0°  10  -  1  10*  10  a  10*  -  _  >%  o  .'•\'v^. >  o o  o  I"*  •'•Jv-"-  •-  ft  o o 10'  o  Thymus DN2  o  87  0.1  c  o  CD C D  4  O) IM  °  crioj  Jig  Z  10°  10  1  10  2  10°  o  10  4  4  CD3  NK1.1  10°  10 1 0 10 10 10° 10 10 10 NK1.1 NK1.1 1  2  3  4  1  2  3  10*  120 110 100  ±  90  u  80  Dl RJ  70  C  60  S  y  6 D_  • DN1 •  so 40  DM2  • NKP  -H  30 20 10 0  of  gP  a*  4*  /  4T  Figure 4.2. T h y m u s D N 1 a n d D N 2 progenitors have the potential to give rise to N K cells d u r i n g in v i t r o cultures. B M N K P s and D N 1 and D N 2 progenitors were cultured on OP9 in N K cell differentiation conditions. (a,b) Profiles o f cells at the end o f cultures, (a) values are representative o f one culture set. (b) Averages of receptor expression each from at least 4 cultures (except 2B4). 74  4.2.2. DN2-derived N K cells are  Ter/  The D N l - and DN2-derived N K cells were sorted and T C R y gene rearrangements were examined. Semi-quantitative genomic P C R showed that 25-50% o f DN2-derived N K cells had Vy2-Jyl rearrangements while D N l - d e r i v e d N K cells had low T C R y gene rearrangement with about 5% of cells being positive (Fig. 4.3a,b). This is to be expected since T C R gene rearrangement only begins at the D N 2 cell stage. Therefore, Tcry N K cells likely arise from D N 2 cells in vivo while other thymus N K cells likely arise from the earlier D N l progenitors which have not begun T C R y gene rearrangement yet.  d) o Z  Z  TJ aj  TJ  •c  •c  T>  TJ  > 0)  %y8T cell DNA 100 50  25  10  5  —  T  0  z Q  0) 0  o  ^ z •a  0)  > 0)  (N  z Q  0) o  =  a>  j* %  s  " 2 -a % rr  k  z  z  -  CM  Z  Z  Q  Q  z TJ  1 s c  100 50  25  10  5  c a) TJ T CN Q ° D  %y5T cell DNA  XJ  0  tot  H Genomic PCR control  Figure 4.3. Thymus DN2 derived N K cells have rearranged T C R y genes. N K cells produced during in vitro cultures o f D N l and D N 2 thymocytes were tested by genomic P C R for T C R y gene rearrangements. Genomic P C R ( V y 2 - J y l ) was performed with 8yT cell D N A and fibroblast D N A mixed at various ratios and with N K cell D N A . (a) Ethidium bromide stained gel o f genomic P C R and control N K G 2 A P C R . (b) Southern blot o f genomic P C R with J y l probe. Two exposures o f the same membrane are shown.  4.2.3. Thymus N K cells are phenotypically different from N K cells in other tissues In the previous chapter, it was shown that almost 50% o f thymus N K cells have rearranged T C R y genes, suggesting that the majority o f thymus N K cells develop i n the thymus. To determine i f the thymus-dependent pathway of N K cell development produces N K cells that 75  differ from those in other tissues, we performed F A C S analysis o f N K cells from the thymus and compared them to N K cells from the spleen, liver, lung, B M , and mesenteric lymph nodes. The N K cell profile is difficult to visualize in the wild type thymus with such a large percentage o f C D 3 cells, and depletion o f T cells seems to skew the N K cell profile as well. +  Therefore, thymuses from T C R P " ^ " mice were used. N o T cells or N K T cells are present in 7  7  these mice and the N K cell population is easily visualized. A s expected, N K cell receptor expression on splenic N K cells from these mice is the same as B 6 splenic N K cells (Table4.1b). There are many significant differences in receptor expression on thymus N K cells compared to those o f other tissues. Thymus N K cells exhibit both extremes o f expression when compared to other tissues (Table 4.1a).  76  Table 4 . 1 . Global F A C S analysis of N K cells from the thymus, L N , B M , spleen, liver and lung. (a) Averages o f percentages o f receptor expression on N K cells from various tissues determined by F A C S analysis. Values are from triplicate (or more) experiments. Values marked with * are significantly different than spleen N K cells with a p-value <0.05. (b) Percentages o f receptor expression on N K cells from the spleen o f T C R p ^ " mice. 7  a Thymus  LN  BM  Spleen  Liver  Lung  NKG2ATC/E  74.3 ± 10.4*  55.7 ± 1.5  47.7 ± 5.7  45.0 ± 5.6  56.8 ± 5.0  59.0 ± 4 . 6  CD94  83.7 ± 1.2*  56.3 ± 0 . 6  52.3 ± 3 . 5  47.7+4.0  57.0 ± 5.6  59.7 ± 6 . 0  Ly49A  4.0 ± 1.0*  16.5 ± 2.3  17.3 ± 2.9  14.5 ± 0.7  13.3 ± 2.5  11.0+1.0  Ly49G  19 ± 3.6*  43.8 ± 2.5  51.0 ± 2.8  47.7 ± 6.8  41.0 ± 0.0  35.0 ± 5.2  Ly49C/l  8.3 ± 1.5*  25.8 ± 2.2  29.0 ± 5 . 3  46.0 ± 2.6  33.0 ± 9.5  23.0 ± 3.0  Ly49D  . 11.0 ± 3.5*  34.6 ± 6.6  41.5 ± 0.7  43.5 ± 3.3  n.d  n.d.  Ly49H  21.0 ± 1.7  12.3 ± 3.5  17.7 ± 2 . 1  27.3 ± 5.9  n.d.  n.d.  2B4  94.0 ± 2.0  81.0 ± 6.1  81.3 ± 12.9  87.3 + 9.7  94.7 ± 2.9  82.0 ± 26.0  KLRG1  n.d.  6.0 + 2.9  2.4 ± 2.2  5.0 ± 3.6  4.3 ± 2.5  14.0 ± 1.7  Mac-1  12.0 ± 5.0*  52.5 ± 8.8  59.8 ± 8.2  76.3 ± 9.0  78.0 + 5.7  96.7 ± 0.6  IL7Ra  65.0 + 7.9*  25.5 ± 1.7  12.8 ± 3.7  7.3 ± 3.2  4.3 ± 2 . 2  5.7 ± 2 . 3  b TCRS'-<-6-'- S p l e e n NKG2A/C/E  49.0  Ly49G  48.0  Ly49C/l  50.0  Mac-1  76.0  IL7Ra  5.0  Cells expressing N K G 2 A , C D 9 4 , 2B4, and I L - 7 R a are most abundant among thymus N K cells while those expressing M a c - 1 , L y 4 9 A , C/I, G , and D are very rare (Fig. 4.4, 4.5, 4.6). More thymus N K cells express L y 4 9 H than lymph node and B M N K cells but less than splenic N K cells. For the L y 4 9 H values, double staining with 5E6 (Ly49C/I) and 1F8 (Ly49C/I/H) m A b s was performed. The L y 4 9 H value represents L y 4 9 H single positive cells, so the value is lower  77  than reported values since double positive ( L y 4 9 H L y 4 9 C / I ) cells were excluded. +  +  Interestingly, the percentage o f I L - 7 R a cells is very high in thymus N K cells and L N N K +  cells are the only other population with a moderate percentage o f I L - 7 R a cells (Fig. 4.4). +  Also, thymus N K cells have the lowest Mac-1 expression, with L N N K cells a distant second. All  thymus N K cell receptors have significantly different expression pattern from splenic N K  cells except L y 4 9 H and 2B4. The intensity o f receptor staining did not differ between N K cell populations. These results suggest an immature N K cell phenotype ( C D 9 4 / N K G 2 L y 4 9 M a c hl  lo  1 °) of thymus N K cells and also a most closely shared phenotype with L N N K cells.  120  Thymus  Lymph node  BM  Spleen  Liver  Lung  Figure 4 . 4 . Thymus N K cells appear 'immature' ( M a c - l I L - 7 R a ) compared to other tissue N K cells. F A C S analysis o f Mac-1 and I L 7 R a expression on N K cells from various tissues. The percentages o f I L - 7 R a and Mac-1 expression in the thymus N K cells are statistically significant from spleen N K cells with a p-value >0.05. l o  78  h l  Figure 4.5. Thymus N K cells have the highest percentages of N K G 2 A / C / E and CD94 expression and average 2B4 expression. F A C S analysis o f N K G 2 A / C / E , C D 9 4 , and 2B4 expression on N K cells from various tissues. The percentages o f N K G 2 A / C / E and CD94 on thymus N K cells are statistically significant from spleen N K cells with a p-value >0.05.  60  Ly49A  Ly49G  Ly49C/l  Ly49D  Ly49H  Figure 4.6. Thymus N K cells have the lowest percentages of Ly49A, G , D, and C/I expression. F A C S analysis o f L y 4 9 expression on N K cells from various tissues. The expression o f L y 4 9 A , L y 4 9 G , L y 4 9 C / I and L y 4 9 D are all significantly different between thymus and spleen N K cells.  4.2.4. T C R y gene rearrangement in L N N K cells suggests a link with DN-derived thymus N K cells To determine whether the N K cells produced v i a the thymus-dependent developmental pathway migrate to other tissues, we purified N K cells from lung, liver, and both mesenteric and peripheral L N s and checked for presence o f T C R y gene rearrangement by genomic P C R since this is a marker o f N K cells that originated from the thymus pathway rather than the B M pathway. Tissues with a significant Tcry N K cell population have N K cells that originate v i a the thymus-dependent pathway and selectively migrate to the tissue. A s shown i n chapter 3, 50% of thymus N K cells have T C R y gene rearrangement while less than 5% o f splenic and B M N K cells do (Fig. 4.7c). Lung N K cells have negligible levels o f T C R y gene rearrangement as do liver N K cells, except on one occasion where a visible band was detected. Both peripheral and mesenteric L N N K cells, on the other hand, consistently had T C R y gene rearrangement (Fig. 4.7a). Semi-quantitative genomic P C R showed that approximately 2025% o f L N N K cells are positive for T C R y gene rearrangement (Fig. 4.7b).  80  b  a  0) T3  o  0) 0) CL  V)  CL CO.  E >i  _l  r-n  i_  3 _ l  _ l  C  &>  .>  o X  %yoT cell DNA  CN 100  TJ  50  25  10  z z  %y8T cell DNA Thymus:  50%  LN:  15-20%  BM:  5%  Spleen:  1%  Lung:  1%  Liver:  1%  100  50  25  10  Figure 4.7. L N N K cells have the highest percentage of Tcry N K cells other than thymus N K cells. (a) Equivalent amounts o f D N A from N K cells from each tissue were examined by genomic P C R for the presence o f T C R y gene rearrangement. Agarose gel electrophoresis and ethidium bromide staining o f genomic P C R is shown, (b) Southern blots o f genomic P C R with J y l probe was carried out as in Figure 3.5 to estimate the percentage o f T C R y gene rearrangement in mesenteric ( M - L N ) and peripheral ( P - L N ) L N N K cells. Genomic P C R (Vy2-Jyl) was performed with 8yT cell D N A and fibroblast D N A mixed at various ratios and with N K cell D N A . (c) A comparison o f the percentage o f T C R y gene rearrangement in N K cells o f multiple tissues.  4.2.5. D N l and pre-DN2 cells in LNs give rise to N K cells in culture The above results suggested thymic origin o f L N N K cells. However, it was unknown whether they develop in the thymus and migrate to the L N or whether immature thymocytes migrate to the L N and differentiate into N K cells within the L N . Terra et al.  141  previously showed that  L N s have D N l and pre-DN2 ( C D 4 4 C D 2 5 ) progenitors present. The D N l cells differ from +  l0  the E T P profile o f the thymus, and no C D 2 5 D N 2 or D N 3 cells are present in the L N . hl  Furthermore, L N D N cells i n normal mice are unable to differentiate into T cells whereas their  a 56  12  L.I ,. ,  Lin cocktail  CD25  b !-  |  :„™..™„„_„,„. -  O  NK1.1  NK1.1 Z  %y8T cell DNA 100  50  I (WPWHr  25  10  0  z  I  |  z  g  CL  I  Figure 4.8. D N 1 and pre-DN2 progenitors are present in the lymph node and they possess N K cell potential in vitro. (a) Lin" cells i n the L N were gated and D N 1 and pre-DN2 profiles are shown, (b) D N 1 and pre-DN2 cells were sorted and cultured in N K cell differentiation conditions as in Figure 4.1. F A C S profile o f cells at the end of one representative culture is shown, (c) Southern blot of genomic P C R with J y l probe of N K cells that differentiate during the culture was used to estimate the percentage of N K cells with T C R y gene rearrangement. Genomic P C R (Vy2-Jyl) was performed with 5yT cell D N A and fibroblast D N A mixed at various ratios and with N K cell D N A .  N K cell potential has not been addressed. To test the possibility o f thymus-derived progenitors migrating to the L N and producing thymus-dependent N K cells, the N K cell potential o f the L N D N 1 and pre-DN2 cells was tested. The D N cells were purified and cultured for N K cell differentiation as described for D N thymocytes. The majority o f cells recovered from the 82  cultures were N K cells very similar to those from D N thymocyte cultures. These N K cells expressed a detectable level o f L y 4 9 receptors as well (Fig. 4.8a,b). L N DN-derived N K cells were purified by cell sorting and T C R y gene rearrangement was checked by genomic P C R . A s seen in the thymus cultures, D N l - d e r i v e d N K cells had negligible levels o f rearrangement while pre-DN2-derived N K cells had significant levels o f rearrangement (Fig. 4.8c). To confirm that the N K cells were not from contaminating N K cells that expanded during the culture, the same in vitro culture was performed with D N l and pre-DN2 cells from IL-15" " 7  mice which have greatly reduced N K cell numbers. The end result o f these cultures was the same as the w i l d type cultures. Once again the majority o f cells were N K cells and they expressed detectable levels o f L y 4 9 receptors (Fig. 4.9).  IL15' DN1 92  i  9  3  -:io! ' : .•.:  -W fl.2H -  .10* •  i  IL15-/- Pre-DN2 91  12  •if*  R2  *>  18V. 10' • 19? • ' •:1D. :5SC-W.: ..  Ly49G  CD3  Figure 4.9. D N progenitors from IL-15" " mouse LNs still show N K cell potential in vitro. D N progenitors from IL-15" " mice, which lack N K cells, were cultured i n N K cell differentiation conditions as in Figure 4.1. F A C S profiles o f cells recovered from a representative culture. 7  7  83  4.2.6. The D N cells and Ter/  N K cells in the L N are thymus-dependent  To confirm that the Tcry N K cells i n the L N were thymus-dependent rather than being derived through an alternate B M developmental pathway, we checked for Tcry N K cells in the L N s o f nude mice, which lack a thymus. The N K cells in nude L N s lack T C R y gene rearrangement. Therefore, the population from the thymus-dependent pathway is absent in these mice (Fig. 4.10b). In addition, we found that D N l and pre-DN2 cells are greatly reduced i n nude mouse LN.  In w i l d type mice, approximately 50% of Lin" cells are D N l whereas i n the nude mouse,  only 3-9% o f Lin" cells are D N l progenitors. Pre-DN2 cells were 7 fold fewer in nude L N s (Fig. 4.10a). These results indicate that L N D N cells derive from the thymus.  a  <u o 0)  R2  1  2  >1  o o E  •xt "xt  0J TJ 3  Q O  UJ  10°  10'  10*  10"  Lin cocktail  10*  Z  2  Vy2-Jy1 2 CD25  control  Figure 4.10. Nude mouse LNs have lower levels of D N l and pre-DN2 progenitors than wild type LNs and nude L N N K cells do not have T C R y gene rearrangement. (a) Lin" cells were gated and then D N l and pre-DN2 profiles from nude mouse mesenteric L N s were examined, (b) Agarose gel electrophoresis and ethidium bromide staining o f genomic P C R to test for T C R y gene rearrangement in nude mouse L N N K cells and N K G 2 A genomic P C R control.  4.2.7. Preliminary results suggest that L N DN progenitors give rise to N K cells in vivo L N D N l and pre-DN2 progenitors were sorted from IL-15" " donors and were injected 7  intraperitoneally (i.p.) into three N o d Scid IL-2 receptor gamma " mice, which are deficient for 7  all lymphocytes including N K cells. After three weeks, the spleens, thymuses, B M , and L N s  84  were examined for the presence o f CD3"NK1.1 express N K 1 . 1 and therefore all N K 1 . 1  cells (Table 4.2a). The host strain does not  cells are donor derived. L y m p h nodes o f untreated  hosts were invisible. Donor-derived N K cells were undetectable in the B M but detectable in the L N , thymus, and spleen. Mouse 1 and 3 had small numbers o f C D 3 " N K 1 . 1 cells while +  mouse 2 did not. Mouse 3 had the highest percentage o f N K cells in all tissues.  Interestingly,  when C D 3 " D X 5 cells were examined, the population was much higher (Table 4.2b). The host +  mouse N K cells express D X 5 but not N K 1 . 1 . However, N K cells are deficient in the host mice and undetectable in the spleen. Therefore, the C D 3 " D X 5 cells i n the transplanted mice were +  likely donor derived, but it remains to be confirmed. The reason for the difference in N K 1 . 1 versus D X 5 expression is unknown. It is possible that the L N D N progenitors selectively give rise to a D X S ^ N K l . l " cells. Although such cells are readily detected in normal B 6 mouse L N (data not shown), their identity is unknown. To conclude, L N D N progenitors do appear to give rise to N K cells in vivo although the number o f L N DN-derived N K cells is low.  A second study was performed with Pep3b mice as D N progenitor donors and N o d Scid IL-2 receptor gamma" " mice were once again recipients. T w o mice were intravenously (i.v.) 7  injected with L N D N cells. Four weeks later, the spleen and B M were examined for donorderived N K 1 . 1 N K cells (Table 4.2a) as well as D X 5 N K cells (Table 4.2b). Once again, +  +  small numbers o f N K cells were observed.  85  Table 4.2. L N DNl/pre-DN2 cells produce small numbers of N K cells following one i.p. and one i.v. transplantation. (a) L N D N cells ( L i n " C D 4 4 C D 2 5 " ) from B 6 mice were purified by F A C S sorting and 2 x 1 0 cells were injected i.p. or i.v. into each N O D Scid g c " mouse. Tissues were examined 3 weeks (i.p.) or 4 weeks (i.v.) after transplantation for donor-derived N K cells by N K 1 . 1 expression, (b) The spleens (and B M for i.v.) o f each mouse were examined for D X 5 N K cells. A l l values are minus the background staining in control mice. +  /l0  4  +/  +  Percentage of CD3NK1.1* cells:  Percentage of CD3-DX5 cells:  i.p. injection:  i.p. injection:  Mouse 1 2 3  +  Spleen  Thymus  LN  BM  0.3 0.05 1.13  0.34 0 1.58  0.35 n.d. 2.5  0.11 0 0.08  2  Spleen 26.6 1.4 8.04  i.v. injection:  i.v. injection:  Mouse 1  Mouse 1 2 3  Spleen  BM  0.55 0.27  0.18 0.22  Mouse 1 2  Spleen  BM  5.6 10.24  0.3 1.24  4.2.8. Thymus N K and L N N K cells produce lower levels of IFNy upon stimulation N K cells from thymus, L N , and spleen were stimulated with IL-12 and IL-18 for 24 hours and their level o f intracellular IFNy was determined by F A C S analysis. About 18% o f thymus N K cells and 33% o f L N N K cells produced IFNy whereas 45-48% o f splenic N K cells produced IFNy (Fig. 4.11). Although the fractions o f IFNy-producing N K cells i n the thymus and L N were significantly smaller than those in the spleen, the amount o f IFNy, as determined by the fluorescence intensity, i n the IFNy-positive N K cells from the thymus and L N was higher than that of spleen N K cells.  86  P-value= <0.05  P-value= <0.05 1  1  •• •  o a.  ~»  Thymus N K .  Spleen N K  Figure 4.11. L N N K cells and thymus N K cells produce lower levels of IFNy after IL-12 and IL-18 stimulation than spleen N K cells. L N cells, thymocytes and spleen cells were stimulated with IL-12 and IL-18 for 24 hours, stained for C D 3 , N K 1 . 1 and intracellular IFNy, and the percentage o f N K (CD3"NK1.1 ) cells that were positive for intracellular IFNy over isotype matched control antibody staining was determined. Values are representative of triplicate experiments. Both L N (p-value= 2.40947E" ) and thymus (p-value= 1.03E" ) N K cell IFNy levels are significantly lower than spleen N K cells +  13  3  4.2.9. Thymus N K cells have normal levels of cytotoxicity Spleen cells and thymocytes were cultured with IL-2 to generate L A K cells and their cytotoxicity was tested. A cytotoxicity assay measures N K cells' ability to k i l l M H C class I deficient target cells or in other words, their natural cytotoxicity level. N K cells are the effector cells and they are mixed with target cells at various ratios and the number of target cells killed is observed. L A K cells were mixed with CFSE-labeled R M A - S cells at various effector:target (E:T) ratios (1:1, 1:2, 1:5, 1:10 and 1:20) for four hours and their cytotoxicity was measured by F A C S analysis. The level of cytotoxicity was determined as the percentage of C F S E target cells that were P I (minus the background level). Although thymus N K cells +  +  have slightly lower cytotoxicity than spleen N K cells, the difference is not significant. The pvalue from the student's t-test was lower as the E : T ratio increased but at 1:20 E : T ratio the pvalue=0.08 and was therefore still not significantly different (Fig. 4.12).  87  A  Spleen  —•—Thymus  1  2  5  10  20  E:T ratio (:1)  Figure 4.12. Spleen and thymus N K cells have similar levels of cytotoxicity. Spleen cells and thymocytes were cultured with IL-2 to stimulate and amplify N K cells. U n stimulated N K cells recovered from the cultures were mixed with C F S E labeled R M A - S cells at ratios o f 1:1, 1:2, 1:5, l : 1 0 a n d 1:20. Cytotoxicity was determined by the percentage o f P I cells among C F S E ( R M A - S ) cells by F A C S analysis. Values are representative o f quadruplicate experiments.  +  +  4.3. Discussion The results presented i n this chapter have shown that significant fractions o f N K cells in the thymus and lymph nodes o f normal mice have rearranged T C R y genes. The lack o f these N K cells in nude mice indicates that they are thymus-dependent. Therefore, the thymus-dependent developmental pathway proposed in chapter 3 is not restricted to thymus N K cells but a subset of L N N K cells also develop through this pathway. The thymus N K cells most probably derive from D N 1 and D N 2 thymocytes that readily differentiate into N K cells in vitro while those derived from D N 2 , but not D N 1 are  Tcry.  D N progenitors are also present in L N s  1 4 1  . We have demonstrated that they have N K cell  potential and can specifically give rise to Tcry N K cells in vitro. Out o f the Lin" cells in the L N , 56% o f cells are C D 4 4 C D 2 5 " (DN1) and 6.7% are C D 4 4 C D 2 5 +  +  88  l 0  ( p r e - D N 2 ) . It has 141  been reported that the D N l progenitors in the L N differ from those i n the thymus. In the thymus there are two progenitors that can generate T cells, the E T P s , which are L i n " S c a - l c +  K i t I L - 7 R a " and then also, L i n " S c a - l c - K i t I L - 7 R a cells. Only the second population is hl  +  lo  +  located in the L N s . These D N l cells in the L N progress to a pre-DN2 stage with CD44 CD25 +  l 0  expression but then become cell cycle arrested in the G l phase. This is because  they lack expression o f Wnt4 which is needed to upregulate certain genes and downregulate others for the cells to progress i n T cell development . 141  Although we can conclude that a subset o f L N N K cells is thymus-dependent, it is still unclear whether N K cells that develop i n the thymus migrate to the L N or D N thymocytes migrate to the L N and differentiate into N K cells. Since thymus N K cells and L N N K cells significantly differ in their phenotypes, it is unlikely that thymus N K cells significantly contribute to the thymus-dependent L N N K cell population. It is more likely that thymus-derived D N progenitors migrate to L N and differentiate into TCRy+ N K cells. This is supported by the fact that L N D N l and pre-DN2 cells efficiently differentiate into N K cells during i n vitro cultures. They also give rise to N K cells, albeit in small numbers, in vivo as revealed by our preliminary transplantation studies. The L N environment also seems to be supportive for N K cell development. The gut (mesenteric L N s ) has high IL-7 expression  142  which would likely allow  extrathymic T C R y gene rearrangement to occur. A l s o , human L N D C s produce IL-15 upon stimulation , which is crucial for N K cell maturation. Finally, early T cell progenitors do 143  have the capacity to migrate out o f the thymus without losing their differentiation potential or their capacity to colonize primary lymphoid organs. Until very recently it was believed that only mature T cells could exit the thymus. However, a new study has shown that early T cell progenitors, specifically D N 2 and D N 3 progenitors, migrate to the gut and produce CD8aoc T+  89  IELs in the gut cryptopatches . The thymus is only important for the establishment of the 142  C D 8 a a pool during development because i f a thymectomy is performed after the neonatal +  period, this cell population still developed normally. It therefore seems that the D N 2 and D N 3 cells that migrate and settle in the crytopatches have a longer lifespan and can self renew.  One contradiction to the hypothesis that L N D N progenitors produce N K cells is that the L N D N l phenotype resembles C L P s , rather than the thymic T cell precursors, which share N K cell potential. It would be expected that since the cells both share a thymus-dependent pathway via common precursors, that the D N l phenotype i n the L N would be the same as that i n the thymus. One possibility may be that the phenotype o f D N l changes as the cells migrate from the thymus and settle in the L N .  It is still unknown what fractions o f L N N K cells develop through this pathway. Although about 20% o f normal N K cells are Tcry, it is very likely that more L N N K cells are thymusdependent, because not all thymus-dependent N K cells are Tcry.  Both D N l and D N 2  thymocytes show N K potential i n N K differentiation cultures in vitro while only DN2-derived, but not D N l - d e r i v e d , N K cells are Tcry. D N l , L N cells i n vitro are Tcry.  Similarly, N K cells generated from pre-DN2, but not  Since no marker is available to identify D N l - d e r i v e d N K  cells, the extent o f D N l -derived N K cell development cannot be determined.  The thymus-dependent pathway that produces thymus and L N N K cells very likely plays a unique role i n the generation o f N K cells that differs from the B M pathway o f development. It produces N K cells with unique phenotypes and functional abilities. F A C S analysis of N K cells from the spleen, B M , liver, lung, L N and thymus revealed that each tissue exhibits specific  90  differences in N K cell phenotype. Strikingly, thymus N K cells are the most different with receptor expression values on both extremes: the highest levels o f certain receptors and the lowest levels o f others. The thymus N K cell phenotype resembles that o f an immature N K cell at an intermediate stage o f development. They have high expression o f N K G 2 A 7 C / E and C D 9 4 and l o w expression o f most L y 4 9 receptors. A l s o , immature N K cells are normally classified as Mac-1 C D 4 3 l o  l 0  and cells at this stage can also express I L - 7 R a . Thus, most  thymus N K cells have the immature phenotype. Whether the thymus N K cell are 'immature' or whether this phenotype does not represent an immature developmental stage in the thymusdependent pathway remains to be further studied.  While L N N K cells are not as different from other tissue N K cells (and presumably B M derived N K cells) i n receptor expression, they are the population that is the most similar to thymus N K cells based on the expression o f I L - 7 R a and M a c - 1 . Their C D 9 4 / N K G 2 and Ly49 receptor expression patterns are more similar to those o f other tissue N K cells but other similarities exist between the L N and thymus N K cells as well. They are both poor IFNy producers but have regular cytotoxicity compared to splenic N K cells. Since a thymusdependent pathway is shared between thymus N K cells and at least a subset o f L N N K cells, one may expect the phenotype to be the same for the two populations. This need not be the case as the difference i n thymus and L N profiles could either be due to further maturing of thymus N K cells after they migrate to the L N or most likely due to D N progenitors giving rise to N K cells in the L N . The difference i n the L N developmental environment (and perhaps D N l progenitor identity) likely creates the differences in N K cell phenotypes.  91  Chen, et al.  144  previously showed that L N N K cells produced lower IFNy levels than splenic  N K cells but they tested this following poly I:C stimulation in vivo, which stimulates N K cell effector functions v i a T L R 3 l i g a t i o n . The L N N K cell cytotoxicity was not tested in our 145  study because it has already been shown to be the same as spleen N K c e l l s  144  . Although we  tested thymus N K cell cytotoxicity following L A K culture (to increase the cell number for the assay), Carlyle et al. previously examined the cytotoxicity o f fresh fetal and adult thymus N K cells without stimulation and showed that they were functionally mature and lysed Yac-1 target cells in a cytotoxicity assay . Therefore, the IL-2 stimulation did not induce functional 146  maturation in the cells and they are cytotoxic as fresh cells as well.  In this study, we defined thymus D N 1 cells for our in vitro cultures as CD4"CD8~CD3" C D 4 4 C D 2 5 " , which is very heterogenous and includes committed B cell progenitors, myeloid +  cells, and D C s in addition to T cell progenitors. Balciunaite et al? compared c-kit and c-kit" 6  +  D N 1 and D N 2 cells and the lineage marker expression difference between C D 1 1 7 and CD117" +  D N cells was striking. The C D 117" cells expressed a broad array o f markers ( C D 19, N K 1 . 1 , C D 1 l c , and C D 1 l b ) but the C D 117 cells did not express any o f these. They further +  characterized C D 1 1 7 D N 1 and D N 2 cells to show that they have T cell and N K cell potential +  but lack B cell potential. A l s o , early thymus progenitors (ETPs), which are the earliest and most efficient T cell progenitors, fall within the C D 1 1 7 D N 1 classification. More than half o f +  the cells express C D 4 and therefore C D 4 was not included in our lineage marker cocktail for lineage negative selection, to ensure that the E T P / D N 1 cells were not removed. The novel classification o f E T P / D N 1 is L i n C D 2 5 " C D l 17 . These cells are uniformly C D 4 4 so this lo  +  +  marker does not need to be included. Although we used the older classification o f D N 1 cells, which includes the committed cells in the heterogenous m i x , it is not likely that the N K cells in  92  our culture arose from these other cell types. Firstly, committed N K cells were removed since N K 1 . 1 was included i n the lineage marker cocktail. Secondly, Porritt, et al.  XA1  further divided  the C D 1 1 7 D N 1 population into D N l a - e by C D 2 4 expression. Porritt demonstrated that the +  true E T P is D N l a (CD117 CD24"CD127"). These cells have the slowest kinetics o f +  differentiation, the highest proliferative burst potential, and the ability to home to the thymus following i.v. transplantation. Most importantly for our study, only the true T cell progenitor populations ( D N l a and its progeny, D N l b ) have N K cell potential. Therefore, by excluding mature N K cells from the sort and gating on D N l cells only as L i n " C D 4 4 C D 2 5 " , only the true +  T cell precursor within the broader D N l classification possesses N K cell potential . 147  Therefore, when we demonstrate that Tcry ' N K cells arise from thymus D N l (and D N 2 ) cells, 1  they are truly arising from the same progenitors that produce T cells.  To conclude, this chapter revealed that the thymus-dependent developmental pathway produces N K cells i n both the thymus and L N s . It is very likely that thymus-dependent D N progenitors, in both the thymus and L N , give rise to Tcry (and Tcry' cells, which cannot be measured) N K cells. These N K cells are distinct and likely have unique roles that differ from BM-derived N K cells.  93  5 G E N E R A L DISCUSSION  The most significant outcome o f this thesis research is the discovery o f novel pathways of N K cell development. W e have demonstrated that T C R y genes are rearranged i n subsets of N K cells i n the thymus and L N and that they develop v i a thymus-dependent pathways. Furthermore, the phenotypes and functions o f N K cells in the thymus are different from those of spleen N K cells. Most studies on mouse N K cell development have focused on N K cells i n the spleen, liver, and B M , and it has commonly been assumed that a l l N K cells developed v i a the same pathway in the B M . Differences in receptor expression on N K cells in various tissues have been thought to be due to different developmental stages/maturation levels or due to migration o f certain subpopulations o f BM-derived N K cells. Our results now suggest that various developmental pathways and locations may produce different types o f N K cells. Our updated N K cell developmental pathway scheme is shown i n Figure 5.1.  94  Figure 5 . 1 . Revised model of N K cell development in the mouse. The B M and thymusdependent pathway o f N K cell development are illustrated. Participation o f the L N in the thymus-dependent pathway is shown as well as potential locations o f other unidentified N K cell developmental pathways.  In a very recent review, D i Santo addresses the topic o f N K cell development in other tissues 46  as well. Two important points from the review are that N K P s have been identified in L N s and spleen as well as the previously described fetal thymus and adult B M locations. Presence o f N K P s in L N s strengthens the hypothesis that N K cells develop from thymus-derived D N progenitors in the L N . Other developmental intermediates need to be looked for in the L N . Our laboratory is also currently examining the possibility o f N K P s in lungs. Secondly, D i  95  Santo addresses the fact that B M ablation greatly affects N K cell numbers in the periphery. 46  Although this is often cited as evidence that the B M is required for N K cell development perhaps these treatments may affect the capacity o f N K precursors to respond to maturation signals that they would encounter elsewhere in the body. The B M may also only be involved in the initial steps o f N K cell differentiation, with other sites required for N K cell maturation.  5.1. Thymus-dependent N K developmental pathway Our studies with in vitro differentiation o f D N 1 / D N 2 thymocytes and D N l / p r e - D N 2 L N cells into N K cells clearly demonstrated that they have N K cell potentials. Furthermore, T C R y gene rearrangement i n large fractions o f thymic and L N N K cells i n normal mice and the lack of such cells in nude mice indicate that N K cell potentials o f these cells are realized in normal mice. Regardless o f the origin o f the L N Tcry N K cells, whether it is thymus D N progenitors that migrate to the L N and give rise to L N N K cells or immature thymus N K cells that migrate to the L N s and further mature, both L N Tcry N K cells and thymus N K cells are thymus dependent and develop from a shared pathway. This pathway includes T cell precursors that retain N K cell potential up until a stage after V ( D ) J recombination has begun and this pathway differs from the B M pathway in location, precursors, and output.  A s discussed in the previous chapter, not all thymus-dependent N K cells are Tcry , and no specific marker is available to identify all thymus-dependent N K cells. Therefore, the extent o f the contribution o f this developmental pathway in the overall N K population is difficult to assess. However, it should be noted that neonatal as well as adult nude mice have normal number o f N K cells in the spleen. Therefore, it seems likely that the contribution o f thymusdependent N K cell development is tissue-specific. We have found that Tcry N K cells are  96  abundant in the thymus and L N s and rare in spleen, liver, lung and B M , but we have not examined other tissues. It is possible that Tcry N K cells may localize in other tissues. Gut+  associate lymphoid tissues are o f particular interest, because D N 2 (and D N 3 ) thymocytes have been shown to migrate and become a unique T cell population i n gut crytopatches. Perhaps N K cell potential is realized in this location as well.  It is of great interest to determine what the significance o f this pathway is. This pathway could simply be an evolutionary remnant since studies have suggested that the N K cell fate may be the default fate o f thymocytes . W h y does the thymus retain N K cell potential? Perhaps the 35  thymus is a main supplier o f N K cells to the peripheral lymphoid tissues during the fetal/neonatal stages since N K cells are generated first in the F T during ontogeny. This is supported by the bipotent T / N K progenitors in the fetal development and by our studies which show higher T C R y m R N A expression in neonatal N K cells than i n adult N K cells. The pathway may remain i n the adult while most hematopoiesis switches to the B M . Another option is that there may just be no necessity in cutting off N K cell potential before entering the thymus since it does not seem to disturb T cell generation. Since the N K cell potential is retained in D N progenitors and few signals are likely required to switch the cell's fate from a T cell to an N K cell and vice versa, it may be that the N K cell potential is occasionally realized in vivo. A s discussed i n more detail later, it is possible that there are certain areas i n the thymic environment that are more permissive of the N K cell lineage decision. A l s o , the N K cell fate may be adopted when a crucial signal for the T cell lineage is not received or perhaps when T C R gene rearrangements are not successful. T w o findings argue against this hypothesis. First, the N K cell fate is lost after the D N 2 stage, when only complete T C R y and T C R 8 gene rearrangements are seen but rarely TCR(5 gene rearrangements.  97  Secondly, one would expect to  see more incomplete T C R 8 gene rearrangements as well as T C R y gene rearrangements in N K cells. Regardless o f how the pathway arose, it does not seem to be redundant, because the thymus-dependent pathway produces distinct N K cells.  5.2. Mouse N K cell subsets We hypothesize that the developmental pathway and location influence the N K cell phenotype and function resulting in distinct N K cell subsets. Our F A C S analysis o f N K cells in various tissues indeed showed that surface phenotypes o f N K cells i n the thymus are significantly different from those in spleen, liver, lung and B M . However, the interpretation o f this finding is complicated, because N K cells i n a given tissue are heterogeneous with respect to their maturity. K i m et al. identified distinct steps in N K cell development in the B M based on the expression o f various receptors. According to their scheme, immature N K cells are M a c - 1 whereas mature N K cells are M a c - l . Hayakawa et al. h l  148  N K cell subpopulations, namely a M a c - l C D 2 7 h i  h i  10  have recently defined two mature  and a M a c - l C D 2 7 h i  l 0  population. They  differ in receptor expression, tissue distribution, proliferation, function, and chemokine sensitivity. C D 2 7  10  cells have higher Ly49C/I levels, which are the s e l f - M H C I receptors in B 6  mice, and higher K L R G 1 expression. They are also long-lived, senescent cell and they localize in peripheral tissues such as spleen, liver, lung and blood. C D 2 7  cells show greater  h l  responsiveness to activatory ligand on tumor cells and are cytotoxic to tumor cells even in the presence o f M H C class I expression. These cells also respond more strongly to IL-12 and I L 18 stimulation as well as to stimulation directly from D C s . Finally, C D 2 7  h l  cells express high  levels o f C X C R 3 which is crucial for L N migration and they localize in the L N s and B M . Importantly, the cells in each tissue retain the same receptor profiles that accompany the M a c l CD27 h l  h i / l 0  phenotypes. Immature N K cells and two subsets o f mature N K cells defined by  98  Hayakawa et al. coexist in individual tissues. These N K subsets differ from each other in the pattern o f L y 4 9 and other cell surface marker expression, resulting i n high degree o f heterogeneity in the overall N K population in individual tissues. Since the majority o f thymus N K cells are M a c - 1  10  and I L - 7 R a and only small fractions express L y 4 9 , they can be +  considered immature. O n the other hand, the immature phenotype is less obvious with L N N K cells. Almost 50% o f them are M a c - l , and their Ly49 expression pattern is similar to that of h l  spleen N K cells. Therefore, it still remains to be determined whether thymus-dependent developmental pathway generates unique population o f N K cells and whether the location o f N K development influences the phenotype and functions o f the products.  5 . 3 . Human N K cell subsets and L N pathway of development There are two main subsets o f N K cells in the human (reviewed in CD56  dltn  1 4 9  ) . One subset, defined by  , has high cytotoxicity but low cytokine production. The other subset, C D 5 6  b n g h t  , has  poor natural cytotoxicity but plays an immunoregulatory role and are potent cytokine producers. Interestingly, C D 5 6  b r i g h t  blood. It is likely that the C D 5 6  b r i g h t  N K cells are ten times more frequent i n the L N s than in the subset develops i n the L N s since C D 3 4 N K precursors,  which differentiate i n vitro into C D 5 6 LNs  1 4 3  +  b r l g h t  N K cells, are found i n the T cell rich region of  . A follow up study showed that L N s contain all developmental intermediates spanning  from C D 3 4 progenitors to C D 5 6 +  b n g h t  N K c e l l s . Each population is capable o f downstream 150  N K cell differentiation ex vivo. T cell and D C potential are lost during progression through the developmental stages and N K cell cytoxicity, cytokine production and N K cell receptor repertoire is acquired during development . These studies demonstrate that a distinct subset 150  of N K cells develops in human L N . The correlation between this developmental pathway in humans and the thymus-dependent pathway we found in mice is unclear. N K progenitors in  99  human L N likely derive from B M whereas Tcry N K cells in mouse L N seem to develop from D N 2 thymocytes or pre-DN2 cells in the L N . N o T C R rearrangement i n human L N N K cells has been reported. Nevertheless, it is o f interest that N K cells generated from adult C D 3 4 progenitor cells i n B M stroma-free cultures with IL-15 resemble the C D 5 6  b n g h t  +  N K cells with  high C D 9 4 / N K G 2 expression but little or no C D 16 or K I R expression. This phenotype is similar to mouse thymus N K cells, which are mostly C D 9 4 / N K G 2  h l  and Ly49" (mouse  counterpart o f K I R ) . However, the functions o f mouse thymus N K cells are different from human LN-derived N K cells as the former are poor producers o f IFNy upon stimulation with IL-12 and IL-18. Further analyses are needed to determine how these subsets relate to each other.  5 . 4 . T h y m u s N K cells While we have identified their precursors and characterized their phenotype, we have not defined the role o f thymus N K cells. It does not appear that thymus N K cells are hyporesponsive because their cytotoxic abilities are at normal levels even though IFNy production is low. Although L y 4 9 C / I expression is very low, N K G 2 A and C D 9 4 expression are very high. The three s e l f - M H C class I specific receptors i n C 5 7 B L / 6 are L y 4 9 C , Ly49I, and N K G 2 A / C D 9 4 . Therefore, it is likely that these cells passed through the stage of self tolerance induction without becoming (disarming model) or remaining (arming model) hyporesponsive. Perhaps since there are fewer cells (i.e. D C s ) to influence with IFNy i n the thymus, only a lower level is required to be effective.  Thymus derived N K cells may be locally involved in the regulation o f thymopoiesis. L A K cells generated by I L - 2 stimulation o f D N progenitors  100  1 5 1  selectively killed D P progenitors but  not any other progenitors. Therefore, N K cells may aid i n the elimination o f the majority o f D P cells that do not proceed through positive selection. D P thymic autoreactive clones are expected to be "activated" after interaction with their substrate and thus their cell membrane may look 'foreign' due to expression o f activation antigens. Ballas et al.  151  sought indirect  evidence in support o f this hypothesis. Thymocytes were activated by P H A stimulation and tested for their L A K susceptibility. Such activated thymocytes were readily lysed by D N derived L A K , suggesting that L A K may indeed play a role i n thymic selection and elimination of D P thymocytes. The D P cells that are marked to die may be lysed by N K cells. Small cortical thymocytes ( D N progenitors) have l o w or no M H C class I expression while subcapsular thymocytes ( D P progenitors) and thymus migrant cells (mature SP thymocytes) all express M H C class I receptors . However, Ballas et al. 152  151  did not detect killing o f the D N  progenitors by the thymus L A K cells. It should be noted that IL-2-stimulated thymic N K cells were used in these studies to demonstrate their cytotoxicity. Whether thymic N K cells in vivo display any cytotoxicity is currently unknown. Moreover, intrathymic T cell development appears to be normal i n IL-15-deficient mice and Id2-deficient mice, both o f which have greatly reduced number o f N K cells. Therefore, it seems unlikely that N K cells play a major role in intrathymic T cell development.  Another very likely role o f thymus N K cells is to monitor the thymus for infectious agents and tumor cells. D i Santo suggested that N K cells may be important for detection o f thymocyte transformation . N K cells could lyse early thymocytes and could be involved in 46  immunosurveillance o f rapidly dividing thymic precursors. The thymus is a target o f certain infectious agents. Despite the blood-thymus barrier, viruses ( H I V , S I V , lymphocoriomeningitis), parasites (T. cruzi), and fungi (Paracoccidioides  101  brasiliensis) have  1 S} -  been detected within the thymus combat against H I V and T. cruzi  . It is well documented that N K cells play a role in the  which have both been detected in the thymus. One way to  test the role o f N K cells i n the thymus is to infect IL15~ mice or thymuses directly with T. A  cruzi, for example, and determine i f the outcome is different in the absence of N K cells.  5 . 5 . L N N K cells Human C D 5 6  b r i g h t  N K cells are located in the paracortex where circulating lymphocytes enter  the L N s via high endothelial venules ( H E V s ) and where T cells and D C s interact. Approximately 0.5% o f total lymphocytes are N K cells in resting mouse L N s . Many N K cells migrate to L N s following stimulation. While only 20% of resting L N N K cells have rearranged T C R y , this may be due to the lower number o f D N 2 cells in the L N . It is possible that the remainder o f resting L N N K cells derive from D N 1 progenitors. W e hypothesize that the N K cells in resting L N s are thymus-dependent N K cells, whereas the massive migration of N K cells following stimulation is from BM-derived N K cells. Therefore, to study N K cells that are thymus-dependent, one would have to use resting L N s . In L-selectin, E-selectin, and P-selectin triple deficient mice there are very low numbers of N K cells i n the L N s . N K cells in these mice cannot migrate to the L N s so the cells remaining are resident N K cells. These mice may be useful to study thymus-dependent N K cell function. Another option would be to examine the L N s o f nude mice which would lack thymus-dependent N K cells and compare functional abilities and phenotype to w i l d type mice. I f a function or phenotype were changed, one could deduce that the function was normally controlled by the thymus-dependent N K cells that are now absent.  102  Most studies that have examined the role o f L N N K cells focused on those that migrate to the L N , rather than those that are resident in the L N . These N K cells play a role i n controlling tumor formation. When N K cell migration to the L N s is blocked in L-selectin" " mice, there is /  aggressive tumor formation i n L N s following B16 melanoma injection . Another study 144  focused on an N K cell role in priming T h I c e l l s . Recruitment o f N K cells correlates with 154  induction o f T h I responses as they provide the early source o f IFNy which is necessary for the T h I polarization.  Confocal microscopy o f L N sections shows that N K cells reside in two areas o f the L N : the medulla and the T cell rich paracortex . In resting L N s 44% o f N K cells are in the paracortex 155  and 56% are in the medulla. B y intravital imaging it was shown that N K cells are relatively immotile. After stimulation, N K cells accumulate i n the paracortex whereas the N K cell numbers i n the medulla remain the same. N K cells were also shown to make long steady contacts with D C s  1 5 5  . One possible resident (we hypothesize thymus-dependent) N K cell  function may be to monitor the L N for pathogens. One clue to this is that some N K cells reside in the paracortex and can interact with blood borne cells and molecules entering the L N s . From here they interact with D C s as has already been shown. L i k e l y , lymph borne pathogens include Yersiniapestis  (the bubonic plague), H I V , mycobacteria and anthrax.  5 . 6 . T cell v s . N K cell lineage commitment The discovery o f DN2-derived N K cells poses an interesting question o f what determines whether D N cells differentiate into T cells or N K cells. Obviously the T cell pathway is dominant over the N K pathway in the thymus whereas the T cell pathway o f D N cells appears  103  to be blocked in the L N possibly leaving only the N K pathway open. There are two candidates that may play critical role i n the T vs. N K commitment, namely Notch and Id2.  5.6.1. Notch signals The Notch pathway is crucial for commitment to the T cell lineage. Notch receptors interact with two kinds o f ligands, Delta and Jagged, and although they both activate the same downstream target, their developmental responses differ. N K cells can develop in the presence of Notch ligands. When D N 1 and D N 2 progenitors are cultured on O P 9 or OP9-Deltal stroma, both conditions produce N K c e l l s . The N K cell potential is decreased in the 140  presence of Notch signaling but it is not fully removed, especially for D N 1 progenitors. When Notch inhibitor is added to O P 9 - D L 1 cultures, there is a striking increase in N K cells generated as the inhibitor concentration increases. It appears that the Notch ligand and therefore signal strength are determinants o f the switch between T and N K cell fate. Delta 1 signals are stronger than Jagged 1 signals. Both Jagged 1 and Delta 1 can inhibit B cell potential but only Deltal signals are strong enough to induce T cell potential . N K cells are the favored fate o f 156  intermediate Notch signals. D N 1 cells grown on OP9-Jagl do not produce T cells or B cells but preferentially develop into N K cells (20-65%). Jaggedl signals may only promote N K cell development by default since they inhibit B cell potential and fail to promote T cell potential. Or, they may induce commitment to the N K cell lineage . While 5% o f multipotent 157  progenitors become N K cells after culture on O P 9 stroma, 30% o f cells become N K cells following OP9-Jagged2 culture. The absolute N K cell numbers were 7.5 fold higher in the OP9-Jagged2 cultures than the O P 9 cultures . Perhaps the N K cell fate is revealed during 157  normal development in areas o f the thymus where Notch ligands are sparser or in areas of the thymus that are more concentrated with Jagged rather than Delta ligand. Jaggedl is expressed  104  in the correct anatomic location to influence this decision in vivo. Precursors enter the thymus through blood vessels near the corticomedullary junction and Jagged 1 is expressed at high levels on endothelial cells lining the thymus blood vessels . 156  5.6.2. E proteins  vs.  Id2 1 SR  Class I H L H transcription factors play key roles in T cell development (reviewed in  ). These  transcription factors, also known as E proteins, bind to a palindromic D N A sequence called the E box site. E boxes are found in promoter and enhancer regions o f many T cell specific genes including C D 4  1 5 9  and p r e - T a  160  . The E protein family includes four members i n vertebrates:  E12, E47, H E B , and E2-2. The Id (inhibitors o f D N A binding) proteins act as dominantnegative H L H proteins and can regulate the level o f E proteins. Members o f the Id group lack D N A binding domains but they heterodimerize very well with E proteins and block their D N A b i n d i n g . E protein expression begins to rise at the D N 2 stage o f T cell development and Id 161  expression is also high in D N progenitors . H L H proteins play a direct role in the T vs. N K 158  cell lineage decision. Ikawa et al.  162  demonstrated that Id2 directly controls the production o f  N K P s from fetal D N l T cell precursors. This group previously used a clonal culture system that supports both fetal T and N K cell development to show that D N l and D N 2 cells contain bipotent T / N K progenitors as well as pT and N K P s and that D N l progenitors can be divided into C D 122" and C D 1 2 2 populations, with the C D 1 2 2 cells representing N K P s . Ikawa et +  al.  162  +  6 3  showed that i n Id2" mice, the N K cell percentage i n the fetal thymus was 0.5% compared /_  to 2.5% o f a w i l d type thymus. A l s o , the C D 1 2 2 population o f D N l cells is absent in Id2" " +  /  mice. To analyze the commitment status o f the D N l C D 122" cells, they cultured them in the progenitor culture for a single cell assay. A s previously reported, the W T D N l C D 122" progenitors contained a l l three progenitors (pT, N K P s , and T N K P s ) but the Id2" D N l C D 122" /_  progenitors contained only pT cells. Interestingly, the number o f pT i n the D N l C D 122"  progenitors o f 162''' mice corresponded to the total number o f pT, N K P , and T N K P s cells o f W T mice. Loss o f Id2" failed to support commitment o f bipotent precursors to N K P s and /_  therefore they all committed to the T cell lineage  . E proteins promote the T cell lineage and  their functional inactivation by Id promotes the N K cell lineage. While lack o f E proteins inhibits T cell commitment, too high levels o f E proteins can cause G l arrest at stages when this is not required (i.e. proliferation stages), therefore, a balance o f E proteins and Id proteins is important. Interestingly, rearrangement o f specific fetal types o f T C R y and T C R 5 chains are facilitated by lower levels o f activity o f E 2 A and are disfavored by higher l e v e l s . 163  Both Notch signaling and E protein activity are crucial for the D N stages o f T cell development but at a later stage, the y5 T cell lineage does not require as high levels. It seems that there is a continuum in levels required between a P T cells, y8 T cells, and N K cells. Therefore, in D N 1 and D N 2 precursors, a weak Notch signal and l o w E protein/high Id levels appear to be crucial in producing thymus-dependent N K cells.  5.7. M e d i c a l relevance A pathway that produces T C R y N K cells in humans is likely. A rare blastic N K - c e l l like +  leukemia/lymphoma with T C R y gene rearrangement has been described i n several patients " 164  1 6 6  . These T C R y N K cells may be equivalent to the population we have described in mice. +  It has been recently reported that 3.2-36% o f human peripheral N K cells have incomplete T C R 5 rearrangements . T C R y rearrangements on the other hand were much rarer. These 167  Tcrtf cells likely occur through a thymus-dependent pathway shared with early T cell progenitors. In humans, the earliest T cell progenitor in the thymus ( C D 3 4 C D l a " ) has begun +  V ( D ) J recombination with immature T C R 8 gene rearrangements although they still retain N K 106  cell and D C potential . In humans, T C R y rearrangements do not begin until the next stage 168  ( C D 3 4 C D l a ) when N K cell potential is much l o w e r +  +  169  . Therefore, this appears to be very  similar to what we have described in mice, with N K cells splitting from the T cell lineage at the very early stages o f V(D)J recombination, except with  T C R 8 rearrangement  first. It seems  likely that some N K cells still branch at the later stage since the T C R y N K cell +  leukemia/lymphoma have been described.  5.8. F i n a l conclusions A t least two pathways o f lineage commitment produce N K cells i n the adult mouse. We discovered a thymus-dependent pathway that has shared progenitors with T cells and T C R rearrangement. This thymus and LN-specific pathway gives rise to N K cells with unique phenotype and functional capabilities. It is likely that thymus-derived T cell precursors migrate to the L N where they give rise to N K cells although it is still possible that thymus N K cells migrate to the L N and further mature. This study refutes the idea that all N K cells develop in the B M and it also brings to light the possibility that multiple pathways o f N K cell development may occur i n multiple locations. The division o f N K cells into distinct subsets is relatively new and it is possible that the subsets of N K cells are the result o f multiple developmental environments. W e hypothesize that differences in tissue environment influence the final N K cell phenotype and function, resulting in multiple subsets of N K cells throughout the body. This may be how the body produces distinct N K cell subsets that may each play specific roles in the immune response. For example, the thymus-dependent N K cells may fill a specific niche which requires them to have low levels o f M H C - c l a s s I inhibitory receptors and IFNy production. The B M pathway may be unable to fill this niche since its environment produces N K cells with a different phenotype (i.e. B M stroma stimulates higher  107  L y 4 9 receptor  expression levels). Future characterization o f N K cell subsets and the environments that produce them may reveal developmental triggers specific to each subset. If different subsets play distinct roles i n the immune response, the specific N K cell subset that is best suited to combat the threat at hand (i.e. tumors, pathogens, viruses) may be used i n treatments by stimulating their development or proliferation.  108  BIBLIOGRAPHY 1.  Smyth, M . J. et al. Activation o f N K cell cytotoxicity. M o l Immunol 42, 501-10 (2005).  2.  Cooper, M . A . et al. In vivo evidence for a dependence on interleukin 15 for survival o f natural killer cells. B l o o d 100, 3633-8 (2002).  3.  Kennedy, M . et al. A common precursor for primitive erythropoiesis and definitive haematopoiesis. Nature 386, 488-93 (1997).  4.  Screpanti, V . , W a l l i n , R. P., Ljunggren, H . G . & Grandien, A . A central role for death receptor-mediated apoptosis in the rejection o f tumors by N K cells. J Immunol 167, 2068-73 (2001).  5.  Takeda, K . et al. T R A I L identifies immature natural killer cells i n newborn mice and adult mouse liver. B l o o d 105, 2082-9 (2005).  6.  Hayakawa, Y . & Smyth, M . J. N K G 2 D and cytotoxic effector function in tumor immune surveillance. Semin Immunol 18, 176-85 (2006).  7.  Raulet, D . H . & Vance, R. E . Self-tolerance o f natural killer cells. Nat Rev Immunol 6, 520-31 (2006).  8.  Veinotte, L . L . , W i l h e l m , B . T., Mager, D . L . & Takei, F. Acquisition o f MHC-specific receptors on murine natural killer cells. Crit Rev Immunol 23, 251-66 (2003).  9.  Lodoen, M . B . & Lanier, L . L . V i r a l modulation o f N K cell immunity. Nat Rev Microbiol 3, 59-69 (2005).  10.  K i m , S. et al. In vivo developmental stages in murine natural killer cell maturation. Nat Immunol 3, 523-8 (2002).  11.  Mandelboim, O. et al. Recognition o f haemagglutinins on virus-infected cells by N K p 4 6 activates lysis by human N K cells. Nature 409, 1055-60 (2001).  12.  X i o n g , N . , Baker, J. E . , Kang, C . & Raulet, D . H . The genomic arrangement o f T cell receptor variable genes is a determinant o f the developmental rearrangement pattern. Proc Natl A c a d Sci U S A 101, 260-5 (2004).  13.  Lodoen, M . B . & Lanier, L . L . Natural killer cells as an initial defense against pathogens. Curr Opin Immunol 18, 391-8 (2006).  14.  Degli-Esposti, M . A . & Smyth, M . J. Close encounters o f different kinds: dendritic cells and N K cells take centre stage. Nat Rev Immunol 5, 112-24 (2005).  109  15.  Andrews, D . M . , Scalzo, A . A . , Yokoyama, W . M . , Smyth,.M. J. & Degli-Esposti, M . A . Functional interactions between dendritic cells and N K cells during viral infection. Nat Immunol 4, 175-81 (2003).  16.  Weissman, I. L . , Anderson, D . J. & Gage, F. Stem and progenitor cells: origins, phenotypes, lineage commitments, and transdifferentiations. A n n u Rev Cell Dev B i o l 17,387-403 (2001).  17.  Adolfsson, J. et al. Upregulation o f Flt3 expression within the bone marrow Lin()Scal(+)c-kit(+) stem cell compartment is accompanied by loss o f self-renewal capacity. Immunity 15, 659-69 (2001).  18.  Akashi, K . , Traver, D . , Miyamoto, T. & Weissman, I. L . A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404, 193-7 (2000).  19.  Kondo, M . , Weissman, I. L . & Akashi, K . Identification o f clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91, 661-72 (1997).  20.  Rothenberg, E . V . & Dionne, C . J. Lineage plasticity and commitment in T-cell development. Immunol Rev 187, 96-115 (2002).  21.  Akashi, K . et al. Transcriptional accessibility for genes o f multiple tissues and hematopoietic lineages is hierarchically controlled during early hematopoiesis. Blood 101,383-9(2003).  22.  Kumar, V . , Ben-Ezra, J., Bennett, M . & Sonnenfeld, G . Natural killer cells in mice treated with 89strontium: normal target-binding cell numbers but inability to k i l l even after interferon administration. J Immunol 123, 1832-8 (1979).  23.  Seaman, W . E . , Gindhart, T. D . , Greenspan, J. S., Blackman, M . A . & Talal, N . Natural killer cells, bone, and the bone marrow: studies in estrogen-treated mice and in congenitally osteopetrotic (mi/mi) mice. J Immunol 122, 2541-7 (1979).  24.  Wiktor-Jedrzejczak, W . W . , Ahmed, A . , Szczylik, C . & Skelly, R. R. Hematological characterization o f congenital osteopetrosis in op/op mouse. Possible mechanism for abnormal macrophage differentiation. J E x p M e d 156, 1516-27 (1982).  25.  Waskow, C , Paul, S., Haller, C , Gassmann, M . & Rodewald, H . Viable c - K i t ( W / W ) mutants reveal pivotal role for c-kit i n the maintenance o f lymphopoiesis. Immunity 17, 277-88 (2002).  26.  Waskow, C . & Rodewald, H . R. Lymphocyte development i n neonatal and adult c-Kitdeficient ( c - K i t W / W ) mice. A d v E x p M e d B i o l 512, 1-10 (2002).  110  27.  Borghesi, L . et al. B lineage-specific regulation o f V ( D ) J recombinase activity is established in common lymphoid progenitors. J E x p M e d 199, 491-502 (2004).  28.  Igarashi, H . , Gregory, S. C . , Yokota, T., Sakaguchi, N . & Kincade, P. W . Transcription from the R A G 1 locus marks the earliest lymphocyte progenitors in bone marrow. Immunity 17, 117-30 (2002).  29.  Kouro, T., Kumar, V . & Kincade, P. W . Relationships between early B - and N K lineage lymphocyte precursors in bone marrow. B l o o d 100, 3672-80 (2002).  30.  Rosmaraki, E . E . et al. Identification o f committed N K cell progenitors i n adult murine bone marrow. Eur J Immunol 31, 1900-9 (2001).  31.  G i r i , J. G . et al. Utilization o f the beta and gamma chains o f the I L - 2 receptor by the novel cytokine IL-15. Embo J 13, 2822-30 (1994).  32.  Vosshenrich, C . A . & D i Santo, J. P. Interleukin signaling. Curr B i o l 12, R760-3 (2002).  33.  Douagi, I., Colucci, F., D i Santo, J. P. & Cumano, A . Identification o f the earliest prethymic bipotent T / N K progenitor in murine fetal liver. B l o o d 99, 463-71 (2002).  34.  Ikawa, T. et al. Identification o f the earliest prethymic T-cell progenitors in murine fetal blood. B l o o d 103, 530-7 (2004).  35.  Masuda, K . et al. Thymic anlage is colonized by progenitors restricted to T, N K , and dendritic cell lineages. J Immunol 174, 2525-32 (2005).  36.  Balciunaite, G . , Ceredig, R. & Rolink, A . G . The earliest subpopulation o f mouse thymocytes contains potent T, significant macrophage, and natural killer cell but no B lymphocyte potential. B l o o d 105, 1930-6 (2005).  37.  Wang, B . et al. A block in both early T lymphocyte and natural killer cell development in transgenic mice with high-copy numbers o f the human C D 3 E gene. Proc Natl Acad S c i U S A 9 1 , 9402-6(1994).  38.  Flamand, V . et al. Delayed maturation o f C D 4 - C D 8 - F c gamma RII/III+ T and natural killer cell precursors in F c epsilon RI gamma transgenic mice. J E x p M e d 184, 1725-35 (1996).  39.  Tabatabaei-Zavareh, N . V . , Linnea L . ; Fumio Takei. in Altered Immunoregulation and Human Disease (ed. Gorczynski, R. M . ) 33-59 (Research Signpost, India, 2005).  40.  Williams, N . S. et al. Generation o f lytic natural killer 1.1+, L y - 4 9 - cells from multipotential murine bone marrow progenitors in a stroma-free culture: definition o f  111  cytokine requirements and developmental intermediates. J E x p M e d 186, 1609-14 (1997). 41.  Williams, N . S. et al. Differentiation o f N K 1 . 1 + , Ly49+ N K cells from flt3+ multipotent marrow progenitor cells. J Immunol 163, 2648-56 (1999).  42.  Vosshenrich, C . A . et al. Roles for common cytokine receptor gamma-chain-dependent cytokines i n the generation, differentiation, and maturation o f N K cell precursors and peripheral N K cells i n vivo. J Immunol 174, 1213-21 (2005).  43.  Robbins, S. H . , Tessmer, M . S., Mikayama, T. & Brossay, L . Expansion and contraction o f the N K cell compartment i n response to murine cytomegalovirus infection. J Immunol 173, 259-66 (2004).  44.  Moore, T. A . , von Freeden-Jeffry, U . , Murray, R. & Zlotnik, A . Inhibition o f gamma delta T cell development and early thymocyte maturation in IL-7 -/- mice. J Immunol 157, 2366-73 (1996).  45.  He, Y . W . & Malek, T. R. Interleukin-7 receptor alpha is essential for the development of gamma delta + T cells, but not natural killer cells. J E x p M e d 184, 289-93 (1996).  46.  D i Santo, J. P. Natural killer cell developmental pathways: a question o f balance. A n n u Rev Immunol 24, 257-86 (2006).  47.  M c K e n n a , H . J. et al. M i c e lacking flt3 ligand have deficient hematopoiesis affecting hematopoietic progenitor cells, dendritic cells, and natural killer cells. Blood 95, 348997 (2000).  48.  Colucci, F. & D i Santo, J. P. The receptor tyrosine kinase c-kit provides a critical signal for survival, expansion, and maturation o f mouse natural killer cells. B l o o d 95, 984-91 (2000).  49.  Lodolce, J. P. et al. IL-15 receptor maintains lymphoid homeostasis by supporting lymphocyte homing and proliferation. Immunity 9, 669-76 (1998).  50.  Cao, X . et al. Defective lymphoid development in mice lacking expression o f the common cytokine receptor gamma chain. Immunity 2, 223-38 (1995).  51.  DiSanto, J. P., Muller, W . , Guy-Grand, D . , Fischer, A . & Rajewsky, K . Lymphoid development i n mice with a targeted deletion o f the interleukin 2 receptor gamma chain. Proc N a t l A c a d Sci U S A 92, 377-81 (1995).  52.  Kennedy, M . K . et al. Reversible defects in natural killer and memory C D 8 T cell lineages in interleukin 15-deficient mice. J E x p M e d 191, 771-80 (2000).  112  53.  Wang, J. H . et al. Selective defects in the development o f the fetal and adult lymphoid system in mice with an Ikaros null mutation. Immunity 5, 537-49 (1996).  54.  Nichogiannopoulou, A . , Trevisan, M . , Neben, S., Friedrich, C . & Georgopoulos, K . Defects i n hemopoietic stem cell activity in Ikaros mutant mice. J E x p M e d 190, 120114(1999).  55.  Scott, E . W . , Simon, M . C , Anastasi, J. & Singh, H . Requirement o f transcription factor P U . l in the development o f multiple hematopoietic lineages. Science 265, 15737(1994).  56.  DeKoter, R. P., Lee, H . J. & Singh, H . P U . l regulates expression o f the interleukin-7 receptor in lymphoid progenitors. Immunity 16, 297-309 (2002).  57.  Colucci, F. et al. Differential requirement for the transcription factor P U . 1 in the generation o f natural killer cells versus B and T cells. B l o o d 97, 2625-32 (2001).  58.  Barton, K . et al. The E t s - l transcription factor is required for the development o f natural killer cells i n mice. Immunity 9, 555-63 (1998).  59.  Bories, J. C . et al. Increased T-cell apoptosis and terminal B - c e l l differentiation induced by inactivation o f the E t s - l proto-oncogene. Nature 377, 635-8 (1995).  60.  Heemskerk, M . H . et al. Inhibition o f T cell and promotion o f natural killer cell development by the dominant negative helix loop helix factor Id3. J E x p M e d 186, 1597-602 (1997).  61.  Spits, H . , Couwenberg, F., Bakker, A . Q., Weijer, K . & Uittenbogaart, C . H . Id2 and Id3 inhibit development o f CD34(+) stem cells into predendritic cell (pre-DC)2 but not into p r e - D C l . Evidence for a lymphoid origin o f pre-DC2. J E x p M e d 192, 1775-84 (2000).  62.  Yokota, Y . et al. Development o f peripheral lymphoid organs and natural killer cells depends on the helix-loop-helix inhibitor Id2. Nature 397, 702-6 (1999).  63.  Ikawa, T., Kawamoto, H . , Fujimoto, S. & Katsura, Y . Commitment o f common T/Natural killer ( N K ) progenitors to unipotent T and N K progenitors i n the murine fetal thymus revealed by a single progenitor assay. J E x p M e d 190, 1617-26 (1999).  64.  Kubota, A . , Kubota, S., Lohwasser, S., Mager, D . L . & Takei, F. Diversity of N K cell receptor repertoire in adult and neonatal mice. J Immunol 163, 212-6 (1999).  65.  Fraser, K . P. et al. N K cells developing in vitro from fetal mouse progenitors express at least one member o f the L y 4 9 family that is acquired in a time-dependent and  113  stochastic manner independently o f C D 9 4 and N K G 2 . Eur J Immunol 32, 868-78 (2002). 66.  V a n Beneden, K . et al. Expression of L y 4 9 E and C D 9 4 / N K G 2 on fetal and adult N K cells. J Immunol 166, 4302-11 (2001).  67.  Salcedo, M . et al. Role o f Qa-l(b)-binding receptors in the specificity o f developing N K cells. Eur J Immunol 30, 1094-101 (2000).  68.  Waldmann, T. A . & Tagaya, Y . The multifaceted regulation o f interleukin-15 expression and the role o f this cytokine in N K cell differentiation and host response to intracellular pathogens. A n n u Rev Immunol 17, 19-49 (1999).  69.  Tassara, C . , Pepper, A . E . & Puck, J. M . Intronic point mutation i n the I L 2 R G gene causing X - l i n k e d severe combined immunodeficiency. H u m M o l Genet 4, 1693-5 (1995).  70.  Pepper, A . E . , Buckley, R. H . , Small, T. N . & Puck, J. M . T w o mutational hotspots in the interleukin-2 receptor gamma chain gene causing human X - l i n k e d severe combined immunodeficiency. A m J H u m Genet 57, 564-71 (1995).  71.  Minagawa, M . et al. Enforced expression o f Bcl-2 restores the number o f N K cells, but does not rescue the impaired development o f N K T cells or intraepithelial lymphocytes, in IL-2/IL-15 receptor beta-chain-deficient mice. J Immunol 169, 4153-60 (2002).  72.  Ranson, T. et al. IL-15 is an essential mediator o f peripheral N K - c e l l homeostasis. B l o o d 101,4887-93 (2003).  73.  Grund, E . M . , Spyropoulos, D . D . , Watson, D . K . & Muise-Helmericks, R. C . Interleukins 2 and 15 regulate E t s l expression v i a E R K 1 / 2 and M N K 1 in human natural killer cells. J B i o l Chem 280, 4772-4778 (2005).  74.  Ito, D . , Back, T. C . , Shakhov, A . N . , Wiltrout, R. H . & Nedospasov, S. A . M i c e with a targeted mutation in lymphotoxin-alpha exhibit enhanced tumor growth and metastasis: impaired N K cell development and recruitment. J Immunol 163, 2809-15 (1999).  75.  Iizuka, K . et al. Requirement for membrane lymphotoxin in natural killer cell development. Proc Natl A c a d Sci U S A 96, 6336-40 (1999).  76.  W u , Q. et al. Signal v i a lymphotoxin-beta R on bone marrow stromal cells is required for an early checkpoint o f N K cell development. J Immunol 166, 1684-9 (2001).  77.  Lian, R. H . et al. A role for lymphotoxin in the acquisition o f L y 4 9 receptors during N K cell development. Eur J Immunol 34, 2699-707 (2004).  114  78.  Ogasawara, K . et al. Requirement for IRF-1 in the microenvironment supporting development o f natural killer cells. Nature 391, 700-3 (1998).  79.  Ohteki, T. et al. The transcription factor interferon regulatory factor 1 (IRF-1) is important during the maturation o f natural killer 1.1+ T cell receptor-alpha/beta+ ( N K 1 + T) cells, natural killer cells, and intestinal intraepithelial T cells. J E x p M e d 187,967-72(1998).  80.  Caraux, A . et al. Natural killer cell differentiation driven by Tyro3 receptor tyrosine kinases. Nat Immunol 7, 747-54 (2006).  81.  Lee, K . N . et al. V D U P 1 is required for the development o f natural killer cells. Immunity 22, 195-208 (2005).  82.  Lacorazza, H . D . et al. The E T S protein M E F plays a critical role i n perforin gene expression and the development o f natural killer and N K - T cells. Immunity 17, 437-49 (2002).  83.  L i , Q. & Verma, I. M . NF-kappaB regulation in the immune system. Nat Rev Immunol 2, 725-34 (2002).  84.  Caamano, J. & Hunter, C . A . NF-kappaB family o f transcription factors: central regulators o f innate and adaptive immune functions. C l i n M i c r o b i o l Rev 15, 414-29 (2002).  85.  Samson, S. I. et al. Combined deficiency in IkappaBalpha and IkappaBepsilon reveals a critical window o f NF-kappaB activity in natural killer cell differentiation. Blood 103, 4573-80 (2004).  86.  Townsend, M . J. et al. T-bet regulates the terminal maturation and homeostasis of N K and V a l p h a H i N K T cells. Immunity 20, 477-94 (2004).  87.  Pearce, E . L . et al. Control o f effector C D 8 + T cell function by the transcription factor Eomesodermin. Science 302, 1041-3 (2003).  88.  Samson, S. I. et al. G A T A - 3 promotes maturation, IFN-gamma production, and liverspecific homing o f N K cells. Immunity 19, 701-11 (2003).  89.  Murphy, K . M . & Reiner, S. L . The lineage decisions o f helper T cells. Nat Rev Immunol 2, 933-44 (2002).  90.  Mullen, A . C . et al. H l x is induced by and genetically interacts with T-bet to promote heritable T(H)1 gene induction. Nat Immunol 3, 652-8 (2002).  115  91.  Ito, A . et al. Inhibitory effect on natural killer activity o f microphthalmia transcription factor encoded by the mutant m i allele o f mice. B l o o d 97, 2075-83 (2001).  92.  Orange, J. S. et al. Deficient natural killer cell cytotoxicity in patients with I K K g a m m a / N E M O mutations. J C l i n Invest 109, 1501-9 (2002).  93.  Kaisho, T. et al. Impairment o f natural killer cytotoxic activity and interferon gamma production in C C A A T / e n h a n c e r binding protein gamma-deficient mice. J E x p M e d 190,1573-82 (1999).  94.  Brady, J., Hayakawa, Y . , Smyth, M . J. & Nutt, S. L . IL-21 induces the functional maturation o f murine N K cells. J Immunol 172, 2048-58 (2004).  95.  Janeway, C . A . T., Paul; Walport, Mark; Schlomchik, Mark. Immunobiology (Garland Science, N e w Y o r k and London, 2001).  96.  Carding, S. R. & Egan, P. J. Gammadelta T cells: functional plasticity and heterogeneity. Nat Rev Immunol 2, 336-45 (2002).  97.  Hayes, S. M . & Love, P. E . Distinct structure and signaling potential o f the gamma delta T C R complex. Immunity 16, 827-38 (2002).  98.  Pennington, D . J., Silva-Santos, B . & Hayday, A . C . Gammadelta T cell developmenthaving the strength to get there. Curr Opin Immunol 17, 108-15 (2005).  99.  Kronenberg, M . Toward an understanding o f N K T cell biology: progress and paradoxes. A n n u Rev Immunol 23, 877-900 (2005).  100.  Capone, M . , Hockett, R. D . , Jr. & Zlotnik, A . Kinetics o f T cell receptor beta, gamma, and delta rearrangements during adult thymic development: T cell receptor rearrangements are present in CD44(+)CD25(+) Pro-T thymocytes. Proc Natl Acad Sci U S A 95, 12522-7(1998).  101.  Livak, F., Tourigny, M . , Schatz, D . G . & Petrie, H . T. Characterization o f T C R gene rearrangements during adult murine T cell development. J Immunol 162, 2575-80 (1999).  102.  Hayday, A . C . et al. Structure, organization, and somatic rearrangement o f T cell gamma genes. Cell 40, 259-69 (1985).  103.  Garman, R. D . , Doherty, P. J. & Raulet, D . H . Diversity, rearrangement, and expression of murine T cell gamma genes. Cell 45, 733-42 (1986).  104.  Dudley, E . C , Girardi, M . , Owen, M . J. & Hayday, A . C . A l p h a beta and gamma delta T cells can share a late common precursor. Curr B i o l 5, 659-69 (1995).  116  105.  Kang, J. & Raulet, D . H . Events that regulate differentiation o f alpha beta T C R + and gamma delta T C R + T cells from a common precursor. Semin Immunol 9, 171-9 (1997).  106.  Livak, F., Petrie, H . T., Crispe, I. N . & Schatz, D . G . In-frame T C R delta gene rearrangements play a critical role i n the alpha beta/gamma delta T cell lineage decision. Immunity 2, 617-27 (1995).  107.  Wilson, A . , de Villartay, J. P. & MacDonald, H . R. T cell receptor delta gene rearrangement and T early alpha ( T E A ) expression in immature alpha beta lineage thymocytes: implications for alpha beta/gamma delta lineage commitment. Immunity 4, 37-45 (1996).  108.  Dudley, E . C , Petrie, H . T., Shah, L . M . , Owen, M . J. & Hayday, A . C . T cell receptor beta chain gene rearrangement and selection during thymocyte development in adult mice. Immunity 1, 83-93 (1994).  109.  Burtrum, D . B . , K i m , S., Dudley, E . C , Hayday, A . C . & Petrie, H . T. T C R gene recombination and alpha beta-gamma delta lineage divergence: productive TCR-beta rearrangement is neither exclusive nor preclusive o f gamma delta cell development. J Immunol 157, 4293-6 (1996).  110.  Mertsching, E . & Ceredig, R. T cell receptor-gamma, delta-expressing fetal mouse thymocytes are generated without T cell receptor V beta selection. Eur J Immunol 26, 804-10 (1996).  111.  Vicari, A . P., M o c c i , S., Openshaw, P., O'Garra, A . & Zlotnik, A . Mouse gamma delta T C R + N K 1 . 1 + thymocytes specifically produce interleukin-4, are major histocompatibility complex class I independent, and are developmentally related to alpha beta T C R + N K 1 . 1 + thymocytes. Eur J Immunol 26, 1424-9 (1996).  112.  Gellert, M . V ( D ) J recombination: R A G proteins, repair factors, and regulation. A n n u Rev Biochem 71, 101-32 (2002).  113.  Bergman, Y . , Fisher, A . & Cedar, H . Epigenetic mechanisms that regulate antigen receptor gene expression. Curr Opin Immunol 15, 176-81 (2003).  114.  Candeias, S., Peschon, J. J., Muegge, K . & Durum, S. K . Defective T-cell receptor gamma gene rearrangement in interleukin-7 receptor knockout mice. Immunol Lett 57, 9-14(1997).  117  115.  Huang, J., Durum, S. K . & Muegge, K . Cutting edge: histone acetylation and recombination at the T C R gamma locus follows IL-7 induction. J Immunol 167, 6073-7 (2001).  116.  Schlissel, M . S., Durum, S. D . & Muegge, K . The interleukin 7 receptor is required for T cell receptor gamma locus accessibility to the V ( D ) J recombinase. J E x p M e d 191, 1045-50 (2000).  117.  Heilig, J. S. & Tonegawa, S. Diversity o f murine gamma genes and expression in fetal and adult T lymphocytes. Nature 322, 836-40 (1986).  118.  Pereira, P . et al. Rearrangement and expression o f V z e t a l , Vzeta2 and Vzeta3 T C R zeta genes in C 5 7 B L / 6 mice. Int Immunol 8, 83-90 (1996).  119.  Raulet, D . H . et al. Control o f gamma delta T-cell development. Immunol Rev 120, 185-204(1991).  120.  Goldman, J. P., Spencer, D . M . & Raulet, D . H . Ordered rearrangement o f variable region genes o f the T cell receptor gamma locus correlates with transcription o f the unrearranged genes. J E x p M e d 177, 729-39 (1993).  121.  Havran, W . L . & A l l i s o n , J. P. Developmentally ordered appearance o f thymocytes expressing different T-cell antigen receptors. Nature 335, 443-5 (1988).  122.  Ito, K . et al. Different gamma delta T-cell receptors are expressed on thymocytes at different stages o f development. Proc Natl Acad Sci U S A 86, 631-5 (1989).  123.  A l l i s o n , J. P. & Havran, W . L . The immunobiology o f T cells with invariant gamma delta antigen receptors. A n n u Rev Immunol 9, 679-705 (1991).  124.  Boismenu, R. & Havran, W . L . Modulation o f epithelial cell growth by intraepithelial gamma delta T cells. Science 266, 1253-5 (1994).  125.  Raulet, D . H . The structure, function, and molecular genetics o f the gamma/delta T cell receptor. A n n u Rev Immunol 7, 175-207 (1989).  126.  Baker, J. E . , Cado, D . & Raulet, D . H . Developmentally programmed rearrangement o f T cell receptor V g a m m a genes is controlled by sequences immediately upstream of the Vgamma genes. Immunity 9, 159-68 (1998).  127.  Pereira, P. & Boucontet, L . Rates o f recombination and chain pair biases greatly influence the primary gammadelta T C R repertoire i n the thymus o f adult mice. J Immunol 173, 3261-70 (2004).  118  128.  Kang, J., Baker, J. & Raulet, D . H . Evidence that productive rearrangements o f T C R gamma genes influence the commitment o f progenitor cells to differentiate into alpha beta or gamma delta T cells. Eur J Immunol 25, 2706-9 (1995).  129.  Itohara, S. et al. T cell receptor delta gene mutant mice: independent generation o f alpha beta T cells and programmed rearrangements o f gamma delta T C R genes. Cell 72, 337-48 (1993).  130.  Kang, H . S. et al. Stage-dependent gene expression profiles during natural killer cell development. Genomics 86, 551-65 (2005).  131.  Gallagher, M . et al. Use o f T C R A D V gene segments by the delta chain is independent of their position and o f C D 3 expression. Eur J Immunol 28, 3878-85 (1998).  132.  Pereira, P. et al. Developmentally regulated and lineage-specific rearrangement of T cell receptor Valpha/delta gene segments. Eur J Immunol 30, 1988-97 (2000).  133.  Born, W . , White, J., Kappler, J. & Marrack, P. Rearrangement o f IgH genes in normal thymocyte development. J Immunol 140, 3228-32 (1988)!  134.  Krejci, O., Prouzova, Z . , Horvath, O., Trka, J. & Hrusak, O. Cutting edge: T C R delta gene is frequently rearranged in adult B lymphocytes. J Immunol 171, 524-7 (2003).  135.  Terszowski, G . et al. Evidence for a functional second thymus in mice. Science 312, 284-7 (2006).  136.  Krangel, M . S., Carabana, J., Abbarategui, I., Schlimgen, R. & Hawwari, A . Enforcing order within a complex locus: current perspectives on the control o f V ( D ) J recombination at the murine T-cell receptor alpha/delta locus. Immunol Rev 200, 22432 (2004).  137.  Biron, C . A . et al. Murine natural killer cells stimulated in vivo do not express the T cell receptor alpha, beta, gamma, T3 delta, or T3 epsilon genes. J Immunol 139, 170410 (1987).  138.  Tutt, M . M . et al. Murine natural killer cells do not express functional transcripts o f the alpha-, beta-, or gamma-chain genes o f the T cell receptor. J Immunol 137, 2998-3001 (1986).  139.  Tutt, M - M . et al. T cell receptor genes do not rearrange or express functional transcripts in natural killer cells o f scid mice. J Immunol 138, 2338-44 (1987).  119  140.  Schmitt, T. M . , Ciofani, M . , Petrie, H . T. & Zuniga-Pflucker, J. C . Maintenance of T cell specification and differentiation requires recurrent notch receptor-ligand interactions. J E x p M e d 200, 469-79 (2004).  141.  Terra, R. et al. T-cell generation by lymph node resident progenitor cells. Blood 106, 193-200(2005).  142.  Lambolez, F. et al. The thymus exports long-lived fully committed T cell precursors that can colonize primary lymphoid organs. Nat Immunol 7, 76-82 (2006).  143.  Freud, A . G . et al. A human CD34(+) subset resides i n lymph nodes and differentiates into CD56bright natural killer cells. Immunity 22, 295-304 (2005).  144.  Chen, S., Kawashima, H . , Lowe, J. B . , Lanier, L . L . & Fukuda, M . Suppression o f tumor formation i n lymph nodes by L-selectin-mediated natural killer cell recruitment. J E x p M e d 202, 1679-89 (2005).  145.  Schmidt, K . N . et al. APC-independent activation o f N K cells by the Toll-like receptor 3 agonist double-stranded R N A . J Immunol 172, 138-43 (2004).  146.  Carlyle, J. R., M i c h i e , A . M . , Cho, S. K . & Zuniga-Pflucker, J. C . Natural killer cell development and function precede alpha beta T cell differentiation in mouse fetal thymic ontogeny. J Immunol 160, 744-53 (1998).  147.  Porritt, H . E . et al. Heterogeneity among D N l prothymocytes reveals multiple progenitors with different capacities to generate T cell and non-T cell lineages. Immunity 20, 735-45 (2004).  148.  Hayakawa, Y . & Smyth, M . J. C D 2 7 dissects mature N K cells into two subsets with distinct responsiveness and migratory capacity. J Immunol 176, 1517-24 (2006).  149.  Farag, S. S. & Caligiuri, M . A . Human natural killer cell development and biology. Blood Rev 20, 123-37 (2006).  150.  Freud, A . G . et al. Evidence for discrete stages o f human natural killer cell differentiation in vivo. J E x p M e d 203, 1033-43 (2006).  151.  Ballas, Z . K . & Rasmussen, W . Lymphokine-activated killer ( L A K ) cells. III. Characterization o f L A K precursors and susceptible target cells within the murine thymus. J Immunol 139, 3542-9 (1987).  152.  Scollay, R., Jacobs, S., Jerabek, L . , Butcher, E . & Weissman, I. T cell maturation: thymocyte and thymus migrant subpopulations defined with monoclonal antibodies to M H C region antigens, j Immunol 124, 2845-53 (1980).  120  153.  Savino, W . The thymus is a common target organ i n infectious diseases. P L o S Pathog 2, e62 (2006).  154.  Martin-Fontecha, A . et al. Induced recruitment o f N K cells to lymph nodes provides IFN-gamma for T(H)1 priming. Nat Immunol 5, 1260-5 (2004).  155.  Bajenoff, M . et al. Natural killer cell behavior in lymph nodes revealed by static and real-time imaging. J E x p M e d 203, 619-31 (2006).  156.  Lehar, S. M . , Dooley, J., Fair, A . G . & Bevan, M . J. N o t c h ligands Delta 1 and Jaggedl transmit distinct signals to T-cell precursors. B l o o d 105, 1440-7 (2005).  157.  DeHart, S. L . , Heikens, M . J. & Tsai, S. Jagged2 promotes the development o f natural killer cells and the establishment o f functional natural killer cell lines. B l o o d 105, 3521-7 (2005).  158.  Quong, M . W . , Romanow, W . J. & Murre, C . E protein function i n lymphocyte development. A n n u Rev Immunol 20, 301-22 (2002).  159.  Sawada, S. & Littman, D . R. A heterodimer o f H E B and an E12-related protein interacts with the C D 4 enhancer and regulates its activity in T-cell lines. M o l Cell B i o l 13,5620-8 (1993).  160.  Petersson, K . , Ivars, F. & Sigvardsson, M . The pT alpha promoter and enhancer are direct targets for transactivation by E box-binding proteins. Eur J Immunol 32, 911-20 (2002).  161.  Benezra, R., Davis, R. L . , Lockshon, D . , Turner, D . L . & Weintraub, H . The protein Id: a negative regulator o f helix-loop-helix D N A binding proteins. C e l l 61, 49-59 (1990).  162.  Ikawa, T., Fujimoto, S., Kawamoto, H . , Katsura, Y . & Yokota, Y . Commitment to natural killer cells requires the helix-loop-helix inhibitor Id2. Proc Natl A c a d Sci U S A 98,5164-9 (2001).  163.  Bain, G . , Romanow, W . J., Albers, K . , Havran, W . L . & Murre, C . Positive and negative regulation o f V ( D ) J recombination by the E 2 A proteins. J E x p M e d 189, 289300 (1999).  164.  Aoyama, Y . et al. Blastic N K - c e l l lymphoma/leukemia with T-cell receptor gamma rearrangement. A n n Hematol 80, 752-4 (2001).  165.  L i u , X . Y . , Atkins, R. C , Feusner, J. H . & Rowland, J. M . Blastic N K - c e l l - l i k e lymphoma with T-cell receptor gene rearrangement. A m J Hematol 75, 251-3 (2004).  121  166.  L i n , C . W . , Lee, W . H . , Chang, C . L . , Yang, J. Y . & H s u , S. M . Restricted killer cell immunoglobulin-like receptor repertoire without T-cell receptor gamma rearrangement supports a true natural killer-cell lineage i n a subset o f sinonasal lymphomas. A m J Pathol 159, 1671-9 (2001).  167.  Fronkova, E . et al. Lymphoid differentiation pathways can be traced by T C R delta rearrangements. J Immunol 175, 2495-500 (2005).  168.  B l o m , B . et al. T C R gene rearrangements and expression o f the pre-T cell receptor complex during human T-cell differentiation. B l o o d 93, 3033-43 (1999).  169.  Spits, H . et al. Early stages in the development o f human T, natural killer and thymic dendritic cells. Immunol Rev 165, 75-86 (1998).  122  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100554/manifest

Comment

Related Items