FUNCTIONS OF CD45 IN TCR SIGNALING IN CD4+CDS+ DOUBLE-POSITIVE THYMOCYTES
Gordon W. Cheng
A thesis submitted in conformity with the requirements for the degree of Master's of Science Graduate Department of Immunology University of Toronto
O Copyright by Gordon W. Cheng (1997) Acquisitions and Acquisitions et Bibliographie Services semices bibliographiques 395 Wellington Street 395, rue Wellington Onawa ON KIA ON4 Onawa ON K1A ON4 Canada Canada Your fib Volro re:erence
Our Ida Norre iefeisnce
The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sel1 reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de rnicrofiche/film, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or othenvise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. - a Thymocytes Master's of Science (1997) Gordon W. Cheng Graduate Department of Immunology University of Toronto
Abstract
T ceIl receptor (TCR) signals are essential for normal T cell developrnent, and CD45 is thought
10 be essential for coupling TCR to the intracellular signaling rnachinery. However, T cell
development is only partially compromised in CD45-deficient mice. This thesis describes a
CD45-deficient CD4+CD8+double-posi tive (DP) thymoma called 3T7. 1 show that in 3T7 ce1 ls and DP thymocytes CD45 is necessary for the TCR-induced protein tyrosine phosphorylation, as well as changes in CD5, MG-1 and CD41CD8 expression levels. However, CO-aggregation of TCR and either CD4 or CD8 induced signaling events in a CD45-independent rnanner, providing a ralionale for the developmental phenotype observed in CD45-dericient mice.
Surprisingly, 1 found no disferences in the overall phosphorylation slatc of Lck or Fyn in CD45
deficient cells versus CD45-positive cells. However, 1 show that loss of CD45 is accornpaiiied by a hyperphosphorylation of TCRS in 3T7 cells and converscly a hypophosphorylation of TCRC in CD45-deficient thyrnocytes, providing onc possible biochemical mcchanism for ~hc TCR signaling dcficits dcscribed. Collectively, these observations providc a conccplual and
esperimental framcwork for understanding the role of CD45 in TCR signaling at the DP slagc of' T ce11 dcvelopment. Acknowledgments
1 would likc to thank al1 those without whom this work would not have been possible.
First and foremost, 1 thank my supervisor Cindy, for her guidance and support. For sharing thc reagents rvhich allowed this work to be done: Drs. Pauline Johnson for the CD45 construct; Michael Julius for the a-Lck antisera; Phi1 Branton for the a-CD45 antisera; Andrc Veillcttc for the a-Fyn antisera; and Josef Penninger for the CD45-/-mice. To my cornmiitcc, Drs. Michacl Julius and Rob Rottapel for their scientific input. Jayne Danska, for her big piclure and conceptual input. The Danskonians (Danny, Case, Priscilla, Chns ti ne, Ildico) for their group supporUtherapy. To my sister, Serena, for always listening & understanding - 1 thank you. To my parents as always, your patience, lovc and confidence in me have allowcd mc to complctc this, and mainlain somc semblance of sanity. And last and most importantly, 1 mus1 lhank my Iab mütcs pas1 and prcscnl (a spial thanks to Tim, you wcrc thcre from thc bcginning to thc end and whüt a joumcy its becn!!; to thc rest, Dianne, Patti, Trang, BJ and thc oihers who movcd on) for making thc daily grind bearable, and dare 1 say, enjoyabie. Paqe Abstract II . . . Acknowledgements III Table of Contents iv-v List of Abbreviations List of Figures and Tables Chagter 1) INTRODUCTION A) T Cell Development i) Overview ii) Positive Selection and Negative Seiection iii) Markers of Positive Selection B) T Cell Receptor (TCR) Signal Transduction Mechanisms i) TCR Stucture ii) Proximal/Distal Evcnts in Mature T ce1 1s iii) TCR Signal Transduction in DP Thymocytes C) Regulation and Function of Tyrosine Phosphatase, CD45 i) Background ii) CD45-Deficient Cell Lines iii) CD45 Substrates iv) CD45 in T Cell Dcvelopment D) Thcsis Objective Chapter 2) MATERIALS AND METHODS A) Ce11 lines and Cell Cullurc B) Retroviral Gene Transfer C) Antibodies and Second Stage Reagents D) Flow cytomelry E) Stimulation of cells for assessment of tyrosine phosphorylation F) Stimulation of cells for assessment of surface phenotypic changcs G) Immunoprccipitations H) Cc11 Surfacc Biotinylation 1) SDS-PAGE and Immunoblotting J) RNA preparation and Northern analysis K) cDNA synthesis and RT-PCR analysis Chaptes 3) RESULTS 3 1-59 A) Charictcnzation of a CD45-dcficienl thymoma ccll linc, 3T7 B) TCR signaling defcct in 3T7 celis II l2illy CVG111,4 - IIIUULLIUII U1 IYlUblilC: ~IIUh~IlUI~liLLII~Il ii) Latc cvcnts - phcnotypic maturation, RAG- 1 downmodulation Differential ability of CO-rcccptorsto ovcrcome TCR signaling defect Presence of other phosphatases in 3T7 cells Defect in CD45 gene espression in 3T7 cclls Re-espression of exogenous CD45 in 3T7 cells Rescue of TCR signaling defects by CD45 i) Early events - induction of tyrosine phosphorylation ii) iate events - phenotypic maturation, RAG- 1 downmodulation Biochemical basis for rescue of TCR signaling by CD45 Analysis of thymocytes from ~~45-I-mice Chapter 4) DISCUSSION A) CD45-Dependent versus CD45-Independent TCR Signaling Pathways B) Molecular Targets of CD45 in TCR Signal Transduction C) Role of CD45 in Positive Selection of DP Thymocytes D) Future Studies Chapter 5) REFERENCES BSA: Bovine >ci-iiriialhiin~iri cDN A: Coinplcniciirary cicox yrihoiliickic acid CTL: Cytotoxic T lymphocyic DAG: Diacylglyccnd DEPC: Dictiiyl pyiocaihoilatc DMF: Dinlcthyl Iormmidc DN: Doublc ncptivc dNTP: Dcoxyribonuclcosidc 5'-triphosphate DP: Doublc posirivc DR: Dilhiothrcilol ECL: Enhanccci chcmilumincscci~sc EDTA: Ethylcncdiaminc tctrüacictiç ad FACS: Fluorcsccncc aciivatcd cc11 sostcr FCS: Fctial cdl' scrum FITC: Fluorescein isothiocyanak GM: Growth nicdium HBS: HEPES bufkrcd saline MEPES: N-2-hydroxycthylpipcr~~inc-NI-2-clhancsulli~nicacid HRP: Horsc radish peroxidasc IP: immunoprccipilii~ion IPj: Inositol tri-phosphate ITAM: Iinm~inorcccptortyrosine-hascd activation motif LB: Lysis buflCr mAb: Monoclunal aiilibudy MAE: MOPS, sodiiim iicciatc, EDTA 2-ME: 7-1ncrcaptoctl1~111ol MFI: nican 17~ioscsccnccinlcnsily MI IC 1: Ma,ior Iiistocc,nipa~ibililyccii~~plcx class 1 MHC II: Maior histoconipatibilily complcx class II mRNA: Mcsscngcr RNA MOPS: 3-(N-niorpho1ino)-propanc sulfonic acid PBS: Phosphate bull'ci'cd salitic PCR: Polyincrasc chain iciictiori PE: Phycocrytliiin PMA: Phorbol 12-niyi-istritc 13-acclatc L '0. ' ",LU" 'J"'.""" """"7"
PTf'iisc: Pi'0t~iiliyili~ili~ ~iho~~)hiilii~~ p-tyr: Phospho~yrosiiic K AG: Rcuvnhin:isc ircriv;itirig pc RNA: Kibonuclcic acid RT: Rcvcrsc lranscripli~w SDS-PAGE: Sodium dodccyl sullà~cpolyacrylamidc gcl clcctrophorcsis SHI : Src-homology 1 domain SH3: Src-honmlogy 7 domain SH3: Src-homology 3 domain SP: Singlc positive TCR: T-cc11 antigcn rcccptor Tyr: Tyrosinc V(D)J: Variahic, divcrsily, joining List of Figures
Fur1 : Surlàcc phcnotype of 317 cclls. Figurc 2: ElTcct of CD45 dcficicncy on TCR-rilcdiatcd sisna1 ~ransduciionin 3T7 cclls. (A) Induclion of tyrosine phosphorylation in 377 cclls aller TCRP or TCRP + co- rcccpior crosslinking. (B) CD5 induction in 3T7 cclls alter TCRp or TCRP + co-rcccpior crosslinking. (C) CDS induclion in 3T7 cclls ~rcatcdwith thc PTK inliibilor. Hcrhimycin A.
Figure 3: Molccular basis OC difl'crcntial signaling hctwccn CD4 and CD8 co-rcccptrirs in 3T7 cclls.
(A) Difkrcntial association of Lck wi~tiCD4 and CD8 co-rcccp~orsin 3T7 cclls.
(B) Ev;ilu;ilion of CD8 corcccpor isoforms cxpi-csscd in Yi7 cclls.
Figure 4: El'lcci oc thc tyrosine phosphatase inhibitor, pcrvanadatc on 3T7 cclls. (A) Induction of lyrosinc phosphoryla~ionin 3l7 cclls aficr pcrvanadaic &calmeni. (B) Induction of CD5 in 3T7 cclls akcr pcrvanadalc trcatrncnt.
Figurc 5: Analysis or CD45 gctic cxprcssion in 3T7 cclls.. (A) Northcrn analysis of CD45 mRNA expression in 3T7 cclls. (B) Surlàcc CD45 inducibly cxprcsscd in 3T7 cclls. (C) RT-PCR anülysis of CD45 isoforms inducibly cxprcsscd in 3T7 cclls.
Figurc 6: Re-expression of cndogcnous CD45 coirclatcs with rcstorütion of TCR rcsponsivcncss.
Figure 7: Expression of cxogcniius CD45 in 317 cclls by rctrovirnl-nicdialcd gcnc 1ransCcr. (A) Sclicrnatic rcprcscnlation of CD45 rclroviral construcl.
(B) Wcstci-n blot anülysis of CD45 prorcin lcvcls in 3T7 inl'cçtan~s. (C) Surface CD45 expression in (3418-rcsis~ani3T7 infcctants.
Figure 8: El'l'cct of exogciious CD45 cxprcssion on 'TCR-mcdiatcd signal trarisduc~ionin 3T7 cclls. (A) Exogcnously cxprcsscd CD45 rcstorcs TCR-mcdiatcd protcin Lyrosinc phosphorylaiion in 317 cdls.
(B) Exogci~uuslycxprcsscd CD45 rcstorcs TCR-induced changes in cc11 surfacc phcnolypc.
(C)Extigcnously cxprcsscd CD45 rcstores TCR-induccd RAG- 1 downmodulalion.
Figurc 9: Tyrosinc phosphorylalion indcx of potcntial CD45 suhstraics in 3T7 celIs. (A) Aniilysis of TCR phosphorylation status. (B) AnaIysis oSLck phosphosphorylntion status.
(C) Andysis of Fyn phosphorylation status.
Figure 10: Anülysis of'TCR< tyrosinc phosphorylation in CD45-/-thymticytcs.
List of Tables
Table 1. TCR-Mediatcd CD5 induction in Parental 3T7 cclls versus CD45 Infcctanb CHAPTER 1
INTRODUCTION Introduction
T cclls arc the priniriry oi-clicstrators ol' tlic imii-iunc systcni. rcspimsihlc l'or thc co- ordinat~rcgulation of boih c~ll~~liiiand Ii~inioriil inini~inily. To ;icconiplisli this. T cclls havc cvolvcd a complcx surlàcc rcccptor. the T cc11 aiiiigcii rcccptor (TCR) wliich rccogiiizcs antigcnic pcptidcs hound to self niqior Iiistoçompütihility complcx (Ml-IC) molcculcs. Maturc T cdls mcdialc tlicir ~SfcctorS~~nctiolis. which involvc cclliilar prolilkraiion. rclcasc ol' hici-activc niolcculcs. and recruiinicnt of othcr cc11 types. via signals Lhrough the TCR. A criiical niolcculc jnvolvcd in rcgiila~ingTCR signal transduction is the tyrosiiic phosphatase, CD45 Studics of maturc T cclls have shown thüt CD45 acts to dcphosphorylütc nçgativc rcgulntory tyrosine rcsiducs on S~-c-~iniilykinascs, thc activation of which initiates thc TCR signal tciiisduction cüscadc. H«wcvcr, ihc rolc of CD45 in TCR signüling in immaturc T cclls has not yct hccii fiilly dclïncd. T cc11 devclopmcnt takes place in thc thymus, whcrc a complcx aiid inconiplctcly undcrstod devclopmcntal pathway is Iollowcd. It is at ihe CD4+CD8+ double-positivc (DP) stqc of dcvclopmcnt whcrc a critical test of thc thyiiiocytc's TCR spccil'iciiy ocçurs. This dcvclopiiicnial chcckpoiiit znsurcs that oiily thc appropriatc T cclls niütuie via die pioccss of positivc sclcctioii, whilc riddinp thc organisrn of thymocytcs whosc spccilïcity niay hc Iiarmiùl via thc proccss 01' ncgaiivc sclcction.
Thc work prcscntcd hcrc aticmpts to addrcss thc rcilc of CD45 in rcgiilating TCR sigiiüls at thc critical DP stage of dcvclopmcnt. Two mode1 systcnis arc uscd: i) a CD45-dcîïcicnt thyniic lyniphoniii DP cc11 hic callcd 3T7, and ii) DP thymocytcs froiii CD4.5-1- micc. Tlic
çhanictcrizatioii of the I'uncticiiial and phcnoiypiç conscqucnccs of thc lack of CD45 in thcsc two sysiciiis providcs iniportmt iiisipht iiito the rcilc of this niolcculc during T cc11 dcvcloprnciit. This iiiisoduchm will hcgin with an ovcrvicw of T ccil dcvclopmciit lixusiiig oii positivc sclccticii~, fiillowcd hy a rcvicw oIsTCR signal traiisdoctioii nicchaiiisms, 1iiglili;litiiig dill'crcnccs hciwccii m;iiusc and imniitturc T cclls. Thc I'inal scction will tiisc~issCD45 linc!ion ;iiid rcg~ilation. l'oçusiiig titi its sole dusing T cdl dcvcli~pniciit. A) T Cell Development
i) Overview
The primary site of T cell development takes place in the thymus, wherc CD4-CDS- (double-negative, DN) precursors bccomc DP before finally maturing in10 eithcr CD4+ or CD%+ (single-positive, SP) cells (reviewed in Guidos, 1996). While expression of CD4 and CD8 arc widely used to definc thymocyte populations, it is important to remember that each population an bc further subdividcd on the bais of other markers. For example, thc earlicst thymocytcs exprcss low levels of CD4 and have not rearranged their TCR genes (Wu et al., 1991). Thesc cells are the immediatc precursors of DN thymocytes, which can further bc subdivided into CD44+CD25+, CD44-CD25+,and CD44-CD25subsets (Godfrey et al., 1994). Signals through a putative pre-TCR ccimplex, consisting of pre-TaiTCRplCD3 componenls, are critical at the CD4.4-CD25+ slagc, and rcgulatc the clonal expansion during the DN to DP transition (reviewcd in Lcvclt and Eichmann, 1995). In most mouse strains, a CD~-CD~-CD~]~~dcvclopmental intermcdiate exists, which gives rise to DP blast cells, expressing low lcvels of clonotypic
TCRup on the surface (Guidos et al., 1989). A com plex, and as yct incomplctely undcrstood, TCR-mediated sclection process ensues. Most devcloping thymocytes will die from ncglect because they exprcss TCR incapable of recognizing self-MHC. Thc remainder can follow onc of two dcvelopn~entalfatcs: i) A srnaIl subset of thymocytes bearing a TCR capable of inlcrilcting with sclf-MHC arc rescucd from active cell death in a process known as positive sclcction. This cnsurcs the crcation of a functicinal rcpertoirc of mature T cclls (rcvicwcd in Jamcson ct al.,
1995). ii) Altcrnali\dy, thymocytcs with self-reactive TCRs against self-MHC arc dcstroycd or lunctionally inactivatcd in a pi-occss known as negütive seleclion (rcvic\vcd in Nossal, 1994).
Togcthcr thcsc Iwo proccsscs, both mediated by signals delivcrcd through thc TCR, dictatc whcthcr a givcn thymocytc dics or livcs Io continuc ihc maturation proccss. Positivcly sclcclctl
DP thymacytcs will then furlhcr dif'f'crcntistc into CD4 or CD8 SP thyinucytcs. Thc proccss of + L. csprcssion of CD4, whcrcas thosc bcaring class 1 MHC-spccifrc TCR will rctain cxprcssion of
CD8, although thc csact mechanism of this proccss rcmains unresolvcd (rcvicwcd in von
Bochmer, 1996).
ii) Positive Selection and Negntive Selection
Positive selection was discovered by thc analysis of bone marrow radiation chimeric micc. Whcn bonc marrow from an (AxB)FI heterozygous mouse (A and B representing diffcrent MHC haplotypes) was used to rescue a lethally irradiated homozygous A parent. T cells responded preferentially to antigens presented by the host MHC (A) antigen preseniing cclls
(Bevan, 1977). This MHC rcstriction was shown b be mediated by H-2 antigens expresscd on thc thymus (Fink and Bevan, 1978; Zinkernagel et al., 1978). Negative selection was definitively demonstraled by the clonal deletion of Mls-reactive T cells in the thymus (Kappler cl al., 1987). With the advent of TCR transgenic mice, analysis of selectivc evcnts bccame technically easier (Teh et al., 1988). Analysis of transgenic mice expressing a TCR spccific for the male transplanlation antigen HY revealed thit when lhe HY antigen was expressed (ic. in mdc mice) in conjunction with the appropriate MHC restriction element, massive clonal dcletion of TCR transgcnic thymocytes occurred, dramatically confirming thc existence of ncgalive sclcction (Kisielow et al., 1988). In constrast, when the HY autoantigen was not expressed (ie. in fcmalc mice), positivc selection of TCR transgcnic Lhymocyles occurred, bascd on MHC haplotypc
(Kisiclow ct al., 1988; Huesmann et al., 1991). Gcnctic studics havc dcfincd al least two "devclopiiiental chcckpoinls". Thc fïrst chcckpoint occurs at the DN to DP transition, and is mediaied by putativc signals through thc pi-c-
TCR complex to ensurc that a funclional TCRp chain has becn made (rcvicwcd in Lcvcll and
Eichmann, 1995). Thc sccond chcckpoint occurs at the DP 10 SP transition, during which ~hc
thymocytc undcrgocs a critical tcsl of its TCR spccificily, rcsulting in positivc and ncgativc
sclcction. In contrat to the first dcvelopmental checkpoint, positivc and ncgative sclcclion arc 2 Li u thymocytes with the appropriate seceptors on their surface develop, both in terms of Sunction and spcci ficity.
iii) Mnrkers of Positive Selectiott
The early studies described above laid the groundwork for a more detailed examination of thc developmental events that accompany selection events. A major consequence of positive selection is the alteration in the life span of TCRap+ thymocytes. Kinetic studies havc deduced the average Me span of a DP thymocyte 10 be 3.5 days, after which the thymocytes cithcr dic by apoptosis, or are positively selected for further maturation into cells that are long-lived (Egerton et al., 1990; Huesmann et al., 1991; Kisiclow and Maizek, 1995). The upregulati«n of thc cell survivül gene, bcl-2, during positive selection provides an attractive mechanism for this phenomenon (Veis et al., 1993; Linette et al., 1994). Another major consequence of positive selection is the termination of the TCR rearrangement process. TCRap DP thymocytcs still express the recombinase-activating genes RAGl and RAGS, but expression of thcse gencs is rapidly shut off in response to TCRap signais (Turka et al., 1991b; Brandle et al., 1992). Thus, rearrangement can occur at the second TCRa allele if rearrangement of the first allele was not successful (Petrie et al., 1993), increasing the likelihood that the developing DP lhymocyte will be positively sclccted. Positive selection also triggers an ordered series of phenotypic changes. Adoptive transfcr of defined subsets of thymocytes into the thymus providcd a uscful tcchniquc Ibi- studying the developmental potential of particular thyrnocyte populations. Using this approach, i t was Iound that DP TCRI0 blast cclls have the developmental potential for furthcr diffcrcntiation, bccoming TCRI~CD~+CD~-or TCR~IWCDS+ T cells (Guidos et al., 1989). Transitional intcrmcdiates on thc CD4 or CD8 differentialion pathways wcrc suggcstcd to havc a
TCR"~~CD~+CD~~~or TcR1nd~D4l0c~8+ phenotypc (Guidos cl al., 1990). Howcvcr thc csact lincagc rclationships among these transitional intcrmediales rcmains controvcrsial. Somc stuarcs haVC SUggCSted tnat tne transirionai ceiis on rnc LUY iincagc may nor ncccssariij7oc comrnitted to the CD4 lincagc, and may still develop into CD8 SP thymocytcs (Lucas et al., 1995;
Lundbcrg et al., 1995; Suzuki ct al., 1995). The quantal uprepulation of TCR espression on developing DP thymocytes likcly rclatcs to "sub-stagcs" of diffcrentiation (Guidos et al., 1990; Ohashi ct al., 1990; Shortman ct al., 1991), suggesting that positive selection maybe a multi-step process. Importantly, whcn puriricd
DP thymocytcs are stimulatcd through their TCR iri vitro, only a subset of maturationai proccsscs associated with positive selection occurs (Kearse et al., 1995; Groves ct al., 1997). Whilc increased expression of CD5, CD69, and Bcl-2 and termination of RAG- I and prc-Ta expression were observed, clona1 deletion and CD4JCD8 Iineage cornmitment wcre not seen (Kcarse ct al.,
1995; Grovcs et al., 1997). These observations are consistent with the notion that multiple, or sustained TCR engagements may be necessary for the complete developmental progression associated with positive selection (Kisielow and Maizek, 1995; Wilkinson et al., 1995). AI tcmativel y or addi tionall y, non-TCR derived signals present in the thy mic rnicrocnvironmcnt (including growth factors, adhesion molecules, CO-stimulatorymolecules, etc.) maybc rcquircd for further developmental progression. Late events associated wi th maturation i ncludc the downrcgulalion of heat stable antigen (HSA) and Thy- 1 (Fowi kcs ct al ., 1988; Grovcs cl al .,
7 997; Lucas et al., 1994). Whilc thcsc markcrs provide a useful experimental tool for following positivc sclection, somc caution must bc used in their interpretation. For cxample, CD69 is an activation markcr uprcgulated following TCR engagement, and ihus could also be a markcr of negativc sclcciion (Kishimoto ct al., 1995). The biologic functions of mosi of thesc moleculcs rcmain ill-dclïncd. Onc of thcsc markcrs, CD5, appcars to bc a ncgativc rcgulator OS TCR signaling (Tarakhovsky ct al., 1995). Thc identification of other molecules capable of impinging on TCR signaling paihways at this stagc of dcvcloprncnt will have important çonscqucnçcs on thcsc sclcciivc processes. Onc such molcculc, CD45, is thc focus of this thcsis. Belorc rcvicwing thc fiinction and rcgulation of CD45, a rcvicw oSTCR signal transduction mcchanisms is prcscntcd. B) TCR Signal Transduction Mechanisms
i) TCR Structure
Thc vanable domains of a and p chains of the TCR mediate recognition of antigcniç pcptides bound to MHC molecules on the surface of antigen prcsenting cclls (APCs). Thc cxireme diversity of TCRap specificities is mediated by somatic gene rearrangcment of n~ultiple variable (V), diversity (D), and joining (J) gene segments in a process known as V(D)J recombination (reviewed in Lewis, 1994). The TCRa and TCRP subunits are non-covalently associatcd with the invariant CD3 y$, and E chains, and either TCRS-< homodimers or <-q hcterodimers. The cxact stoichiometry of the TCRICD3 complex is not known, but is thought to consist of a disulfide-linked TCRa$ heterodimer, in association with a CD3ey and a CD3t6 pair, and cither a TCRCS homo- or TCRh heterodimer. The TCRS and CD3 components perhrm two critical functions: i) ensuring the proper assembly and surface expression of the TCRICD3 corn plex, and ii) coupling antigen recognition to the intracellular signali ng machinery (revicwcd in
Malissen and Malissen, 2996). In addition to the TCWCM complex, the CD4and CD8 CO-receptorsplay important rdcs in T ce11 recognition and signaling. CD4 is a monomeric integral mcmbranc glycoprotein, whi lc
CD8 cxists as ü disulfide-linked a-a homodimcr or a+ hetcrodirncr (reviewcd in Julius ct al., 1993). Thc CO-rcceptorsact as adhesion moleculcs by recognizing non-polymorphic rcgions on MHC class II (CD4) (Doylc and Strominger, 1987), and MHC class 1 molccules (CD8)
(Normcnt ct al., 1988),thus stabilizing interactions betwccn thcT ceIl and the APC. In addition, the co-rcccptors can transducc bi ochcmical signals during T-cc11 activation, by vi rtuc of thci r association wi th thc tyrosine kinasc Lck (Veillette et al., 1988).
Thc TCR a and P chahs possess only short cytoplasmic domains of Iïvc arnino acids and arc thcmsclvcs unli kcl y to bc capablc of cou pling to inlraxllular signaling pathways. I n con trast, tlic CD3 chains and TCRt chains contain larger cytoplasmic domains, which couple anti gcnlMHC rccogni lion to signal transduclion paihways (rcvicwed in Chan et al ., 1994a). This . , - . - - - .. . - - L... d fi subunits, including thc TCRICD3, BCRIIgalIgp, and FCsRIy, which rcvcaled a highly conscrvcd motif in thc various receptor systems (Relh, 1989). This ITAM motif (irnmunoreçeptor tyrosine-based activation motif: DIEX7DIEX2YX2LX7Y X2L) was found to bc both necessary and sufficient for receptor signaling (Romeo et al., 1992; Irving et al., 1993). The ability of thcse ITAM motifs to be inducibly phosphorylated on thcir two tyrosine rcsiducs crmtes a binding site for the SH2 domain of various intracellular signaling molccules. Binding to phosphotyrosine sites can affect SHZcontaining proteins in multiplc ways, including dircct stimulation of enzymatic activity, cellular relocalization, and enhanccd tyrosine phosphorylation
(reviewed in Pawson, 1995).
ii) ProxirtiallDistnl Events in Mature TCR Signai Tramdiution
Effector functions induced during T ce11 activation include lymphokine secrebon, cellular proliferation, and cellular differentiation. The signaling events leading to these changes in genc expression involve a complex set of biochemical events, which have been relatively wcll- charactcrizcd (reviewed in Cantrell, 1996; Wange and Samelson, 1996). The carliesl dctcclüblc evcnt in TCR signaling is thc induction of tyrosine kinase activity, leading to protcin lyrosinc
phosphorylalion (Samelson cl al., 1986; Hsi et al., 1989; June et al., 1990b). Two farnilics of
PTKs have becn implicated in TCR signaling. Lck and Fyn arc lymphocytc-spccific membcrs of
thc Src-kinase fainily (Veillcttc and Davidson. 1992). The çonservcd struct.ura1 katurcs 01' this
fmily include: 1) an amino-terminal glycine, which is required for myrisloytation and mci-iibr~~nc association; 2) a unique domain of roughly 60 amino acids, which allows spccific inlcractions wilh cellular regulators; 3) the SH3 domain, a region involved in recognition of prolinc rich
sequcnccs (Rcn ct al., 1993); 4) the SH2 domain, which mcdiatcs inlcractions with
phosphotyrosine-containing proteins (Pawson and Gish, 1992); 5) the catalytic SHI domain,
coniüining an ATP-binding and an autophosphorylation site; and 6) ihc carbosy-terminal ncgativc
rcgulütory domain. Zap-70 is a member of another class of cyk)plasmic PTKs, thc SyklZap-70 ~--- ~ - - .~-" --.,, ---- > -~ . -, - -~---- , i they lack an SH3 domain and instcad havc two SH2 domains; 2) thcy lack niyristoylation si tcs; and 3) they lack thc carboxy-tcrminal negativc rcgulatov dornain. These three PTKs have been implicated in TCR signaling by both biochcmical and gcnctic mcans. Experiments with T cell lines lacking Lck (Straus and Weiss, 1992) or csprcssing mutant forms of Lck (Abraham et al., 1991; Luo and Sefton, 1992) indicatc that Lck can participatc in
TCR signaling. Furthermore, mutant mice overespressing a dominant negativc Ick transgenc (Levin et al., 1993) or with a targeted disruption of the lck gene (Molina et al., 1992) possessed profound defccts in thymocyte development and diminshed TCR function. A rolc for Fyn in TCR signaling is supported by studies in which SP thymocytes from mice overexpressing afyrr transgene were hyperstimulable to TCR ligation (Cooke et al., 1991). Conversel y, SP thymocytcs from mutant mice lacking Fyn displayed grcatly diminishcd TCR rcsponscs (Applcby et al., 1992; Stcin ct al., 1992). Finally, Zap-70 has been shown to be activated by TCR ligation and Sound to associate with the tyrosine-phosphorylated ITAMs in TCRL (Chan et al., 1992). Futhermore, T cells from mice (Negishi et al., 1995) and human patients (Arpaia et al., 1994; Chan et al., 1994c; Elder et al., 1994) defective in Zap-70 expression are severel y impaired in their response to TCR ligation. A role for Syk in TCR signaling is less clear. Engagement of TCR results in Syk activation in thyrnocytes (Chan et al., 1994). Howcver disruption of thc Syk gene does no1 affect maturation of af3 T cells (Cheng et al., 1995; Turner el al., 1995), suggcsting that Syk is dispensable during T cell development. Currcnt inodcls of TCR signal transduciion posit a sequcntial activalion of PTKs following TCR engagement, in which Lck andior Fyn are initally activated to phosphorylalc
ITAMs in CD3 componcnts (Iwashima ct ai., 1994; van Ocrs ci al., 1996). This allows rccrui tmcnt of Zap-70 by binding phosphorylated ITAMs (Chan ct al., 1993), and ils rapid aciivation by tyrosine phosphoryIation (Chan et al., 1995; 1washima ct al., 1994). Molcculcs subsequently phosphorylated by the TCR-proximal PTK signaling cascadc includc phospholipw
Cy 1 (PLCy 1) (Mustclin ct al., 1990), p95 Vav (Gulbins et al., 1993), MAP kinasc (Ettchadich cl al., 1992), and PI-3 kinasc (Ward et al., 1996). Following lhcsc immcdiatc carly signaling -. -..'") . -- , . --.-.,.A-u ...- --.."W. -. -- -.-.----.., ,------O ...r-.- (IP3),and diacylglycerol (DAG) (Weiss and Liltman, 1994). These two second mcssengers arc i-csponsible for thc TCR-induced rise in cytoplasmic free calcium and activation of' protein kinasc C (PKC), respective1y. A critical regulatory mechanism for Lck and Fyn involvcs the C-terminal negativc rcgulatory tyrosine. It has been demonstrated that Csk can phosphorylate this site (Bcrgrnan et al., 1992), making Csk a negative regulator of the Src-family kinascs. When phosphorylatcd, the C-terminal tyrosine binds to the SH2 domain of the same kinase molecule, thereby forcing it intoan inactivc conformation (Cooper and Howell, 1993). In contrast to Csk, CD45 is thought to positively regulate the Src-family kinases by dephosphorylating the C-terminal tyrosine (revicwed in Trowbridge and Thomas, 1994). The recent resolution of the crystal structures ol'c-
Src (Xu et al., 19971, Lck (Yamaguchi and Hendrickson, 1996), and anothcr membcr of thc Src- family, Hck (Sicheri et al., 1997), have provided further insight into the structurallfunclionaI rclationships of Srk-family PTKs. These studies have revealed thal not only is lhc phosphorylated C-terminal tyrosine bound by the SH2 domain, but thc SH3 domain is also involved in an intramolecular interaction with the SHNinase linker domain (Sicheri et al., 1997; Xu cl al., 1997). Thus, the crystal structures raises the possibility that compelilivc interactions with SH3 or SH2 ligands could also activate the molecules by displacing thc inhibitory intrarnolecular interaction, without necessarily involving C-terminal dephosphorylation.
Following these TCR-proximal events, second messengers are thought Lo aclivatc at lcast twoTCR-dislal signüling palhways. Onc involves thc activation of calcincurin by clcvatcd Ic\~ls of intracellular calcium. Calcineurin, a calcium/calmodulin-depcndent serinelthrcminc phosphahsc, stiinulritcs translocation of thc NF-AT(c) transcription factor from thc cytoplasm to ltic nuclcus (Flanagan et al., 1991). A sccond TCR-distal pathway involvcs tlic Ras andtor PKC- mediaicd activation of' the ERKIMAPK cytoplasmic serinelthreonine kinasc family. Thcsc kinascs translocatc to the nuclcus and rcgulatc changcs in gcne cxprcssion by phosphorylating a varicty of' transcription factors (Hill and Trcisman, 1995). In this way, changcs in gcnc csprcssion arc incurred such that T ce11 activation occurs. iii) TCR Signal Trartsdrcctiort in Zrltmatrlre Thy~rmcytes
In contrast to mature T cells, the elucidation of signal transduction pathways used in immarurc thymocytes has only just begun. Many of the same signaling molecules which havc been described in mature T cell signal transduction have been shown to participate in TCR signajing in DP thymocytes. Most of the receptor molecules, as well as the intracellular PTKs involved in thc proximal TCR signal transduction cascade, have been genetically disrupted, with varying consequences on T cell development. For esample, micc dcficient in the Src-family kinases Lck or Fyn posscss dramatically different phenotypes. Lck-deficient mice have a 10-lold reduction in the number of thymocytas, due to a partial block at the DN to DP expansion stage, and mature SP cells are greatly reduced (Molinaet al., 1992). In striking contrast, Fyn-deficien1 mice have no gross abnormalities in either the number or phenotype of thymocytes (Appleby ct al., 1992; Stein et al., 1992). Yet another phenotype is observed in Zap-70 knockout micc, where there is a block in the development of both CD4+ and CD8+ SP T cells (Negishi et al., 1995). Finally, mice lacking expression of Itk, a cytoplasmic PfK bclonging to the BtklTec farnily, have a modcrate reduction in the numbcr of mature T cells (Liao ct al., 1995). Howcvcr, this cffect is morc promincnl in Itk-deficient mice expressing transgcnic TCRs, wherc positivc sclection appcars to bc irnpaired (Liao et al., 1995). These gcnctic ablation studies suggcst thü~ the various PTKs are diffcrentially rcquircd during developmcnt, raising the possibility thüt L~C inccha~iismof TCR signal lransduclion müy change over the course of T ccll dcvclopmcnl. Several lincs of cvidence suggest that TCR signaling evenls in DP thymocytes inay dill'er in somc respccts to lhosc characterized in mature T cells. For esamplc, aggrcgalion of TCR or
CD3 on frcsh ex vivo DP thymocytcs produces only marginal incrcascs in Lyrosinc phosphorylation and intraccllular calcium levels (Finkel et al., 1987; Nakayama ct al., IWO;
GilliIand el al., 1991; Turka ct al., 1991a; Sancho el al., 1992). Onc mcchanism for this bluntcd rcsponsc has bcen proposed by Singcr in a series of studics dcmonstrüting that TCR signaling is ncgütivcly rcgulatcd by CD4-associüted Lck (McCarthy et al., 1988; Nakayrirnü ct al., 1089; .-a"-, - -, " - . - - -. . .. - - .-. - - ---.> --- - -.------., -- - - 7 -- -..> -.", .. .------.., .rr .. ,. Iniercstingly, thc othcr co-rcccptor, CD8 docs not display this activity. While this may bc attributable to thc intrinsic prefcrential association of Lck with CD4 over CD8 (Wiest ct al..
I993), there is also a specific developmentally controlled expression of CD8a. Through a mechanism of alternative splicing, two polypeptide chains, a and al, that differ from onc anothcr in the lcngths of their cytoplasmic tails, are expressed in T cells (Zamoyska and Parnes, 1988).
1 rnmaturc T cells cxprcss both CD8a and CD&rqforms on their ce11 surface, whilc maturc T cclls cxpress on their cc11 surface predominantly the heterodimer containing CD8a (Zamoyska and Parnes, 1988). This may have important Iuunctional implications, as the CD&' polypeptide is unable to associate with Lck (Zamoyska et al., 1989), suggesting that DP thymocytcs arc intrinsically morc capable of delivering CD4-mediated signals than CD8-mediaied signals. Another difference inTCR signaling between DP thymocytes and mature T cells involves their requirements for co-stimulation. In the classic mode1 of T ceIl activation, two signals arc required: the first via the TCRICD3 complex and the second provided by CD281B7 ligand- receptor system (Robey and Allison, 1995). In striking contrast, DP thymocytes do not appear to rcquirc co-stimulation. Mice deficient in CD38 conlain normal numbers of T cclls, and apparcntly normal developmental profiles (Shahinian et al., 1993). Furthcrmore, whcn CD78- deficient mice were made to cxpress a$ transgenic TCR, no obvious deficiencics in cilher positivc or ncgütivc sclection were observed (Walunas et al., 1996), suggesting that alternative co- stimuiülory pathways exist, os that the two-signal hypothesis does not apply to DP thymocytcs. Thus several differences cxist in the TCR signal response of DP thymocytes as comparcd to maturc T cells. To complicate malters further, several groups have suggcstcd ihat biochcmically distinct signal-transduction palhways may distinguish positive and ncgüiivc selection. Various groups have looked for qualitative differenccs in TCR signaling Icading to cilher positivc or ncgativc selection. Two groups found positivc sclcction to bc prcfcrcnhlly inhibiled by the calcincurin inhibitors, cyclosporin A or FK506, and not ncgcilivc sclcclion
(Anderson cl al., 1995; Wang ct al., 1995), suggcsting that ncgativc selcclicin acts via a calcineunn-indcpcndcnlpathway. Similarily, ovcrcxprcssion OS doininant-ncgativc Ras or MEK- - selection intact (AI berola-IIa cl al., 1995; Swan et al., 1995). Howcvcr, thc lailure to addrcss whethcr there was a complctc or partial block in the signaling pathways, leavcs opcn Lhc interprclation that quantitalive, ralhcr than qualitative, differences in signaling may distinguish positivc from negative selection. As studies îurther define, by both genctic and biochcmical means, the signaling molecules invohed in the TCR signal transduction pathway in DP thymocytes, a greater understanding of how a thymocyte undergoes selective processes will bc achieved. One critical molecule, CD45 has been shown to be important in regulating TCR signals in DP thymocytes as well as in mature T cells. The final section will review the regulaiion and function of this molecuie.
C) Regulation and Function of the Tyrosine Phosphatase, CD45
CD45 (also known as leukocyte common antigcn, Ly5, T200 in T cells and B330 In B cclls) is a transmembrane glycoprolein, highly expressed on al1 nucleated cells of hematc~poieiic origin (reviewed in Trowbridge and Thomas, 1994; Okumura and Thomas, 1995; Frcarson and
Alcxander, 1996). The function of CD45 rcmained unidcntificd until in 1989, ivhcn significiinl homology bctwecn the cytoplasmic domain of CD45 and a tyrosinc phosphatasc, mP-1B was identificd, suggcsting that CD45 is a receptor phosphtase (PTPasc; Charbonneau ct al., 1989). CD45 has an cxtcnsivcl y gl ycosylated amino-terminal extcrnal domain, a single mcmbranc- spanning rcgion, and a largc cytoplasmic domain containing two 300 amino acid tandcinly rcpcatcd PTPase domains, with the proximal domain containing lhc calalytic aclivity (Johnstin cl al., 1993).
CD45 is csprcsscd as multiple isoSorn~sranging in molcculür mass from 180-335 kDa, which scsult from allcrnativc splicrng of four variable csons (csons 4, 5, Ci and 7) in tlic cstraccllular N-terminal domai n of thc moleculc. CD45 isoSorms conlaining cson 4-, cson 5,or , , , u u , '.L, 7 " , TA,CL 7.. U,\,21CLI.I1\III .II diffcrcnl CD45 isoforms is ce11 type-specific and depends on the diffèrenliation and aclivalion statc of the lymphocyte (Lefrancois and Goodman, 1987; Hathcock et al., 1992). The majority of developing thymocytes and activated T cells esprcss the CD45RO 180 kDa isoform (lacking csons 4, 5 and G), whcreas CD4+ and CD8+ SP thymocytes and naive pcriphcral T cclls can cspress various 1 eson forms. In contrast, B lymphocytes espress the CD45ABC 330 kDa isoform (Cot'fman and Weissman, 1981). The finding thal diffcrent isoforms dif1'cr in their ability to participate in antigcn-rnediatcd slimulation in a mode1 cc11 line systcm (Novük ct al., 1994) has Icd to the hypothesis that changes in isoform expression can direçtly alter the signaling characteristics of the T cell, although the physiologie significance of this remains to bc determined.
ii) CD45-Deficieat Cell Lirzes
Dircct cvidence for the involvement of CD45 in regulating TCR signal transduction \vas first providcd by studics of CD45-deficient T cclls. By mutagcnizing an antigcn-spccifïc CD4+ murinc T ccll linc and selccting for variants which lost surfacc CD45 cxprcssion, Pingel & Thomas corrclalcd loss of CD45 with an inability to prolifcrate in responsc to antigen or lo CD3 l igation. A s poniancous CD45 rcvcrtant rcgaincd KR-mcdiated ac~ivationresponscs (Pi ngcl and
Thomas, 1989). Thc rcquirement for CD45 was soon estcnded by analysis of olhcr CD45 dcficicntccll linc systcms: a CD8f cytolyticT-ccll clonc (Weavcr ct al., 1991), a human DP T ccll lcukcmic linc (Koretzky ct ai., 1990), a DN T lymphoma Iinc (Volarcvic ct al., 1993), hunirin
CM+furkat Icukcinic T cclls (Korctzky et al., 199 l), a B ce11 plasniacytoma cc11 linc (Juslcnicnl cl al., 1991) and a nalural killcr ceIl linc (Bell et al., 1993). In al1 of thcsc syslcins, antigcn rcccplors wcre uncouplcd from signaling pathways and thcir downstrcam outcomcs. Thus, thc
CD45 molcculc plays an obligalory rolc in rcgulaling anligcn rcccplor signaling in a varicly 01' 1ymphocytc lincagcs. " .--..---.- -.. -m.- -..-. O"-'-."C --.--..-- -7 .- --..---.-- . ---. ....-- ..-" ..---.. -,.--...,.. -. . charactcrixcd. CD45-deficicnt T cells arc unablc to prolifcratc or to produce cytokines, such as
IL-?. in rcsponsc to antigcn or to TCR ligation (Pineel and Thomas. 1989). Thc signaling c~~cadcinitiated by TCK ligation is interrupted at the earliest stagc in CD45-dcficicnt cells. Most
CD45-dcficicnt cclls show an inability to induce tyrosine phosphorylation of sevcral protcins (Koretzky et al., 1991), including TCRC chah (Volarevicel al., 1992) and PLC-y 1 (Koretzky et al., 1992) in response to TCR ligation. In addition, they fail to increase intracellular calcium (Koretzky et al., 1990; Volarevic et al., 1992), generate inosilol phosphates (Koretzky et al.,
1990; Volarevic et al., 1992), and activate PKC (Shiroo et al., 1992) in response to TCR ligation. However, not al1 studies have found the same manifestations on TCR signaling in thc absencc of CD45 One study found markedly elevated levels of basal tyrosine phosphorylation in a CD45-deficient leukemic ccll line compared to the CD45+ parental cell line (Volarcvic et al.,
1992), suggesting that CD45 might be responsible for maintaining low basal Ievels of Lyrosinc phosphorylation in parental cclls. Other groups Sound that the block in TCR signaling was nnot absolutc. For cxample, the inability Io signal through the TCR could bc overcornc whcn TCR and CD4 or CD8 were CO-aggregated(Deans et al., 1992; Shiroo ct al., 1992). Yct anothcr siudy described a CD45deficient Jurkat T cell line which was fully capable of responding to TCR stimulation despite the absence of CD45 (Peyron et al., 19911, an cffcct that was later shown to be inediated by the Syk PïK (Chu et al., 1996). Taken togethcr, thcsc studics rcveal a critical rolc for CD45 in regulating TCR signais. However, its role may diffcr dcpcnding on thc particular T cell studied, and the types of stimulatory conditions uscd. Various groups furthcr dissected the regulation and funclion of CD45 by rc-csprcssing various mutant forms of CD45 in the CD45-dcficicnt cc11 line systems by gcnc trünsfcr tcchniqucs. Ashwcll's group demonstratcd that the TCR signaling dcl'cct could bc complcmcntcd using thc enzymalically activc inlncellular portion of CD45 (Volarcvic ct al., 1993), so long as il was teihercd to the plasma membrane and catalytically active (Niklinska et al., 1994). Wciss's group made chimcric molecules in which the eslracellular and transmcmbrane domains of CD45 wcrc rcplaccd with those of thc EGF rcccptor (EGFR) (Dcsai ct al., 1993). Once morc, thc c~Lraçc~~u~üruomarn oi ~u43W~S UlSpenSilDlC ICII I~CSCUL: CIL I LK slgnallng. An InLcrcsung finding was that thc addition of EGFR ligands during TCR signaling resulted in a rapid and dramntic inactivation of TCR-mediated signals (Desai et al., 1993: Desai cl al ., 1994)' siiggcsting lhat ligand binding could regulate CD45 activity, although the exacl rnechanism 1-cmains undctïncd. Finally, a heterologous phosphatase from yeast espressed as a chimcric protcin wilh estracellular and transmembrane domains of a MHC clas 1 molecule could restore TCR-medialcd signal transduction, albcit lcss cfficicntly than thc analagous CD45 chimcric protein (Desai ct al.,
1993; Desai et al., 1994). Together, these studies helped define the striictural rcquiremenis ol' CD45 activity in TCR signaling: the intracellular active catalytic domain is absolutely nccessary, whilc the estracellular domain is dispensable. The above results do not rule out a role for the estracellular domain, however, as physiologic TCR signaling is likely to be more subtle than the highly artificial stimulatory conditions used in in vitro signaling assays. Indeed, a comparison of the ability of various CD45 isoforms to promote IL-2 secretion in a thymoma ceIl line stimulated with antigenlMHC rcvealed thal subclones expressing CD45RO were most effeclive while those expressing CD45RABC wcrc Icrist cffectivc (Novak et al., 1994). Notably, this diffcrence was not obscrvcd whcn TCR- spccific mAb was used as the stimulus. Further studies revealed that thc CD45RO isof'orm prefercntially caps with TCRICD4 when compared with CD45RABC, and that ihis interaction does not rcquirc the CD45 cytoplasmic tail (Dianzani et al., 1992; Leitenbcrg et al., 1996). Rcccnlly, CD45-I- micc have bcen generated that express CD45RO or CD45RABC transgcncs
(Kozicradzki cl al., 1997). While both isoforms rcstorcd thc developmcnt of CD4+ and CD8+
SP thyrnocylcs, only CD45RO micc wcre able 10 gencrate cytotoxic T cc11 rcspnscs açüinst viral infeciion (Kozieradzki ct al., 1997). Thus, the CD45 ectodomain can, in soinc circurnstianccs, play a role in modulating TCR signaling.
iii) CD45 Srrbstrntes * &A- "A..,-..- ..m. ---a .A.-- .....m.. --.... -- -- .- . - O - - - - .. . - . .. - - - - . --c---- transduction have been extensively studied. CD45 positivcly rcgulates mcrnbcrs of the Src family of kinases by dcphosphorylation of their C-terminal negative regulatory tyrosine. This was infcrrcd frorn studies of CM5-deficicnt ce11 lines, which were found to posscss dccrcascd IcvcIs of both Lck and Fyn kinase activity due to a hyperphosphorylation of their C-terminal tyrosines (Mustclin ct al., 1989; Ostergaard et al., 1989; Mustelin ct al., 1992; Gervais and Vcilletlc, 1995). Therefore, it is likely that Lck and Fyn are physiological substrates of CD45 Othcr studies indicate a more cornplex role for CD45 in its relationship with Src family kinases. In three different CD45- T-cell lines (YAC-1, SARKTLS, and HPB-ALL), Lck was found 10 be hyperphosphorylated at the C-terminai tyrosine site, consistent with previous studics (Burns ct al., 1994). Surprisingl y, however, the kinases were found to be hyperactive. Further studies demonstrcited that in addition to Tyr-505, CD45 can dephosphorylate Tyr-394 iri virro (D'Oro ct al., 1996). In the absence of CD45,the hyperphosphorylütion of Tyr-394 can cause an increase in kinase activity, despite the inhibitory hyperphosphorylation of Tyr-505 (D'Oro ct al., 1996). Thus, while the exact nature of the inability to signal through the TCR in CD45-deficicnt cclls rcrnains controversial, it seems to strongly correlate with the dysregulation of Lck andlor Fyn. In most of the studies alluded to above, CD45 appeared to differentially regulaie thc two Src-family kinases, Lck and Fyn. In one study, the degree of C-terminal phosphorylation of Lck and FynT in thrcc diffcrenl CD45-deficient ceIl lines (SAKRTLS 12.1, BW5147, NZB. 1) was assessed (Hurley et al., 1993). In each cell line, the C-terminal hyperphosphorylation of Lck was more pronounccd than for Fyn (Hurley ct al., 1993). Similarly, in anothcr CD45-deficicnt ccll line (L3M-93), an 8-fold increase in C-terminal phosphorylation for Lck and a 2-fold incrcasc for Fyn was rcported, dong with decreased kinase activi ty for cach (McFarland cl al ., 1993). 1 n contrat, anolher group studying a diffcrcnl CD45-dcficicntcc11 linc (HPB-ALL), found thrit Fyn had decreased kinase activity but not Lck (Shiroo et al., 1993). Howcvcr, thc statc of C-tcrminal phosphorylation was no1 examincd in that study. Instcad, data was prcscnted suggcsting thxt CD45 could regulate spcçii'ic pools of the Src-family kinases, spccifically thosc pools which wcrc rcceptor-associatcd, ic. TCR-associatcd hr Fyn and CD4 associatcd I'or Lck (Bil'I.cn cl al., cspcrimcntal proccdurcs remains to be rcsolved.
A numbcr of other potential CD45 substntcs have reccntly been dcfincd. CD45 has bccn shown to associatc with both TCRS (Furukawa et al., 1994) and Zap-70 (Mustelin et al., 1995) in i mrnunoprccipitation experiments, and is able to dephosphorylatc them i~ivitro. Final 1y, onc more potential rcgulator of CD45 activi ty may involve a novel protein called CD45 AP. This protein of approximately 30 kDa has been shown to immunoprecipitatc with CD45 and Lck, suggesting that it may be an adaptor protein linking the two enzymes (Schravcn et al., 1994). Clearly much remains to be learned about the nature of CD45 substrates and how CD45 acts to regulate them.
iv) CD45 irc T Ce11 Developmerrt
Definitive evidence that CD45 is involved in thymocyte development camc from thc gcneration of CD45 exon 6-deficient mice (Kishihara et al., 1993). Disruption of cvon 6 might havc been cxpectcd to result in mice deficient on1 y in expression the CD45RC isoform, howevcr cxprcssion of al1 isoforms of CD45 was compromised. Pan-CD45 staining of thymocytcs [rom ~~45-1-mice showed that the DN and DP populations lacked dctectable cxprcssion of CD45, howcvcr, a fraction of SP thymocytes and peripheral lymph nodc T cclls (10-30%) did cxprcss
CD45 (Kishihara et al., 1993). The total nurnber of thyrnocytes in ~~45-1-micc wüs only slightly reduccd as compared with wildtype mice (Kishihara cl al., 1993). CD45-'- micc had slightly incrcüscd frequencics of DN, normal frequencics of DP, and signifiçantly rcduccd
I'rcqucncics of SP thymocytcs as cornparcd to littcrmatc control micc (Kishihan ct al., 1993).
Pcriphcral T cclls wcrc markcdly reduced in iiurnbcr and wcrc rcfractory to CDS-induccd activation (Kishihara ct al., 1993). Morc recently, CD45-nul1 micc wcrc gcncratcd by gcnctiç disrupiion of cson 9, rcsulting in complcte abrogation of CD45 cxprcssion (Byth ct al., 1996). In thcsc micc, thc CD41CD8 dcvelopmental profilc obscrvcd was vcry similar to that sccn in the cson 6-knockout micc, but thc frcquency of T cells in thc peri pherd coinpartincnt wüs morc frcquency of CD#-CD35+ DN cells (Byth et al., 1996). Thus, thc phcnotype of the two indcpcndcntly-dcrivcd CD45 knockout mice suggcsts a minor rolc for CD45 in the DN 10 DP transition rcgulritcd by putativc signals through thc pre-TCR cornplex, and a more important rolc in the DP to SP transition regulated by signals through thc maturc TCRnfliCD3 cornplcs.
Howcvcr, the nccd for CD45 at cither of these checkpoints is not absolutc, as somc DP and SP thymocytes are still observed.
Another line of evidence suggesting a role for CD45 in T cell development came from thc gcneration of mice expressing a CD45RO transgene (Ong et al., 1994). The augmcntcd expression of the phosphatase led to a marked reduction in numbers of DP thyrnocytcs, which correlated with increased Lck kinase activily and enhanced calcium influx in responsc to TCR ligation (Ong et al., 1994). By examining the effect of CD45RO overexpression on a ncgativc sclccting background, investigators found that there was an enhanced MHC-rcstriclcd ncgativc selcction ol'anti-HY TCR-bcaring DP thymocytes (Ong el al., 1994), suggesting that CD45 could rnodulatc TCR signals to cffcct ncgativc sclcction proccsses.
A third linc of evidence implicating CD45 in T ceIl development involved neonatal injection of a pan-reactive anti-CD45 antibody. This treatment was found to inhibit diffcrentiation of DP thymocytcs in10 mature SP thymocytcs (Benveniste ct al., 1994). To examine the role 01' CD45 in positive selcction, the same antibody treatment was perfomed in fernale mice espressing trdnsgcnic TCR spccific for the male HY antigen. The development OC CD&+SP was inhibitcd in a ktal thyinic organ culture system, suggesling thai lhc CD45 engageincnt inhibitcd positivc sclcction (Bcnvcnistc ct al., 1994). IntcrestingIy, anli-CD45 prc-trcalmcnt of primary lyinph ncxic CD4+ T cclls inhibits many TCR-induced activation priramctcrs, such ris Ca2+ mobilixation and DNA synthesis (Maroun and Julius, 1994). In marked contrat, the samc lrcatmcnt had littlc cffcct in CD8+ T cclls (Maroun and Julius, 1994). This implies thal thc two lincagcs havc a diffcrcntial rcquircmcnt for CD45 inT ce11 activation. Thc mechanism bchind lhcsc obscrvaiions has no1 bccn dcfined. Conccivably, thc addition ol'anti-CD45 could inducc a rcdistribution ol'
CD45 such that it no longcr funclions to rcgutate TCR signals. Altcrnativcly, ligat~onol' CD45 - -.-. - - -.. - - . - .-. - - - - -.- - -. .- - - - - . - - - a - .-. - . ------.- - ,, - ---..-.-- - . ..--~. r. - -.---~ ------~ - --- support of thc l'ormcr hypothcsis. By disriipting thc physical association bctwccn CD4 and CD45 in CD4+ T cells (Bonnard et al., 1997). anti-CD45 treatmcnt may spccif'ically prcvcnt thc appropriatc participation of CD45 in theTCRlCD3 complcs. Clearly, regulation of CD45 function at the DP stage of T-ce11 deveiopmcnt has many potcntial implications on the outcomes of positive and negative sclection. Intriguingly, a rcccnt report suggests thal CD45 may be involved in setting thresholds mediating B ccll selcction cvcnts (Cystcr et al., 1996). Investigators assessed positive and negative seleclion in CD45-del'icicnt micc esprcssi ng immunoglobulin transgenes specific for hen egg lysozyme (HEL) in the prescncc or absence of thc autoantipen. Their data showed that the absence of CD45 rendcred B cclls hypo-responsive 10 BCR stimulation iti vitro, thus resulting in the positive, rather than thc normal ncgative sclcction of HEL-spccific B cells in vivo (Cyster ct al., 1996). Thcsc results support a signal-threshold mode1 for B-ceIl selection, and provide a framework for investigating the role ol' CD45 in T-cell sclection. as wcll.
D) Thesis Objective
The goal of my project involves the characterization of signaling proccsses rcgulating positivc selection. To address the rolc of CD45 in TCR signaling in DP thyrnocyes, 1 havc characierized a CD45-dericicnt DP thymomaccll linc called 3T7. Using this modcl systcm, 1 havc esamincd ihc requirements and rcgulatory rolc of this molcculc. By rclating thcsc obscrvations io the itt vivo counterpart, thc ~D45-1-mousc, 1 proposc that thcsc signalrng rcsults providc a salional basis for undcrstanding thc dcvclopmcnhl phcnotypc of CD45-/-niicc. Thcsc ~cslrltsfurthcr dcfinc thc rch of CD45 in rcgulating TCR signals at thc DP siagc of dcvclopnicnt, whilc providing a rational fianicwork Irour basic undcrstanding 01' antigcn rcccptor signaling pi*,ccsscs. CHAPTER 2
MATERIALS AND METHODS 3T7 derivatives are dl subclones of AKR 3T,ü thyrriic DP lymphoma ceIl linc (Grovcs ci al., 1995). 3T7 was derived from a spontaneous thymoma in AKR mice. The AKR strain of mice was bred for high incidence of thymornas, associated wi th endogenous retroviruses (Rowc and Pincus, 1973). Ce11 lines were maintained in growth medium (GM) consisting of RPMI 1640 supplemented with 5% FCS, 5sl0-5 M 2-ME, 10 mM glutamine, and 10 mM HEPES, pH7.4. 3T7.GC 14 was derived by sorting 3T7 cells that had becorne CDS~~aster overnight cross1 inking wi th H57-5971GK 1.5 heteroconjugate antibody (see Anti bodies). CDS~~cells werc defined as those cells with a mean fluoresence intensity (Mm) greater than 100. Unstimulatcd cells contained less than 1% CD* cells and had a MFI of 30. Pervanadate was freshly made by mixing 10 mM sodium orthovanadate (Sigma Chernical Co., %Louis, MO) with 10 mM Hz%.
The 10 mM stock pervanadate solution was incubated at RT for 10 min , prior to addition to thc the ce11 culture. Herbimycin A, generously provided from Dr. P Benveniste (Hospital for Sick
Childrcn, Ont), was added to ce11 culture at a finai concentration of O. 1 PM.
B ) Retroviral Gene Transfer
Retroviral-mcdiatcd genc transfer was used to express CD45RO in 3T7.GC 14. Thc plasmid T200/0 provided by Dr. P. Johnson (University of British Columbia, BC) (Johnson ct al., 1989b) contains Su11 lcngth murine CD45RO cDNA cloncd into thc rctroviral vcctor pARV 1
(McLachlan ct al., 1987). pARV 1 also contains the neom ycin phophotransfcrasc genc (MI) which confcrs rcsiscüncc to G418. Al tcrnative splicing of transcripts drivcn by thc 5' long tcrminal repcat (LTR) gcncratcs CD45RO and neo messengcr RNA (mRNA). DNA conslruçts wcrc transfcctcd by calcium phosphate prccipitation into thc fibroblast ccotropic viral packaging l inc, GP+E (Markowitz ct al., 1988). A pol yclonal retrovirus-producing cc1 l linc, GP+EIT30010 was cstablishcd by growth in the prcscnce of 500 pglrnl G418 (Gibco, Grand Island, NY). To infcct 3T7.GC 14 cells, 5- 10 s 106 cells were CO-incubatedovcrnight with irradiakd (7000 riid) UlTL1 1 -UUIU. LlUllUI IIIILCILUIIL~IYbLI, UULUIIIbU Ur 11111111115 UlIUllUll 111 UlVl L\J11LLI11III1~tlJU tL$llll
G418. Surviving cells wcrc cvaluatcd for expression of' CD45 by tlow cytornclric analysis
(FACS) analysis using a fluorcsccin isothiocyanate (F1TC)-conjugated pan-CD45 spccil'ic antibody (ALI-4A2). Posilive cells were expanded and maintained in GM containing 300 pg/rnl
(3418. As a negative control, paraIIel infections were done as above, espect using thc q2-ncci packaging line, provided by Dr. D. Vaux (WEHI, Melbourne) to generate 3T7.nco subcloncs. The $-ne0 packaging line contains the pLXSN retroviral vector containing the rieo gene dri\..cn by thc SV40 early promoler (Vaux et al., 1988).
C) Antibodies and Second Stage Reagents
Thc following monoclonal antibodies (MAb) were used in this study: anti-CD4 (GK1.5) (Dialynas et al., 1983) and (YTS-191.1); anti-CD& (YTS-169.4) (Cobbold et al., 1984) and (53-6.7);anti-CD5 (53-7.3) (Ledbetter and Herzenberg, 1979);anti-CD69 (H1.2F3) (Yokoyama et al., 1988); anti-TCRp (H57-597) (Kubo et al., 1989); and pan-anti-CD45 (ALI-4A3)
(Spangrude and Scollay, 1990). Antibodies were purified from tissuc culture supcrnatanls coniaining 5% NuScrum (Collabontivc Rcsearch Inc., Bcdford, MA) instead of' FCS, by protcin
G- or protcin A-Scpharose chromatography, and conjugated to FITC or biotin using siandard tcchniqucs. Avidin-phycoerythrin was purchased from Caltag (San Franscico, CA). All antibodics wcrc pi-c-titratcd and uscd at saturating conccntrations. Rat IgGb, rat IgG2.,, and hainstcr IgG isotypc control antibodies conjugated to fiuoresccin, biotin, or phycocrylhrin wcrc purchascd Srom Pharmingcn (San Dicgo, CA), and used at saturating concentrations OS 1-5 pg/1111. The l'ollowing antisera wcrc used Ior weslern blol detcction andlor immunprccipiilitions: anti-lck polyclonal rabbit antiscra, kindly providcd by Dr. M. Julius (University 01' Toroiilo, Toronto, Ont); anti-Fyn polyclonal rabbi1 anliscra for wcslern blotting (Dr. A. Vcillcltc; McGill
University, Montreal, Quc), or immunoprecipi(ations (Santa-Cruz Biotechnology, Santa Cruz,
CA); anti-CD45 polyclonal rabbit anliscra, kindly providcd by Dr. P. Brrinton (McGill d. . - ,. , . , % 1994); and anti-phosphotyrosinc MAb 4G 10 (Upstate Biotechnology, Lake Placid, NY). Sccondary reagents used in western blot detection were: horseradish perosidase (HRP)- conjugated protein A and HRP-conjugated strepavidin, purchased from Amersham Corp. (Arlington Heights, IL); and HRP-conjugated goat anti-mouse IgG, purchased from Bio-Rad Laboratones (Richmond, CA). To generate heteroconjugate anti bodies (H57-5971GK1.5 and H57-597lYTS 169.4) by chemical coupling, H57-597 was modified using 15 pg N-gamma- maleimidobutyryloxysuccinimide (GMBS; Calbiochem, San Diego, CA) pcr mg of antibody, while GK1.5 and YTS- 169.4 were modified using 250 pg 2-iminothiolane HCI (3-IT; Picrce,
Rockford, IL) per mg of antibody (Lcdbctter ct al., 1989). GMBS is a hcierobifunctional crosslinking reagent which introduces thiol-reactive maleimides into the prolein. ModiIïed antibodies were purified through PD-IO columns (Pharmacia, Baie d'Urfe, Que) to removc unreacted GMBS and 2-IT. The GMBS modified antibody then reacts with the second thiol- conlaining 2-IT modified antibody to form a stable thioester crosslink (Fujiwara et al., 1881). This reaction was performed by mixing the GMBS-modified H57-5!37 and 3-IT-modificd GK 1.5 or YTS 169.4 together for 1 h at RT in coupling buffer ( 100 mM Na2HP04, pH6.8; 50 mM
NaCI). The reaction was quenched by adding 5x10-2 M 2-ME for 15 min and ihen O. 1 M N- cthylmaleimide in DMF a further 15 min. Heteroconjugates were purified through PD-IO columns to exchangc coupling buffer with PBS + 10 mM NaN3.
CNBr coupling of antibodies was pcrformcd üccording 10 manufacturer's instructions.
CNBr beads (Pharrnacia, Baie d'Urfc, Quc) were swelled foi- 30 min in 1 mM HC1 and rhen washed thrcc tiincs in coupling buffer (0.1 M Na2HC03, pH 8; 0.5 M NaCI). 1 nig of pui-ificd antibody was added per 79.12 mg ol' CNBr beads (dry wcight). The coupling rcaclion wns pcrl'orined for 3 houis al room Lcmperature with constant mixiiig. Thc proccdurc crcütcs a covülcnt bond bclwccn CNBr-aclivated bcads and ligands conlaining primary amincs. Thc rcaction was qucnched by addilion of 0.1 M Tris, pH 8 for 2 hours al room lemperaturc undct- consianl shaking. Couplcd bcads wcrc washcd three limes in coupling bul'f'cr, Lwicc in PBS, and .-. antibodies to protein-G-sepharosc beads, 10 pg of purified antibody was addcd pcr 50 ol' a
25% protein-G slurry. The beads and antibody wcrc rniscd ovcrnight at 40~,washed thrcc
Limes in PBS and rcsuspcndcd in PBS + O. 1% NaN3 al a final mtibody conccniralion ol 1 p$p1.
D) Flow Cytometry
For llow cytornetric anal.yses, exponentidly growing cells or freshl.y isolated thymocytcs wcre harvested in staining media (SM), consisting of HBSS plus 3% calf scrum and 10 mM
HEPES, pH 7.4. Cells werc stained by incubation of 0.5 - 1.0 x 106 cclls with satunting amounts of FiTC- or biotin-conjugated antibodies at a concentration of 1 x IO7 ccllslml on icc for
20 min. The cells were washed in SM and resuspended at a concentration of 0.5 - 1 .O s 10" cclls/ml in SM plus 1 &ml propidium iodide. Flow cytornetric analyscs werc pcrl'orrned on a FACScan (Becton Dickinson & Co., Mountain View, CA). Data was acquircd on al lcast
10,000 cells/sample and anal yzed using an HP 340 computer and Lysis II software. Dead cclls, idcntificd by thcir low forward scatter and high propidium iodidc fluorcscencc, wcrc escludcd from analysis. Note that x axis labels on al1 histograms have becn converted by the dab analysis software from thc fluorcscencc channel number (nonlinear scalc) to mcan lincar fluorcsccncc intcnsi ty. Because unstained cul tured ceIls exhibi t varying degrces of autolluorcscencc that can intcrfcre wi th discrimination of fluorcscencc due to antibody binding, fluorcscencc coin pensalion was uscd to subtract ccllular autofluorescence (Alberti et al., 1987). This tcchnique I'acilitatcd dctcction of low Icvcls of surfacc staining on highly autofluorcscent culturcd cclls.
E) Stirriuiation of cells for assessment of tyrosine phosphorylation
Cells were harvested in early lo mid-log phase, counted and aliquoted at 30 s 1061tnI prior to staining for 20 min at OC with saturating concentrations of biotinylatcd anli-TCRp
(H57-597), biotinylatcd anli-CD4 (GKIS), biotinylatcd anti-CDXtr (53-6.7), donc or in b\IIIIVIIIULIVII. I IILd bC1I1J VVb1Li IILWIIbU UllU LbIILIO~bIIULdUUL -.J i\ 1" 11111, YlLd I>ULIIIbU LI, J I b 1111
3 min, and cross-linking was pcrformcd for 1 min by adding 5 &ml avidin (Molccular Probcs, Eugcnc, OR). Cc11 stimulation was tcrminated by adding 10 volumcs of icc-cold phosphatc- buffcrcd salinc (PBS) containing 400 pM Na3VO4. Cell pcllets wcrc thcn rcsuspcndcd at 5 s
10~1mlin lysis buffcr (LB) containing 1% NP-40 (Fluka), 50 mM Tris, pH 8.0, 20 mM EDTA,
30 pM Na3VO4, 50 mM NaF, and 20 pglml leupeptin and aprotinin. Nuclci wcre pellctcd by centrifugation (12,000 x g at 4*~for 10 min) and the postnuclear proteins werc scparatcd by
SDS-PAGE gcls (8%non-reducing) (Laemmli, 1970) and then transferred to nitrocellulose, and immunoblotted with anti-phosphotyrosinemAb (4G10).
F) Stimulation of cells for assessment of surface phenotypic changes
Cclls wcrc harvcskd in carly to mid-log phase, and cultured at 37O~overnight in 6-well tissue culture plates (1 x 10~/well;Coming, New York, NY), pre-coated with PBS or various protein A- or protein G-purified antibodies at 5 pglml, the optimal conccntration determincd by titration cxperimcnls, for 2 hours at 37O~.The antibodies used for stimulation werc: anti-TCRp (H57-597), anti-TCRVanti-CD4 heteroconjugate (H57-597/GK1.5), anti-TCRplanti-CD& heteroconjugate (H57-597lYTS-169.4), GK1.5, and YTS- 169.4. To bypass stimulation through the TCR, PMA (Sigma Chemical Co., St.Louis, MO) and ionomycin (Calbiochem, La
Jolla, CA) werc added at 10 nglml and 250 ngiml, rcspcctively. The cclls were hawestcd aftcr 1 or 2 days in culture and stained to quanlitate expression of surfacc CD5 (53-7.3), CD4 (YTS-
191. l), or CD8 (53-6.7)using FITC-conjugated anlibodics.
G) Iinmunoprecipitations
Cclls wcrc lyscd at 5 x lo7/ml in LB for 20 min at OC, csccpl for cclls to be subjcct LO anti-TCRp iinmunoprccipiiations (IP), which were lysed in 1% digitonin (Wako; Richmon, VA)
LB for 1 hr at OC. Lysatcs (10 x 106 cc11 cquivalents) wcrc ccntrifugcd (13,000 s g Ior 10 min ". . -, '- .-*..-. - ...L,Y.-Y.V ...IL 1I.l. ...a- ...W.. r.--.--. -- "J ".--"L'.\'.. "L*.. d.' p..L \,A U -2 'L protcin G-Sepharosc slurry overnight at 40C undcr constant mixing. Precleared matcrial n'as rcmoved by centrifugation (13.000 s ç for 10 min al OC), and lhc rcmaining supernatant was spli t into two equal volumes (5 x 106 cell equivalents) for IP. IPs wcrc performed for 1 hr at
40~with constant mixing using 50 pg of antibody pre-couplcd 10 either protein G-scpharosc bcads (Pharmacia, Uppsala, Sweden) or CNBr-activated-sepharosc beads (Pharmacia, Uppsala, Swcdcn). Control IPs were performed identically but with uncouplcd IP bcads, unlcss othcrwise indicated. Sequential IPs with either anti-CD4 (GK1.5) or anti-CD8 (53-6.7)an11 body ivcrc performcd as abovc esccpt that following the intial IP, the supernatant wu transfcrred to a frcsh tube and re-immunoprecipitated with an equivalent amount of antibody for 1 hr at 40~with constant mixing. The procedure was repeated a toial of five times. Immune complexes wcrc washed 3 times in LE More king resuspended in non-reducing SDS-PAGE loading bulTcr and boilcd for 5 min. The beads were removed by centrifugation (13,000 x g for 2 min) and protcins were resolved by SDS-PAGE (Laemmli, 1970) and then lranslcrrcd to nitroccllulosc mcmbrancs for Western blotting (Towbin et al., 1979).
H) Cell Surface Biotinylation
Cells werc washed three times in ice-cold biotinylation buffer (HEPES buffcred sülinc (HBS), pH8.8; 1 mM MgCl-; and O. 1mM CaCI-) and incubaled at 4OC for 20 min at 20 x 106 ccllslml in bioiinylation buffer containing 0.5 mglm1 sulfo-NHS-biolin (Piercc; Rocklord, IL).
The NHS ester of biotin reacts wi th the deprolonated form of primary amincs (Iysinc rcsidues) to crcate a peptidc bond. The rcaction wüs qucnchcd by addition of biotinylation buffèr cciniaining
35 mM NH4CI. Cclls wcrc ivashed twice and then counted. A small aliquot of biotinylatcd cclls was stained with avidin-PE to assess the degrcc of biotinylation by flow cyloinelry (biotinylatcd cclls typically possesscd 1000-fold higher MF1 over unbiotinylated cclls). Thc rcrnaindcr oi'cclls wcrc lyscd al 5 s lo7 ccllslml in LB with 20 &ml lcupeptin and aprolinin rcx- 20 min at 40~. immunoprccipitation.
1) SDS-PAGE and Irnrnunoblotting
Samples werc boiled for 5 min in SDS sample buffer and elcctrophoresed on 8% or 10.5% SDS-PAGE gels under non-reducing conditions unless othenvise noted. The gcls werc thcn equilibrated in Towbin's transfer buffer (25mM Tris; 193 mM glycine; 30% methanol) and trmferred to nitroccllulose membranes at 72 V for 1.5 h in a Transblor apparalus (Bio-Rad
Lüboratories, Mississauga, Ont). After transfer, blots were blocked for 1 h at room icmpcraturc in TBS-T (O. 1% Tween-30, 10 mM Tris, 2.5 mM EDTA, 50 mM NaCI) with 5% wlv carnation milk and then probed for 1 h wilh primary antibody diluted in 2.5% milk protein (1:5000 I.or anti-lck, 1:800 for anti-Fyn, 1: 1000 for anti-CD45, and 1 &ml foranti-TCR1;). The blois wcrc washed for 10 min with TBS-T and then incubated 1 h with protein A-HRP (15000, Amcrsham
Life Science, Oakville, Ont), after which they were washed five times for a total of 1 h. Antiphosphotyrosine blots were performed as above exccpt that blocking and rintibody incubations were perforrned in 2.5% bovine serum albumin (BSA, I'raction V, ltty acid-frec; Bochringcr Mannhcirn, Indianapolis, IN) in TBS-T. The blots were subsequently developcd with the cnhanced chemiluminescence detection assay (ECL kit, Amersham Lifc Scicncc). Blets
10 bc rcprobcd wcrc strippcd by vigorous washing in 10 mM Tris, pH2.3; 150 mM NaCl for 20 min, f'ollowed by two washes in 10 mM Tris, pH8; 150 mM NaCl for IO min. Dcnsitomctry was performed to asscss relative amounis of protein using a Protcin Databases Inc. Discovciy Scrics DNA 35 Dcnsitomcter.
J) RNA preparation and Northern analysis
Total RNA was crtractcd Sroin cells by thc Trizol acid guanidinium singlc stcp rnclhod
(Molccular Rcscarch Ccntcr, Cincinnati, OH), and quantitatcd by UV spcctropholomctry. RNA 1 \ 'C., .. 3.2M tormaldehydc, 1s MAE) and clcctrophoresed in 1s MAE in a denaturing 1.3%)aagrosc gel conlaining 2.2M formaldehyde. Thc gel was blotted by capillary transfer in 30s SSC onto Zcla- probc GT membrane (Biorad, Mississauga, Ont). Membranes wcre UV cross-linked (1300 J, Stmtalinkcr, Stratagcne, La Jolla, CA) and prehybndized overnight at 420~in of 50% dFA; 5X SSPE; 3X Denhardt's; 0.5% SDS; 100 &ml salmon sperrn DNA. BIots werc ihcn hybridizcd with 32~-d~~~-labelledDNA fragments (106 cpmlml) for 1 hr at 42OC with constant shaking, washed twice with 1s SSC/O.l% SDS at RT, and twice with 0.2X SSC/O.I% SDS al G°C. DNA probes wcre labelled using T7QuickPrime Kit, according to manufacturer's instructions (Pharmacia, Baie d'Urfe, Que). Unincorporated nucleotides were removed by passing the probc through NICK-Columns (Pharmacia). The hybridization probes used in this study werc: human 0-actin cDNA (0.6 kb Pst-1 fragment); murine RAG-1 cDNA (1.0 kb Bgl-II fragment) (Schatz el al., 1989), and murine CD45 cDNA (1.35 kb XbaI fragment), kindly provided by Dr. H. Ostergaard (U. of Alberta, Canada). BIots were exposed to the Molecular Dynamics Phosphorimager (Sunnyvale, CA) system and then analyzed using Imagequant 3.0 software (Molccular Dynamics).
K) cDNA synthesis and RT-PCR analysis
cDNA wüs synthcsizcd iri a 25 pl reaction with 5 pg of totül RNA, Promcga (Madison, WI) RT buffer (final concentration: 50 mM Tris (pH8.3), 75 mM KCI, 3 mM MgCI2, 10 mM
DTT), 0.5 dNTPs (Pharmacia, Baie d'Urfc, Quc), O. 1 mglml BSA (Bochringcr Mannheim,
Indianapolis, IN), 40 U RNAsin (Proii~cga),0.5 FM random hcsamcrs (Gi bco BRL, Grand Island, NY), 2.5 U Promcga AMV RT. For the RT-PCR analysis of CD45 isolorms, oligonuclcotides corrcsponding to thc mutine CW5 cDNA Genbank scqucncc (Johnson cl al.,
1989a) wcrc synthcsized: thc scnsc primcr (ATG ACA GCT GAT CTC CAG ATA TGA CCA
TG) corrcsponding 10 positions 110-134, and lhc anti-scnsc primcr (ATG AGT CGA CAA TCC
TCA TTT CCA CAC TTA GC) corrcsponding to psilions 821-2344, wcrc uscd. For thc murinc L b positions 330-340, and thc anti-sense primer (CAC GCA GCT CAT AGC TCT TCT) corrcsponding to positions 790-8 10, were used. RT-PCR reactions, camed out in 35 PI, containcd Taq buffcr (final concentration: 10 mM Tris (pH9.0), 50 mM KCI, 1.5 mM MgCI2, 0.1% (WIV) Triton-X 100), 200 pM dNTPs
(Pharmacia, Baie d'Urfe, Que), 0.5 pM 3' and 5' primers, 1.35 u of 'Cïrertnirsnqzinticra (kq)
DNA polymerase (purchased from Dr. J. Friesen, Hospital for Sick Childrcn, Ont). Sarnplcs wcre overlaid with 50 pl mineral oil (Sigma Chemical Co., St.Louis, MO) to prcvent condensation and subjected to 37 cycles of amplication using a programmed thermal cycler
(Perkin Elmer Cetus, Nonvalk, CT). For the CD45 PCR, each cycle consisted of denaturation at 940~for 1 min, annealing and extending at 72% for 2.75 min. For p-actin PCR, samples rvcrc subjected to 30 cycles consisting of denaturation at 94O~for 1 min, anncaling at 55OC for I min, extending at 720~for 1 min. A fter amplication, PCR products were anal yzed by electrophoresis in 7.5% agarosc (1.5% NuSicve and 1% Multipurpose) gcl in a TrislAcetate buffer systcm and DNA prcducts wcre visualized by ethidium bromide staining. CHAPTER 3
RESULTS Previous work in our laboratory characterized a panel of DP thymic Iymphomas. which rcspond to TCR cngagcrncnt in vitro by undergoing multiplc maturation cvcnts associatcd with positive selection in vivo, including increased expression of CD5 and Bcl-2, and dccrcased cspression of RAG-1 and RAG-2 (Groves et al., 1995). These ceIl lines offcr a uscful rnodcl systcm to elucidatc molecular mechanisms involved at this cntical stagc of T ce11 developmcnt.
Onc of thcse ccll lincs, 3T7, was identified as king TCR-non-responsive, failing to undcrgo any of the phenotypic changes observed in TCR-responsive ce11 lines (Groves et al., 1995).
Thc defect appeared to be proximal, in that addition of PMA and ionomycin, pharinocologic agents which bypass carly events in TCR signaling, upregulated expression of CD5 and CD69 (Grovcs ct al., 1995). To attcrnpt to idcntify thc TCR signaling defect in 3T7 cells, the surlàcc expression levcls of various cell markers was assessed. Figure I shows that 3T7 cclls express both of the co-receptors CD4 and CDS, and the great majority of cells (90-95%) stably cxprcss medium levels of surface TCR (MF1 = 11 1), while a small fraction (5-10%) express low levcls of surface TCR (MF1 = 5) by flow cytometry. In constrast, 3T7 cells lack dctcctablc surfacc expression of CD45 by flow cytometry (Fig. 1). Western blot, Northern blot, and RT-PCR analysis also revealed no dctcctablc CD45 expression (see Iater). Thus, 3T7 is a CD45- dcfïcient DP lymphoma ce11 linc.
B) TCR signaling defect in 3T7 cells
i) Early events - induction of tyrosine phosphorylation
Thc carlicst detcctablc cvcnl in TCR signaling is Lhc induction of tyrosinc kinase xiivity, as rcvcalcd by induciblc protein tyrosine phosphorylation (Samclson ct al., 1986; Hsi cl al., 1989; Junc et al., 1990b). Givcn the critical rolc of CD45 in rcgulating Src-family PTKs, thc rcquircmcnts Ior CD45 in TCR signaling with or without Lhc ddibcralc CO-aggrcgalion01' Figure 1: Surface phenotype of 3T7 cells.
Ccils wcrc stained wilh FITC-conjugatcd antibodics spccific for CD4 (GK1S), CD8tr (53-A.7), TCRP (H57-597), or pan-CD45 (ALI-4A2). Shadcd hislograms show lhc lcvcl of staining with isotypc-matched control antibodies. stin~ulatedfor onc minute and tyrosinc phosphorylatcd protcins dclcclcd by SDS-PAGE and Wcstcrn blotting with the anti-phosphotyrcisinc antibody (4G 10). In rcsponse to TCR aggrcgation, thcrc was a marginal increase in the tyrosinc phosphorylation of two protcins 01' around 80 kDa and 90 kDa (Fig. 2A). However, when eithcr CD4 or CD8 was delibcratcly co- aggregated with TCR, a striking induction of phosphoproteins, notably 40 kDa, 70 kDa, 80 kDa, and 120 kDa was obsenled. Crosslinking of either CO-receptoralone rcsulled in only a marginal induction of phosphoproteins, simiIar to that observed in response to TCR stimulation alone. Furthermore, TCRICD4 co-aggregation was far more robust in inducing Lyrosinc phosphorylation than TCRICD8 CO-aggregation. These data suggest that the need for CD45 in early TCR signaling is obviated when the TCR and either CO-receptorare CO-aggregated.
ii) Late events - CD5 induction, RAG-1 downmodulation
To detcrmine whether the proximal TCR defects observed extended to Iater cvcnts in
TCR signaling, 1 assessed the ability of 3T7 cel!s to undergo phenotypic maturilion cvcnts.
Prcvious work has indicated that both fresh ex vivo DP thymocytes and various DP thymoma cc11 lines undergo a number of phenotypic changes, including upregulation of CD5, CD69 and
Bcl-2, as well as dccreased expression of RAG-1 and RAG-2, in rcsponse to TCR ligation irr vitro (Turka et al., 1991b; Brandle et al., 1992; Groves et al., 1995; Kearse et al., 1995; Grovcs ct al., 1997). Furthermore, these events have been show to occur during posilive selcction itr vivo (reviewed in Guidos, 1996). Therefore, 1 determined whcther 3T7 cells could incrcasc CD5 and decrcasc RAG-1 expression in response to TCR signais. Following overnight culturc of 3T7 cclls alonc, or in wclls that had bcen precoated with purified antibodics, flow cyloinctry was uscd to quantitatc surfacc cxprcssion of' CD5 as a markcr of maturationlac~iva~iun. 1-Iclcroconjugate aniibodies, generaled by chernical crosslinking of anti-TCRp and cithcr anti-
CD4 or anli-CD8 anlibodies, were used for stimulations involving thc co-cngagenicnt of TCR with cither co-rcçeptor. Whilc TCR engagement donc fàiled to induce CD5 csprcssion in 3T7 a 2i 6 YP:dc4 Stimulation: j ÿ y n u ZkUE-c U
Figure 2: Effect of CD45 deficicncy on TCR-mediatcd signal transduction in 3T7 cclls.
(A) Induction of tyrosine phosphorylation in 3T7 cells aftcr TCRp or TCRp + CO-rcccptor crosslinking. Cclls were cultured for 1 min at 37OC with or without antibody-incdiatcd cross- linking of thc indicatcd surface molecules. Postnuclear lysates rrom cqual cc11 numbcrs
(0.5~106)wcrc scpaiated by 8% SDS-PAGE (non-reducing), traiisferrcd to nitroccllulosc, and probcd with a monoclonal anli-phospholyrosine antibody (4G 10) followcd by goat anti-mousc
HRP, and dctectcd by ECL. Arrows indicate sevcral prolcins that undcrgo TCR-induciblc tyrosinc phosphorylation. Numbcrs on the left indicate the migration of MW standards. , - - -.- --- .. ---.----- . -- .. ------.-..- - .... -. -- . -. --., ,.-.- dclibcrately co-aggrcgated (Fig. 2B). The ligation of eilher CD4 or CD8 co-rcceptors without
TCR lailed to induce CDS expression in 3T7 cells (Fig. 2B). SimiIarly, a decrease in RAG-1 espression was observed when TCR and CD4 were co-aggregated, but not in rcsponse to TCR cngagemenl alone (see later, Fig. 8C). As was observed for protein lyrosine phosphorylation, TCRICD4 co-aggregation was more efficient at inducing CD5 (70-90% became CDS~~)than was TCRICD8 co-aggregation (20-50s became CD*; Fig. 2B). Two populations of cclls were consislently observed: lhose that were responsive (~~518,MFI>100), and those that remained CDJO,MFI<100. This suggests that 3T7 cells may harbour additional delècts bcsidcs CD45, accounting for the inability of some cells to respond. The signaling capaci ty of the CO-receptorsCD4 and CD8 is large1y attribukblc to thcir association with Lck (Veiliette et al., 1988). Thus, 1 hypthesized that thc ability to inducc TCR signaling events by coaggregation of TCR with CD4 or CD8 could bc attributcd to the activation of Lck when brought into close proximity to the TCR complex. To asscss whclher
TCR plus co-receplor signaling was associated with EKactivity, 1 used the tyrosinc kinase inhibitor hcrbimycin A, which has becn shown to markedly diininish TCR signaling (June ci al., 1990a). The induction of CD5 after ovemight stimulation by TCRICD4 was almost completely abolished when cells were cultured with O. 1 PM Herbimycin A (Fig. 2C). This effccl was dose-dependent, with no obscrvablc cffcct when the dnig was added al 0.01 (Fig. 2C). Thus as expected, the generation of TCRlco-reccptor signals not only corrclates wilh KTK activity, but requires it.
C) Differential ability of CD4 versus CD8 CO-receptorsto overcome TCR signaling defect
It has becn rcp~rtedthat CD4 associatcs rnorc strongly than CD8 wiih Lck in wild-typc
DP lhymocytcs (Wicst cl al., 1993). Thus, differential Lck association might csplain thc grcalcr abilily of CD4 to signal than CD8 when coaggregatcd wi th Lhc TCR. To dctcrminc if Stimulation: Stimulation: Stimulation:
TCRPICD4
.'l,:;: !y:.-,..+.:$.. -..., IO* 10' 10' id 10' io* 10' IO' id 10'
Figure 2B: CDS induction in 3T7 cells arter TCRp or TCRp + CO-receptorcrosslinking. Cells were cultured overnight alone or in culture wells coated with the indicated antibodies: anti-TCRp (H57-597), anti- TCRPanti-CD4 (H57-597lGK 1.5 heteroconjugate), anti-TCRf3Ianti-CD8a (H57-597lYTS- 169 heteroconj ugate), anti-CD4
(GK 1S), or anti-CD8a (YTS-169). Cells were stained with RTC-conjugated anti-CDS (53-7.3). Shaded histograms represent staining with an isotypc-matched control antibody. Stimulation: TCRfYCD4 Herbhycin A: Nooe
Figure SC: CD5 induction in 3T7 cells treated with the PTK inhibitor?Herbimycin A.
Cells were cultured overnight alone or in culture wells coaa with anti-TCRP/anti-CD4 (H57-597lGK1.5heteroconjugate)
and the indicated arnount of Herbirnycin A. Cells were stained with FITC-conjugated anti-CD5 (53-7.3). Shaded histograrns represent staining with an isotype-matched control antibody. tliis was the case, I performed sequential immunoprecipitations OS CD4 and CDS. In aggreernent wi th studies of DP thymocytes, CD4 immunoprecipi tates Srom 3T7 cells containcd signilïcantly greater amounts of Lck than the corresponding CD8 immunoprecipitatcs (Fig. 3A). Lck associates with the cytoplasmic tail of CD& (Veillette et al., 1988). An altematively spliced version of CD8a, CD8a' contains a truncation in the cytoplasmic domain such that it no longer associates with Lck (Zamoyska and Parnes, 1988; Zamoyska et al., 1989). Therefore, 1 assessed whether the low stoichiometry of the association between Lck and CD8 might rellect a low ratio of CD8cJCD&ttin 3T7 cells. Cells were biotinylated to label surface protçins prior to
CD8a immunoprecipitation. Immunoprecipitated proteins were then sepanted under reducing conditions on SDS-PAGE to dissociate the disulfide linked CD8 dimers. Two bands of 40 and 35 kDa were observed, corresponding to CD8a and CD&' polypeptides, respective1y (Fig. 3B). The "smearing" appearance of these bands is likely due to the presence of diffcrentially glycosylated forms of the CD8a polypeptide. Also, the CD8p labels poorly in these procedurcs, and is not readily observed in CD8 immunoprecipitates. 3T7 cells expressed significantly greater amounts of CD&' on their surface than CD8a (Fig. 3B). In contrast, immunoprccipitatcs from unfractionated thymocytes revealed roughly equivalcnt surlacc expression of the CD& and CD8a' polypeptides (Fig. 3B), in agreement with previous studies
(Zamoyska and Parnes, 1988). Thus, the lower association of Lck wi th CD8 than CD4 in 3T7 cells can be explained, in part, by the preferential expression of CD8a' which can not associatc with Lck. However, the low stoichiometry of the association between Lck and CD8 in 3T7 cclls could also bc explained by the intrinsically poor association between CD8n and Lck. This is demonstrated in CDS+ T cells from the periphery (cg. lymph nodes), which prcdominantly express the full-length CD8a polypeptide, yel slill have 10-lold lcss associütcd Lck than Sound in CD4+ T cells (Zamoyska and Parnes, 1988). Thymocytcs Parental 3T7
4- CDScx -- &. T% 4- CDSa'
Figure 3: Molecular basis of differential signaling between CD4and CD8 CO-receptorsin 3T7 cells. (A) Differential association of Lck wi th CD4 and CD8 coreceptors in 3T7 cells. Cells were lysed in LB and sequentiall y immunoprecipitated 5 times with protein G-coupled anti-CD4 (GKlS), protein G-coupled anti-CD8 (53-6.7)or with protein G- coupled total rat IgG (control IP). 2x106 ce11 equivalents were used for the sequential immunoprecipitations, and hl06 cell equivarents were used in the lysate control. Proteins were separated by 8% SDS-PAGE (non-reducing), transferred to nitrocellulose, and probed with rabbit anti-Lck antisera followed by protein A-HRP, and detected by ECL. (B) Evaluation of CD8 coreceptor isoforms expressed in 3T7 cells. Parental 3T7 cells or B6 thymocytes were surface labeled with bioiin and then lysed in LB. Lysates from 5 x 106 (3T7) or 5 x 107 (B6 thymocytes) cells were immunoprecipitated with protein G-coupled anti-CD8a (53-6.7)or with protein G beads alone (control IP). Proteins were separated by 10.5% SDS- PAGE (reducing), transferred to nitrocellulose, and probed with strepavidin-HRP followed by ECL detection. Thc ability of 3T7 cclls to signal when TCR and CO-rcccptor wcrc co-aggrcgatcd suggests that CD45 is dispensable under these stimulatory conditions. To addrcss whcthcr altcrnative phosphatases (PTPases) might be espressed in 3T7 cells, the effèct 01' lhc phosphatase inhibitor, pervanadate was assessed. Pervanadate treatment has been shown to mimic T-cell activation via inhibition of PTPases and activation of some FTKs (Secrist ct al., 1993; Imbert et al., 1994). Accordingly, pervanadate treatment led to a dose-dependent induction of several phosphoproteins, notably 40 kDa,46 kDa, 60 kDa, 70 kDa, 120 kDa (Fig.
4A). 1 proceeded to addrcss whether the induction of phosphoproteins extended to later cvenls, such as CD5 induction. Following overnight culture of 3T7 cells alone, or with pervanadatc, Slow cytometq was used to quantitate surface expression or CD5 While highcr concentrütions of thc drug were toxic after overnight culture with 3T7 cells (data not shown), a 5-fold induction of CD5 was observed after treatment with 10 pM pervanadate (Fig. 4B). This induction was less than the 10-fold induction of CD5 observed when surface receptors werc ligatcd (Fig. 2B) and may rclatc 10 thc absence OC a nucleating physical structurc ont0 which ihc induced phosphoproteins can interact. Nonethcless, thesc results suggcst that altcrnativc
FTPases are indecd expressed in 3T7 cclls, and can function to activatc TCR signal Lrünsduction palhways when perturbed.
E) Defect in CD45 gene expression in 3T7 cells
Having charactcrizcd somc SunctionaI consequences of lhc lack of CD45 on TCR signaling in 3T7 ccIls, 1 next wished to idcntify lhc naturc of thc dcfcct in CD45 gcnc csprcssion in 3T7 cclls. As prcviously staled, Northcrn analysis revcaled that rcsting 3T7 cells do no1 contain dctcctablc lcvels of stcady-stale CD45 mcssagc (Fig. 5A). Thus, the csprcssion is no1 constiiulivc. Howevcr, ovcrnighl slimulalion with TCRICD4 or TCRICD8 hetcroconjugütc antibodics induced CD45 mRNA and suri'acc CD45 protein expression (Fig. Pervanadate
(B) Pervanadate: None 10 pM k,w,$".a , ,,w ,, ,kB"$ t<..,:,.c:,. q p.;::, &.; "':::.,.,.-..,*"Ad'. "Fw: \ p:,::.. .-'+'-..+(ib& .: 1. :j 4 .::- 4. .+L Ar..;... l ,,,-,... *$, 1 q'lB' 1 1W "'î
Figure 4: Effect of the tyrosine phosphatase inhibitor, pervanadate on 3T7 cells.
(A) Induction of tyrosine phosphorylation in 3T7 cells after pervanadate treatment. CeIls wert cultured for 10 min at 37'~after addition of the indicated amount of pervanadate. Postnuclear lysate! froni equal ce11 numbers (0.5~109 were sepamted by 8% SDS-PAGE and probed for phosphotyrosini as described in Fig. 2A.
(B) Induction of CD5 in 3T7 cells after pervanadate keatment. Cells were cultured ovemight dont or with the indicated amount of pervanadate. Cells were stained with FITC-conjugated &-CD:
(53-7.3). Shaded histograms represent staining with an isotype-matched control antibody. VL3-3M2 Parental 3T7
Figure 5: Ailalysis of CD45 gene expression in 3T7 cells.
(A) Noi-thei-nandysis of CD45 niRNA expression in 3T7 cells cultureci ovcrnight alone or with indicakd iitimobilized anlibodies. Tolnl RNA was separated on formaldehyde-agarose gels, blottcd onto nylon riieiiibrane, probed witli 33-P-
Iribelled RAG-1 and p-actin cDNA fragiiients, and exposed to a phosphosima~es scieeii. VL3-3M2 (positive control) is a DP ttiyinonin ce11 linc whicli expicsses
high lcvcls of suiflice CD4.5. ~~ ~--- . , C ----- esprcssion of CD45 than was TCRICD8. The pattcrn of CD45 alternative splicing and its changes during lhymacyte development suggcst that it is an important mcchanism for controlling CD45 function. To idcntify which
CD45 isoform is re-expressed, 1 performed RT-PCR analysis on cDNA from rcsting and
stimulakd 3T7 cells. Primers specific for exon 2 and cson 9 were designed such that mRNA containing differcnt combinations of variable exons 4, 5, 6 and 7 would generatc RT-PCR products of different sizes, as indicated in Fig. 5C (top panel). While resling 3T7 cells contained no detectable CD45 mRNA products, stimulated 3T7 cells contained RT-PCR products whose size corresponded to the expected products of thc CD45R(O) and CD45R(-1) isoforms (331 and 257 bp, respectively, Fig. 5C). These bands co-migrated with the PCR products generated frcim the positive control, wild-type thymocytes, which pre-dominanily expressed CD45R(O) and CD45R(-1) isoforms (Chang et al., 1991). Thus, 3T7 cells cxpress littlc or no steady state CD45 message, but can be stimulated to express isoforms of CD45 cxpressed in normal thymocytes. Together these data are consistent with a defect at thc transcriptional level of CD45 However, it is not possible to formally cxcludc a dcfcct in CD45 mRNA stability. I next sought to determine whether CD45 exprcssion in slimulated 3T7 cells was
functional. 3T7 cclls wcrc cultured overnight alone or wiih PMA -t ionomycin to inducc CD45,
and thcn subjccted to f'unctional assays of TCR signaling. 1 chose 10 use PMA + ionomycin to inducc CD45 because these agents bypass surfacc ligation events, and thus avoid the problem of reccpior internalizalinnlblocking that would occur if anti -TCRlanti-CD4 heteroconjugatcs wcrc LISC~.Thc PMA + ionomycin-trcated ceils relaincd surfacc TCRF csprcssion (data not shown) and bccümc lransienlly CD45-positivc (Fig. 6, top panel) bcîorc gradually losing CD45
cxprcssion over the course of 34 hours (data not shown). PMA + ionomycin-lrcatcd 3T7 cclls wcrc now able to rcspond to TCR ligation with the induction or protcin tyrosinc phusphorylalion (data not shown), and also incrcascd CD5 esprcssion in rcsponsc to TCR cngagcmcnt (Fig. 6, bottorn panel). Notc that the PMA + ionomycin trcatcd cclls csprcsscd Stimulation: Stimulation:
1
Figure 5B: Surfacc CD45 inducibly cxprcsscd in 3T7 cells. Cclls wcrc culturcd ovcrnighr donc or in cullurc wclls coatcd wilh thc indicalcd antibodics: anti-TCRp (H57-597), anti-TCRpianti-CD4 (H57-5971GK1.5 hcteroconjugalc), or anti -
TCRplanti-CD8a (H57-597iYTS-169 helcroconjugaie). Cells wcre staincd wilh biolinylatcd pan-anli-CD45 (ALI-4A2) followcd by avidin-PE. Shaded histograms reprcscnt staining with an isolypc-matchcd conlrol anli body. Isoiomi Namc SpUced Producl PCK Product (bp) RAl3C 2-3-45-6-7-8-9 743 €3 2-3-47-8-9 460 RB 2-3-17-8-9 478 RC 2-3-6-7-8-9 472 RAB 2-3-4-5-7-8-9 607 R4C 2-3-4-6-7-8-9 601 RBC 2-3-16-7-8-9 619 R(O) 2-3-7-8-9 33 1
Figure 5C: RT-PCR analysis of CD45 isoforms inducibly exprcssed in 3T7 cells.
Thc schematic represenis the exonlintron structurc of' thc unspliced CD45 prc-mRNA (not t( scalc). The # of nuclcotidcs in cach cson is indicated insidc the boxes. Location ol' lhc prlmcc used for thc RT-PCR analysis arc dcpictcd by thc arrows. Thc cspectcd sizcs of RT-PCI. product from cach isoform arc show in thc table. Cells werc cultured overnight donc or witl anti-TCRpIanli-CD4 (H57-597/GK1.5) hcteroconjugatc antibody. cDNA was rcvcrsl transcri bcd îrom tolaI RNA and PCR-ampli ficd wi th primcrs spccific for CD45 and pactii
(contra! -RT sampIcs wcrc subject to the samc trcatmcnt cxccpt rcvcrsc transcriptase was no added). RT-PCR products wcrc scpamtcd by agarosc gcl and stüincd with clhidium bromidc. Parental 3T7 "CD45-Positive"3T7
Stimulation:
Figure 6: Rc-expression of endogcnous CD45 comelatcs with restoralion of TCR rcsponsivcncss. Cclls wcrc culturcd ovcrnighl alonc or with PMA + Ionomycin, harvcslcd, and a small aliqunt was slaincd with FITC-conjugatcd pan-CD45 antibody (ALI-4A2) or an isolypc niatchcd control antibody and anaIyLed by fl ow cytomelry. Thc rcmaining cclls wcrc rc-çul~urcd ovcinigtit alonc, or wiih immobilizcd anti-TCRP (H57-597). Cclls wcrc haivcslcd and siaincd wilh FITC-conjugatcd anti-CD5 (53-7.3) (open histograms) or an isolypc inalchcd conLr.01 antibody (shadcd histogsains). bdL.
= 100), but this was furthcr incrcascd by anti-TCR stimulation in a subsct of cclls (MF1 =
2000). The rcsults suggest that thc cndogcnous CD45 1s functional, and ablc to rcstorc, at lcast in a subset of cclls, thc ability to rcspond to TCR ligation.
F) Re-expression of exogenous CD45 in 3T7 cells
The above resulls demonstrated that effective coupling of the TCR to TCR signal transduction machinery correlates with the expression of CD45 Horvever, it remained possiblc that other changes induced by the PMA + ionomycin treatmenl could be indirectly responsiblc for restored TCR responsiveness, as opposed to a direct effect of CD45 re-expression. To determine if the primary signaling defect in 3T7 cells was due to Lack of CD45 1 expressed exogenous CD45 and assessed i ts functional consequences on TCR signal transduction. 3T7 celis were infected with a retroviral construcl containing the cDNA encoding the 180-kDa CD45RO isoform of murine CD45 (Fig. 7A), the isoform normally expresscd in DP thymocytcs (Chang cl al., 1991). Aftcr G418 selection, clonal infcctrints (obiaincd by limiting dilution) expressing CD45 were identified by llow cytometry. Western analysis showcd that the two infectanls, 3T7.CD45 C1.l and 3T7.CD45 C1.3 cxprcssed a single specics of CD45 prolcin corrcspanding to thc cxpcctcd 180-kDa band, while parcntal 3T7 cclls and 3T7.nco cclls contained no detectable CD45 protein (Fig. 7B). The siaining characteristics of parcntal 3T7 cells and two clona1 infectants, 3T7.CD45 C1.l and 3T7.CD45 C1.2, are presentcd in Fig. 7C.
Ench infectant cxprcsscd surlacc CD45 protein, while parental 3T7 cells and a negativc conlrol infectant, 3T7.nco did not slain übove background. Clona1 infcctants wcrc also staincd for CD4, CD8, and TCRP and wcrc found to express similar levcls of tliesc surl'acc markcrs, whcn coinparcd to parental 3T7 cclls (data not shown). Figure 7: Expression of cxogcnous CD45 in 3T7 cclls by relroviral-mediated gcne transfer. (A) Schcinatic representation of the CD45R(O) retroviral construct iised in this study. (B) Western blot aiialysis of CD45 protcin levcls in 3T7 inrectants. Post-nuclear supernatants were prcpared from 2.5~106cells lysed in LB. Lysriies were separated by 8% SDS-PAGE (non- rcducing), transferred to nitrocellulose, and probed with CD45-specific antiscra (#788/9-4) followcd by protein A-HRP, and developcd by ECL deiection. (C) Surface CD45 expression in G41X-sesistrint 3T7 infectants. Parental 3T7 cclls, 3T7.nco (3T7 infèctcd with neo-control constsrrct), 3T7.CD45 Cl. I and C1.2 werc stainccl with FITC- corljugatcd pan-CD45 (ALI-4A2). Sliaded histogra~i-isshow ttic lcvcl of stnining with isotype- niatclicd coiitrol nntibody. i) Early events - induction of tyrosine phosphorylation
To esaminc the lunctional consequences of CD45 re-expression on TCR signaling in
3T7 cells, 1 analyzed CD45 infectants for tyrosine phosphorylation upon TCR cngagcmcnt.
Comparing the basal phosphorylation state of proteins in parental 3T7 cells to 3T7.CD45 Cl. 1,1 observed that the overall level of phosphoproteins was similar, except for a 32 kDa protein, which is hyperphosphorylated in parental 3T7 cells (Fig. 8A). This suggests that CD45 csprcssion is not required for the maintenance of the overall tyrosine phosphorylation homeostasis of most phosphoproteins in 3T7 cells. In response to TCR crosslinking, the
3T7.CD45 Cl. 1 infectant regained i ts abili ty to induce protein tyrosine phosphorylation (Fig.
8A). The degree and pattern of phosphoprotein induction was similar to that obscrvcd in parental 3T7 cclls when stimulated with TCRICD4. Similar results were obtaincd in another i ndependently derived infectant, 3T7.CD45 Cl .2 (data not show n). These results indicatc that parental 3T7 cells require expression of the CD45 glycoprotein for the induction of protcin tyrosine phosphorylation in response to KRligation alone.
ii) Late events - phenotypic maturation, RAG-1 downmodulation
To asscss whether the rescue in proximal TCR signaling evcnts observed in 3T7.CD45 infcctants extended to downstream events, 1 evaluated whether TCR engagement coiild indiicc csprcssion of CD5, as well as othcr maluration cvents associatcd with TCR signaling in DP ihymocytcs. As prcviously obscrvcd, TCR engagement of parcnial 3T7 cclls did not causc
CD5 induction. Howcvcr, a largc fraction of the 3T7.CD45 Cl. 1 infcctünts (75%)rcspondcd to
TCRp ligation wiih a IO-Sold induction of CD5 (Fig. 8B). Similarly, a significant proportion of' ihc 3T7.CD45 Cl. 1 infcciants (40%)responded to TCR ligaîion by downmodulating CD4 and
CD8 csprcssion (Fig. 8B), an effect not observed in the parental cell linc. Downrcgulation »I' Parental 3ïS 3i7.CD45 CL1
Figure 8: Effect of exogenous CD45 expression on TCR-rnediated signal transduction in 3T7 cclls.
(A) Esogcnously expressed CD45 rcstores TCR-induccd protcin tyrosine phosphoylaricin in 3T' cclls. Parcntal 3T7 cells or 3T7.CD45 C1.l cells werc subject to antibody-mediatcd cross-linkin; of thc indrcatcd surface inolecules as describcd in Fig. 2A. Arrows indicatc scvcral prolcins tha undcrgo TCR-induciblc tyrosine phosphoylation. Astcrisk (*) indicales phosphoprokin which i dil'fcrcntially phosphorylatcd in parental 3T7 cclls vcrsus 3T7.CD45 Cl. 1 cclls. Stimulation:
ioVo' IO' 10' 10' ioO 10' io'""io3 lo\oO 10' 10' 10' 10'
,$.
10" IO' 10' 10' 10' 1090' io2 10' 10' 10' IO' IO* IO' IO* -CDBa.-+ -- CD5 -b Figure SB: Esogcnously espressed CD45 rcsiorcs KR-induccd changcs in ccll sui-làcc phenotype.
Cells wese culluscd overnighl alone or with irn~nobilizedanli-TCR13 (1-157-597), and ~licri s~ainedwith FITC-conjugatcd antibodics specif'ic for: CD5 (53-7.3), CD4 (YTS-191.1), os CD8a (53-6.7). Shadcd liisrogsams rcprescnt staining wiih isotypc-malchcd conlrol anlibodics. co-rcccptor cspression also occurs wiicn trcshly isolated Dl-' tthytnocytcs arc sliniulalcci ovcrnight with anti-TCR (Grovcs et al., 1995; Kcarsc cl al., 1995; Groves cl al., 1997). Table
1 summarizcs thc levcl of TCR-induccd CD5 expression in several 3T7.CD45 infcctants. Whilc in al1 cascs, some rescue in TCR-induced signaling was observed, thc ability of CD45 to complemcnt the signaling defect varied considerably between clones. This may rclatc to thc additional signaling defect(s) present in 3T7 cells, alluded to before.
1 extended my functional analysis of the CD45 infectants to expression of RAG- 1. To assess whether CD45 infectants also regained their ability to downregulate RAG-1 mRNA in response to TCR ligation, Northern analysis was performed. Parental 3T7 cel!s did no1 downmodulatc RAG-1 message in responsc to TCR ligation alone (Fig. 8C). In contrast, 3T7.CD45 (21.1 cells responded to TCR Iigation with a 3-fold reduction in RAG-1 mcssagc.
However, this was no1 as çomplete as thal obsctved in parental 3T7 cells following TCRICD4 CO-stimulation,whcrc a 30-foId reduciion in the amount of RAG-1 message wüs observcd (Fig. 8C). This likely relates to the presence of a TCR-non-responsive subset of cells, as show in Fig. 4. Collectively, these data formally demonstrate that expression of exogenous CD45 in 3T7 cells rcstores TCR coupling Lo the proximal PTK signal ing pathway and to downslrcam maturation cvents at least in a subset of cells. These data dernonstrate a critical role for CD45 in rcgulating specific TCR-induced maturation events known to occur at the DP stagc of T cc11 dcvelopment.
H) Biochemical basis for rescue of TCR signaling by CD45
To identify thc sitc of action of CD45 in rcgulation of thc TCR signal transduclion pathway, 1 invcstigatcd thc phosphorylalion slatc of various known CD45 substrütcs. Iiilporlant targcts 01' CD45 arc thc Src-l'aniily kinases, Lck and Fyn (Ostcrgaard ct al., 1989; Muslclin ct al., 1993; McFarland cl al., 1993), as ivcll as TCRT, (Furukawa cl al., 1994). Thcrclbrc, I asscsscd thc inllucncc of CD45 cspression on Lhc tyrosinc phosphorylation of Lck, Fyn, and Parental 3T7 3T7.CD4 CL1 u u f! a Stimulation: k B 7.5 -
Figure 8C: Expression of exogenous CD45 restores TCR-induced RAG-1 downmodulation.
Northcrn analysis of RAG-1 and p-actin lranscripts in cclls cultured ovcrnighl alonc or with thc indicatcd immobilizcd antibadies was performcd as dcscribcd in Fig. 5A. Nuinbers on lcft indicatc migration of RNA MW standards. Dcnsitomctric analysis was perlormcd, and thc rcsults wcrc normalizcd for cach cc11 linc by triking thc ralio of RAG- 1 signal to p-actin signal and arbitrarily dcsignating this ratio in unstimulatcd cells as 1.0. Table 1. TCR-mediated CD5 induction in Parental 3T7 Cells versus CD45 Infectants
Mi3 is rcportcd as a mcasurc of surfacc CD45 cxprcssion on thc indicatcd 3T7 subcloncs, as dclected by staining with FITC-conjugated pan-CD45 antibody (ALI-4A2). Control stains wilh an isotype-matched antibody had an average MF1 of 3.0. To rncasurc thc various subclone ccll's funclional response to TCR cross-linking, they were cultured ovcrnight in wells coatcd wilh antihdics specific l'or: TCRp (H57-597) or TCRplCD4 hetcroconjugritc (H57-5971GK1.S), and rvcrc thcn staincd with FITC-conjugated CD5 (53-7.3) and analyzcd by llow cytomctry. CDS~~ cclls wei-c dcl'incd as those cells with a MF1 gseater than 100. Unstimiilated cclls coniaincd lcss ihan 1% ~~5~hlls. - --- . - - .------_.. _ _ _ _._ _ _ . . _ _ _ ------.. ------, .,- -. - - - -. . SDS-PAGE, and transl'crred to nitroccllulosc. The blots wcrc prcibcd f'irst witli unti- phosphotyrcisinc, and lhcn strippcd and re-probcd iising antibodies specific for the midcculc 01' inicrcst. To quantitate the degree of' phosphorylation, densitometric scanning of autoradiographs was performed and the results werc normalized by dividing thc phosphotyrosinc signal of a particular band by the amount of protein in that same band. This ratio was designated as 1.0 in parental 3T7 cells.
A sljght variation on the immunoprecipitation procedure was used 10 asscss thc phosphorylation status of TCRT. Mild lysis conditions (digitonin), which presewe TCRICD3 associations were used, and TCRq was immunoprecipitated indircctly by using antibody specific for TCRp. Thus, TCRp-associated 5 was measured, as oppscd to the total ccllular pool ol' TCRT. As can been seen in Fig. 9A, TCR-associated < chain is 3.5 timcs morc phosphorylated in parental 3T7 cells than in 3T7.CD45 CI. 1. Thus, an inverse relülionship csists between expression of CD45 and TCRS phosphorylation, suggesting that CD45 can regulate, either directly or indirectly, the phosphorylation state of TCRS. This is in agrccment with data from other groups who have found hypcrphosphorylation of TCRC in T-cell Iines deficient in CD45 (Volarevic et al., 1992; Niklinska et al., 19%). In contrast, no apprcciablc differencc was observed in the overall phosphorylation state of either Lck or Fyn in parental
3T7 cclls versus 3T7.CD45 CI. 1 (Fig. 9B, C).
1) Analysis of thymocytes from ~~45-1-mice
Whilc lhe results from the ce11 line syslem off'cr a uscful, inanipulablc tool 101-asking
questions iibout CD45 l'unction and rcgulation, il was impoi-lant 10 coi-1-clakand cossoborak ttic rcsulls in a niore physiologic setling. Thcrcforc, other members of thc laboratoi-y analyzcd TCR signaling in DP Ihymocytes [rom CD45 eson 6-1- niicc (Smilcy P., Grovcs T. and Guidos
C., unpublished rcsults), provided by Josel' Pcnningcr (Kishihara ct al., 1903). Thc i-csulrs
closcly parallclcd thosc rcporlcd in 3T7 cells, in that thyniocytcs froni C~4.5-/-inicc rcspciridcd I';ireril;~l 3T7 31'7.C'I)JS CI.] II' II' II' - Ip, - - "
Illi~ltinpAiilibiid? :
.Iti111 rr (p-tjrl -C *-.; .
înliscrï r e 4- Lch Lc k 45
Figure 9: Tyrosine phosphorylation index of potential CD45 substrates in 3T7 cells. Phosphorylation status of TCRC (Fig. 9A), Lck (Fig. 9B), and Fyn (Fig. 9C) in parental 3T7 cells and 3T7.CD45 C1.l. Cells were lysecl in LB, except in Fig. 9A in which cells were lysed in digitonin LB. Lysates from 5s106 cells were immunoprecipitated with the indicated antibody: anti-TCRp (H57-597),anti-Lck, or anti-Fyn, coupled to CNBr sepharose beads. Control IPs were performed identically using inactivated CNSr beads. Proteins were separated by 12.5% (Fig. 9A) or 10.5% (Fig. 9B, C) SDS-PAGE (non- reducing), and analyzed by sequential 4G10 and TCRS, Lck or Fyn immunoblotting. Densitometnc analysis was performed, and the rcsults wcre normalizcd by taking the ratio of p-tyr signal : TCRC, Lck, or Fyn signal and assigning the value from parental 3T7 cclls as 1 .o. I-----J - - U receptors were CO-engaged,both in tcrms of induction of protcin tyrosine phosphorylation and CD51CD69 induction (Smilcy, Grovcs and Guidos, unpublished rcsults). To asscss thc biochcmical basis of the impaircd TCR signaling obsenred in DP thymocytes from the ~~45-I- mouse, 1 analyzed the phosphorylation state of various CD45 substrates as described Sor 3T7 cclls. No differencesin the overall phosphoryIation state of Lck or Fyn werc obscn~ed(dala not shown). However, TCRp-associated TCR 5 was round to be 5 times less phosphorylated in ~~45-1-thymocytes when compared to ~~45+/-littermate controls (Fig. 10). Additionally, a protcin of approximately 40 kDa (*) was hyperphosphorylated in ~~45-1-thymocytcs whcn compared to cD45+jb littermate controls (Fig. 10). Thus, CD45 acts to regulaic lhc phosphorylation state of several phosphoproteins eithcr dircctly or indirectly. The profound hypophosphorylation of TCRI; in ~~45-1-thyrnocytcs is contrary to thc rcsults oblaincd in CD45-3T7 cells, where a hyperphosphorylation of TCRS was obscrvcd (Fig. 9A). The bwis of this discrepancy remains to bc detcrmined, but müy rcflect differcnccs belwecn studying
Lransformcd culturcd cells grown iri vilru, versus fresh ex vivo thymocytcs. Nonethclcss, ihc disrcgulation of TCRt phosphorylation that accompanies loss of CD45 provides a biochcmicnl corrclate with the TCR signaling defecis observed. Figure 10: Analysis of TCRS tyrosine phosphorylation in ~~45-/-thyniocytes.
Fresliy isolated thyniocytes froni ~~45-1-and CD&+/- littermatecontrols (4months of age) were lyed in digitonin LB and ir~~m~iiioprecipitated(7x 106 cells/lrim) as describcd in Fig. 9. Astcriiik (":) indiciites phosphoprotein whicli is difrercniiully phospl-iorylatecl iii CD4S-1- vetsus CD45+/-thyniocyies CHAPTER 4
DISCUSSION L thc well-documcntcd inübility to signal through the TCR in CD4S-deficient cell lincs, thc immature DPcell line, 3T7, described here is also TCR non-responsive. 1 have shown that the re-expression of endogenous CD45 or an exogenous CD45 construct largcly rcstorcs TCR signaling capabilities in 3T7 cells. CD45 was found to bc necessary not only for thc TCR- triggcred protein tyrosine phosphorylation, but also for several other TCR-induced downstrcarn events, including CD5 upregulation, RAG-1, and CD4lCD8 downregulation. This reprcscnts thc first demonstration of CD45 regulating these specific developmental events, known 10 occur during positive selection in vivo. Intriguingly, 3T7 cells could respond when TCR and co- rcceplor were deliberately co-aggregated, an effect presumably mediated by the co-rcceptar associated kinase, Lck. This CD45independent signaling pathway has also bccn demonstrütcd in
DP thymocytes [rom ~~45-1-micc (Smiley P., Groves T., and Guidos C., unpublished results).
Thus, 1 propose that the partial developmental block observed irr vivo in ~~45-1-rnice can bc rationalized by the existence of CD45-i ndependen t but CD4lCD8-dependent TCR signaling, allowing some DP thymocytes to be positively selected. Finally, 1 demonstrated that lack of CD45 in 3T7 cells results in hyperphosphorylation of TCRC chain. In constrast, the lack of CD45 in thyrnocytcs from the ~~45-/-micc resulls in a profound hypophosphorylation of TCRS chain. Thus, the disrcgulation of TCRC phosphorylation that accompanies the loss of CD45 providcs a biochcmical correlate of the profound TCR signaling deficits obscrved.
A) CD45-Dependent versus CD45-Independent TCR Signaling Pathways
This analysis of TCR signal transduction in a CD45-dcficicnt DP thymoma, 3T7, supports a critical rolc for CD45 in coupling the TCR to the intraccllulas signaling machincry.
Howcvcr, Lhc icquirement for CD45 is not absolute, in that both proximal (PTK activation, Fig.
3A) and distal (phcnotypic maturation, Fig. 2B; RAG-1 downrcgulation, Fig. 8C) outcomcs ol' TCR ligation can be induced by the co-aggrcgation of TCR with cithcr co-rcccptor, CD4 or CDS.
Similar findings havc bccn reportcd by two groups (Dcans ct al., 1992; Shiroo ct al., 1993). In onc study, a subclone 01 the human L)Y leukemic Cell Iine HYB.ALL was isolaled thal IücKcd surl'ace expression of CD45 (possessing a ~ranslationallpost-translationaldefeçt). In agrcemeni with my findings, thcy found that while CD3 ligation alone failed to inducc protcin tyrosine phosphorylation, calcium influs, or PLC-y 1 activation, CO-ligationof CD3 with CD4 was ablc lo induce thesc cvents (Deans et al., 1992). This was correlated with increased activation of CD4 associated Lck in CD45- cells (Deans et al., 1992). In another study, an indepcndently dcrivcd
CD45-negative (possessing a CD45 transcriptional/post-transcriptionaldefect) subclone of HPB.ALL was isolated and compared to a CD45RAB transfectant. Similar iïndings wcrc reported, in that CD3 ligation in the CD45 cells failed to induce protein tyrosine phosphorylation, inositol phosphate production, calcium influx and PKC activation, while CO-aggregationof CD3 with CD4 or CD8 restored these events (Shiroo et al., 1992). The lack of ability to signal when TCR was ligatcd alone was correlated with lower basal Fyn kinase activities in CD4S- cclls, whereas Lck kinase activies were comparable in CD45 versus CD45f cells (Shiroo et al., 1997). Howcver in one report this signaling phenotype was not observed (Biffen cl al., 1994).
Invcstigators found that CD45-cclls (isolated from a human DP T-ce11 linc callcd CB 1) lailcd tri inducc protcin tyrosinc phosphorylation and calcium influx in rcsponse to CD3 as wcll as CD3 s CD41CD8 stimulation (Biffen et al., 1994). The reason for this contradictory result remains undefined, but may relate to differences in the activationldifferentiative state of the various T ccll lincs utilizcd.
A potential cavent of the restored TCR signaling observed when TCR and co-reccptors wcre CO-ligatedis the ability of 3T7 cells to inducibly re-express endogenous CD45 (Fig. 5A, B).
II coiild bc argucd that the rescue in TCR signaling observed was simply due to the rc-esprcssion of CD45 Two lines of cvidcnce however argue against this. First, thc CD45 rc-cxpi'cssion as dctcçtcd by FACS occurs after 24-48 hours. Yct, the ability to inducc an carly responsc
(phosphotyrosinc induction) is dctectcd alkr 1 minute of TCRICD4 aggregation. Thus, il would bc unlikely that CD45 could bc signiricanlly re-cxpressed. Second, sludics of CD45-'- thymocytcs, which do no1 inducibly rc-cxprcss CD45 (Smilcy P., Grovcs T., and Guidos C., unpublishcd results), show similar funclional rcsponscs lo TCRIco-rcccptor stimulation . Thus, L' L' ------". - - - - .. ------, -- -~-~ - .- ..-- - - -.-r--" . . -- .- 7 , U - ~---~ - -.- rathcr is truly CD45-independent. If this is lhe casc, it suggests that the CO-receptorassnciatcd Lck can be activated in thc abscnce of CD45, whilc thc TCR-associated F'TKs (ie. Fyn) can not. Notably, whcn cithcr CO-rcceptorwas ligated without TCR, liltlc or no signaling was obscrvcd
(Fig. ?A, B). This suggests that elements in the TCR comples, namely the CD3 and TCRC chains, are rcquired for the entire signal transduction pathway to occur, perhaps by providing a physical structure ont0 which other molecules can be recruited, such as Zap-70. Several potcntial mechanisms could explain why TCR signaling absolutely requircs CD45 while TCRko-receptor signaling does not. One possible esplanation is ihat co-receptor associated Lck can be activated without C-terminal tyrosine dephosphorylation, in effcct bypassing the requirement for CD45 It is interesthg to note that Moarefi et al. reccntly dcmonstratcd that addition of a SH3 ligand stimulated the activity of purified Hck (a Src-rclated PTK) that is phosphorylated at Tyr 527 (Moarefi et al., 1997). These results suggest that whilc dephosphorylation of C-terminal tyrosine may be a key regulatory switch in Src-làmily kinasc activation, cornpetition for binding of Src-family kinase SH3 domains by cxogcnous ligands may also result in kinasc activation by releasing inhibitory intramolecular conformations (sec later).
Thus, it is possiblc thal C-terminal phosphorylated Lck andfor Fyn in 3T7 cells are nonetheless aclivated when TCR and CO-receptorare CO-aggregatedduc to thc presence of other proteins whicli rclicvc thcir inhibitory conformation. Onc molecule that may act in this capacity is Syk, which has been shown to bc constitutively bound to the TCRICD3 complcx and bccome activatcd in a Lck-independent and CD45-indcpendent mannes (Couture el al., 1994; Pao and Cambicr,
1997), although this remains contentious. It has bcen propsed ihat tyrosine phosphoiylatcd Syk may mcdiate interactions with Lck SH2 domains and lhus recruit Lck io ihc TCRICD3 complcs
(Thomc cl al., 1995). Thus, it is possible that Syk can become autophosphorylatcd in 3T7 cclls, allowing it to compcte for Lck SHZbinding when TCRlco-reccptor are co-aggrcgatcd. This would rclease Lck Srom its inlramolccular inhibition and allow efficicnt signal amplificalion 10 occur. Intercstingly, Syk is cxpsesscd highly in immaturc DP thymocytcs bcforc bcing down- L , . W" .. cclls. Thc diffcrcntial requirement for CD45 in TCR CO-reccptorindcpendcnt signaling vcrsus TCWco-receptor signaling could then be explained by quantitative differenccs in the signaling capacity of thc two stimulatory conditions. Only 1-3% of cellular Fyn is thoughl to bc associalcd with TCRICD3 compIes (Samelson et al., 1990), while greater than 50% of cclluiar Lck associates with the CO-receptors,CD4 and CD8, in DP thymocytes (Wiest et al., 1993). Perhaps
CD45 is a prerequisite when TCR is ligated alone, needed for the efficient initial activation of srnall amounts of Fyn. In contrast, when TCR and CO-rcccptorare CO-ligated,CD45 becomes dispcnsablc, owing 10 the higher arnaunts of Lck recruited to the signaling cornplex and becoming activated by the mechanism proposed above. The observation that TCRICD4 co-aggregation gave consislcntly stronger signals than TCRICD8 CO-aggrcgation(Fig. ?A, B), an effect that direclly correlated with the greater association of Lck with CD4 than CD8 in 3T7 cells (Fig. 3A), lcnds support to this quantitative argument.
A second explanation for the differential requirement for CD45 in CO-receptor-dependent ircrsus co-rcceptor-independent TCR signaling could involve qualitative diffcrcnccs in ~hc aclivation requirements of CO-receptorassociated PTKs versus TCR-associatcd PTKs. For cxamplc, thc CO-receplorassociated PTK activity cauld bc "prc-activated" perhaps by being inücccssiblc to Csk, which itself is predarninantly localized in thc cytoplasm (Nada et al., 1991; Okada et al., 1991; Bergman et al., 1992). In this scenario, the CO-receptor-boundLck would not be cngaged in an intramolccular C-tcrminal phosphotyrosinelSH2 inkraction bccausc thc C- Lcrminal tyrosine is not phosphorylated, and could thus be aclivatcd upon TCRICD4 co- aggrcgation. Intriguingly, a inuhnt rncinbrane-targeted form of' Csk more slrongly inhi bilcd TCR signaling than the WT cytoplasmic form (Chow et al., 1993), suggcsling that thc ccllular localimtions of thc molcculcs can have a critical influcncc on lhcir activiiy. Also, 1 can not rulc out thc possibility that olhcr unidenlilïed PTKs, not rcquiring CD45 for their activation, could bccomc aclivatcd whcn TCiUCD4 arc co-aggrcgatcd. - --.-.--,, -~ -- ,-~---- - lcast one study, thc dcfect in TCR signaling caused by lack 01' CD45 \vas partially conipcnsatcd by csprcssing a hetcrologous PTPase from yeast (Mntto et al.. 1994). suggesting that othcr PTPascs can indced perform somc of CD45 functions. With the identification of ovcr 75 diffcrcnt receptor and cytoplasmic PTPases (Tonks and Neel, 1996), many of which arc cspressed in T cells, the possibility of overlapping funclion becomcs more and morc likcly. In support of this contention, treatment of 3T7 cells wi th the FïPase inhibitor penlanadate lcads to an accumulation of tyrosine phosphoproteins and CD5 upregulation in a dose-dependcnt Sashion (Fig. 4A, B). This indicates that other PTPases are expressed in 3T7 cclls, acting to maintain thc ovcrall phosphotyrosine homcostasis in the cell. When this balance is perturbcd, TCR signal transduction pathways can become activated without dclibente aggregalion of lhc TCR.
B) Molecular Targets of CD45 in TCR Signal Transduction
In this study, no significant differences were observed in the overall tyrosinc phosphorylation statc of Lck or Fyn in 3T7 cells compared to thc 3T7.CD45 inkctants (Fig. 9B,
C). These were unexpected findings, considering several studies have Sound a corrclation bctwcen a lack of CD45 expression and a hyperphosphorylation ol' thc C-terminal tyrosinc of total ccllular LcklFyn (Ostergaard et al., 1989; McFarland ct al., 1993; Sich cl al., 1993). Thc rcsul tant dccrcase in kinase activi ty providcd an allraclivc mcchanism for thc impaircd TCR signaling obscrvcd in CD45-deficient cells. However, in my study neither LckIFyn kinasc activily nor their C-terminal phosphorylalion shlus wcrc assessed. Givcn that Src-l'amily kinascs conlain thrce potcntial rcgulatory tyrosine sites, it is entircly possible that whilc thc ovcrall phosphoiylation slatc of Lck and Fyn may not bc diffcrcnt in lhc prcscncc or absencc of CD45, thc sites of phosphorylation may be diffcrcnt, cither qualitativcly or quantitativcly. Thus, a disrcgulation of LcklFyn may still exist in 3T7 cells. Thereforc, thc asscssmcnt of LcklFyn kinasc activitics in 3T7 cclls vcrsus 3T7.CD45 cells will clarify this mattcr. An altcrnatc hypothcsis is Lhat CD45 may no1 bc invalvcd in rcgulating thc basal LcUFyn kinasc activitics, but ra~ncrmay cscrr ils Iuncuon concurrcnr. w1r.n or jusr ai Lcr r LK iigarion. 1 t may wcii oc inc Daiance of inhibitory vcrsus stimulatory activities of CD45 that determincs its net cffcct during TCR signaling. Nonethelcss, thc results prcsentcd here clearly indicatc that the funciion OS CD45 in
3T7 cclls is more complex: than simply kecping Lck/Fyn in an "on" conf~guration. This is dcmonstratcd by the finding that TCRlco-receptor cmsignal independent of CD45, impl ying that thc activation of CO-receptorLck may occur indepndently of C-terminal dephosphorylation.
Thc comples rolc of CD45 in regulating Lck/Fyn is dernonstrated by scvcral studics whose results wcre incompafible wi th the mode1 of CD45 aclivating LcklFyn simply by dcphosphorylation of C-terminal tyrosine. Two studies reportcd that thc kinase activity of LcWFyn \vas higher in CD45 cells than in CD45+ cells, despite the hyperphosphorylation of thc C-terminal tyrosinc (Dcans et al., 1992; Burns et al., 1994). Also, some investigators have suggested a negative regdatory role Tor CD45 For example, the CO-ligationof CD45 with the TCR cornplex suppressed T ce11 activation (Turka et al., 1997), while the co-ligation of CD45 with CD4 inhibi ted the anti-CD4-induced phosphorylation of Lck and the concomitant incrcasc in Lck kinase activi ty (Ostergaard and Trowbridge, 1990). These antibody-mediated ncgatik~c cllCcts may bc duc to the inappropriatc antibody-rncdiatcd coaggrcgation of CD45 with thc TCRICD3 complcx, lcadi ng to an increased PTPase activi ty which cCfcctivcly prcvcn ts thc accumulation of tyrosine phosphoproteins. Alternatively, ihc inappropriatc dcphosphorylation of
LcHFyn auto-phosphorylation site, which is rcquired for activation, could occur.
T hc considerablc con troversy regardi ng the ef f ec ts of CD45 dc fiçicncy on LçklFy n rcgulation may bc cxplained by the complex activation mcchanisms of Src-family kinases. Thc rcccnt resolution of the crystal struciurcs of c-Src and another Src-family PTK, Hck, havc providcd insights into thc activation requirements of Src-family PTKs (Siclicri ct al., 1997; Xu ct al., 1997). In thesc studics, both the SH2 and SH3 domains arc involvcd in intrimolccular iiitcractions rcsulting in conformational constraints on thc kinase active site (Sichcri ct al., 1997;
Xu et al., 1997). Moiarcfi et nl. rccently proposcd a hypothetical modcl of Hck activation i n which SH3 doinain displacement, SH3 domain displaccmcnt, autophosphorylation and C-
~erininaltyrosinc dcphosphorylation may al1 activatc Hck io difl'cring dcgrccs (MoarclÏ ct al., ., , . ,. ' """,...- . '...--....i "' "'- '-".."., '., D-.- "" "'., "' " .- ". t! ", ", """ rclalc to thc complcment of othcr molcculcs wiLh SH2 and SH3 ligands (incluciing other PTPascs, PTKs and "adapter" moleciilcs) c~pressedin the varioiis T cc11 lines, which mriy var); clcpcnding on the particular developmental and differentiative state of the ccll being studied.
In contrat to the lack of difference in the overall tyrosine phosphorylation of LckIFyn, 1 obscrvcd a hyperphosphorylation of TCRS in 3T7 cells (Fig. 9A). Similar Sindings wcrc rcportcd by lrvo other groups studying CD45-deficient ce11 lines (Volarevic ct al., 1993; Niklinska ct al., 1994). The hyperphosphorylation of TCRC is somewhat surprising considering that Lck and Fyn, which are thoughl to be rcsponsible for TCRI; phosphorylation, would be cspectcd to bc inactivc in 3T7 cclls. Hencc, TCRS should bc largcly unphosphorylatcd in the absencc OS CD45
In Sacl lhis is prcciscly what was observed in thymocytes from CD49 mice, in which TCRC was hypophosphosphorylatcd (Fig. 10). Several potentiül cxplanations cxist for this parados. Thc phosphorylation state of TCRg depends on multiple factors, a number of which may difl'cr bctween 3T7 cells versus fresh ex vivo thymocytes. First, the constitutive CD4 engagement on DP thymocytes by MHC II on thymic stroma in situ has been shown tu result in thc hypcr- phosphorylation of TCRI; (Nakayama et al., 1990). 3T7 cells cul turedin suspension in thc abscnce of MHC II+ cells obviously lack this interaction. Second, 3T7 cells king a transforincd cc11 line, may possess other defects which could affect TCRI; phosphorylation. For esamplc, many transformed ce11 lines possess high PTK activity, and this could resull in a hypcrphosphorylation of TCRI;. The decrease in TCRC phosphorylation thal accompanics CD45 rc-expression in 3T7 cells could thcn bc explained as a dircct cffccl of CD45 which has bccn shown to dephosphorylatc TCRC (Furukawa ct al., 1994).
C) Role of CD45 in Positive Selection of DP Thyniocytes
Thc data prcscntcd in this thcsis support a critical rolc for CD45 in rcgulaling TCR sigiials rit thc DP stagc of dcvclopmcnt. The importance of this molcculc is cvidcnt in thc dcvclopmcnlal phcnolypc of ~~45-1-micc, in ivhich thymocytc dcvclopmcnl is scvcrely impaircd in thc DP to 3r LI~IISILICIII(h1srlinürü el ai., 1 rr3; ~y ln ci ai., IYY~). r el, givcn Lnc cviacncc cnar ~~43- indcpcndcnt TCR signaling can occur when TCR and co-rcceptor arc co-ligatcd, thc scvcrc devclopmcntal block is sornewhat surprising. Physiologically, it has bcen proposcd lhat mosl T cells are activated only when TCR and CO-receptorare brought in10 close prosimity via recognition of the same MHC molecule (Weiss and Littman, 1994). If this TCRIco-rcccptor signaling is CD45-independent, then one might expect that most DP thymocytes could gcneraic the necessary signals in the absence of CD45 to be positively selected. Howcver, thc TCR signals involved in positive selection are likely more subtle than the artil'ical stimulatory conditions used in signaling assays, where antibodies are used to crosslink surface moleculcs in saturating amounts. Thus, while many of the outcomcs of TCR ligation can be induced in the absence of CD45 iti vitro by driving TCR signals using strong stimuli, ii~vivo such ovcrwhclming stimulatory conditions are unlikely to exist. In fact, cven in the presence of ihesc artifical stimulatory conditions, most, but not al1 DP thymocytes from ~~45-1-mice induced CD5 upon TCRIco-receptor crosslinking (Smiley P., Groves T., and Guidos C., unpublishcd rcsults), suggesting that some cells are unable to bypass the requirement for CD45 Similarly, not al1 3T7 cells undergo CD5 induction Iollowing TCRlco-receptor crosslinking (Fig. 2B). I t semains unclcar whether this reflects secondary defects in a subset of cells or a modulation in the signaling characteristics of somc cclls, rendering them non-responsive. In 3T7 cells, 1 consistenily found the CD4 CO-receptorpossessed a grcatcr ability to signal
than thc CD8 CO-rcceplorwhen CO-aggregatedwith TCR (Fig. 2A, B, 5A,B). Thc diffcrcntial
signaling capacity of the two CO-reccplorscorrelated with thcir diffcrcntial association wilh Lck
(Fig. 3A). This resull prcdicls thai CD4+ T cells should bc morc cfficicntly positivcly sclcçicd
than CD8+ 'ï cclls in CD45-1- mice, owing to the strongcr signals dclivcrcd by thc CD4 co- reccptor. This is indeed the case, as the ralio of CD4' to CD8f T cclls in the periphcry of CD45
cson A-/- micc was skewed forn 3: I in WT animaIs to 4: 1 in CD45 cxon 6-1- rnicc (Kishihara ci
al., 19931, suggesting a more profound impairment in ihc dcvclopmcnt of thc CD8 lincagc ihan CD4 lincagc in thc absencc of CD45. In contrast, CD45 cxon 9'-micc appcaicd to have a nioic
scvcre impairn~entin thc devclopment of CD4+ than CDS+ SP T cclls, both in thc thymus and in ~ ~ ------~ C~l~- 1 1 d\, , ~ ~, -, - C --- -- the CD8+ SP thyrnocytes cells were clearly CD~~~and thus likely immature. It is intcresting lo notc that in both CD45 eson 6-1- and exon 9-1- knockout mice, there is ri marked iiprquloticm of
CD4 and CD8 CO-receptormolecules on DP thymocytes (Kishihara et al., 1993; Bytli cl al., 1996). Considcring iny observation that TCRko-receptor signaling is largely CD45-indepcndcnt, i t could be inferred that upregulation of CD4/CD8 expression on DP thymocytes increases thc ability of DP thymocytes to generate TCR signals. Furthemore, CD5, a negative regulator of TCR signals (Tarakhovsky et al., 1995), was shown to be downregulüted on DP thymocylcs l'rom CD45 exon 6-/- rnice (Smiley P., Groves T., and Guidos C., unpublished results).
Collectively, the downregulation of CD5 and upregulation of CO-receptorsmay thus rcprcsent adaptive responses made to overcome the CD45 deficiency, increasing the likelihood that DP thymocytes will be positive1y selected.
The lack of expression of CD45 in 3T7 cells represents an extremc cxamplc of the consequences of CD45 expression on TCR signaling in DP thymocytes. Howcver, under physiologie circumstances more subtle adaptations are likely employed to modulüte thc signaling charactcrislics of thc ccll. Mcchanisticall y, this could involvc changcs in thc surfacc cxprcssion levcl of CD45 The recent finding that CD45 is upregulatcd during positivc seleciion concomitantly with the TCRICD3 complex supports the idea that the thymocyte adapts CD45 lcvels to incrcasing antigcn receptor levels during development (Ong et al., 1994; Kirbcrg and Brockcr, 1996). This increüsc in CD45 expression would be espectcd to lowcr thc signaling thrcshold in thc maturing thymocyte, making it more competent to generate TCR signals.
Altcrnativcly, changcs in CD45 isoform expression could altcr the signaling characteristics 01' the cc11 (Novak ct al., 1994), possibl y by intcracting wilh distinct ligands andlor substratcs (Dimmni ct al., 1992; Leitenberg et al ., 1996). Intcrcstingly, studies of TCR transgcnic micc on sclccling backgrounds showcd that CD45RA and ~~45RBt"d'isofornls specifically werc uprcgulatcd during positivc and ncgalive selection (Wallace et al,, 1992), suggesting that isolOrm patterns do indccd change during thymic sclection cvents. Furthcrmorc, thymocytcs fi-on1 CD45R1113C lrünsgenic micc rcspondcd morc robustly toTCR stimulalion ihan did thymocytcs froni CD45RO """"b'""' """" ,""' -. W.' " " 9,' * a'-"- .-"-"" "- ""' -.. ' -.-"' ..," "'...." ," ,& "'.+,, difl'cr in Lhcir abilitics Lo participalc in TCR signaling, and thus allcr thc signaling chanictcrislics of dcveloping thymocy tes. ln summary, this study has estended our undcrstanding of thc rolc of CD45 in TCR signaling in DP thymocytes. My results suggest that TCR signals can bc gcncratcd in a CD45 independent manner, providing a rational basis for understanding the devclopmentai phcnotype 01'
CD45-'- mice.
D) Future Studies
Tlic results described herc indicate that TCRICD4 and TCRICD8 signaling in DP thymocyics is largcly CD45independent. 1 hypothesize that this signaling is inilatcd via co- rcccptor associatcd Lck. To formally dcmonstrate that Lhis is the case, the following expcrimcnt could be performed. Using the CD4 lineage as an example, I would gcncratc CD~~-/-;CM-;- double-deficient mice and then reconstitute them with a tmncated form of CD4 that can no longcr associate with Lck. If the ability to generate CD4f SP in CD4.5-1- micc is via the postulatcd mcchanism, then in CD4w~t&CD45-~-;CD4-I- mice, no CD4+ SP should be observcd, as thc ncccssary signal can ncit bc transduccd.
Toaltempt 10 further dctïnc signaling pathways, lhe cffccts of ovcrcsprcssicin 01' acli\'atcd forms of molcculcs lhought 10 bc rcgulatcd by CD45 could bc assesscd. Thc complcmcntalion (il'
CD45 defccts by such gain-of-function molecules would provide gcnclic evidcncc of a signaling pathway. Such an approach might involvc csprcssing LckF505 andlos FynT528 transgcncs on thc ~~45-1-background, Lo scc if thymocytc dcvclopmcnt rcvcrts to wild-typc. This approach has been attcinptcd In the 3T7 cclls, whcrc LckF505 or FynT528 wcrc ovcrcxprcsscd, and assesscd for a rescuc in TCR signaling. Whilc a partial rcstoration of TCR function was somclimcs observed in FynT528 ovcrespressors, the resulls wcrc no1 rcproduciblc (data not shown), suggcsting that othcr defects cxist in 3T7 cclls or that CD45 dcficicncy can not bc Sully complcincntcd by constitulivcly activc Fyn. To rcsalvc somc OS Lhc contradiclory rcsults ivith --Q-..-.------.--r--.-,--.-. - - - , - . .. -.. ------.------.. -. - - - - .-. -. . - - - - . .. . .-J --.- - tcrminal tyrosine phosphorylation status of Lck and Fyn in 3T7 cclls versus CD45 inlcctants.
Ah,atternpts could be made to mimic itt sitic effects by ligaiinç CD4 on 3T7 cclls irr vitro, and asscssing if' this restored the TCRS phosphorylation patterns observed in thymocytcs from CD45-
1- micc. Finally, thc expression and activity of Syk in 3T7 cells should be assessed as a first stcp in determining if this PTK could be responsible for the CD45-independent signaling obscrvcd. The defect in CD45 gcne expression in 3T7 ceIIs appears to bc ai the levcl of transcription. This could involve a defcct in trnns, in which resting 3T7 cells lack expression of a positive- acting transcription factor, or the defect couId be in cis, in which mutation(s) in thc regulatory sequcnces of the CD45 locus make it non-transcribablc. Allcrnritivcly, svabilizing factorslsplicing factors neccssary to maintain steady statc message of CD45 could be missing in resting 3T7 cclls, but induced in stimulated cells. The inducible re-expression of CD45 in 3T7 cells suggcsts that differences exist in CD45 expression betwccn resling stale cclls and activated cells. Somc possible future experiments to address the defect in CD45 expression in 3T7 cells might includc ceIl-cell fusion procedures, to see if the defect in CD45 espression can be cornplemcntcd. Thc sclcction of propcr fusion partncrs is critical; idcally it would bc a cc11 which csprcsscs al1 thc transcription factors necessary for CD45 expression, but itself has a disrupted CD45 gcne. If aftcr such a fusion, CD45 was re-expressed, then this would indicatc a dcfect in a transcriptional factor in 3T7 cells In summary, the cxperimcnts outlined in this section would estcnd our undcrstanding of how CD45 acts 10 regulate TCR signal tranduction, and furtlicr our understanding of how TCR signals control T cc11 dcvclopmcnt. CHAPTER 5
REFERENCES Amnam, N., MKCII,NI. L., rames, J. K., ana veiiieue, A. ( IYY 1). mnanccmcnl or i ccri rcsponsiveness by the lymphocyte-spccific tyrosine protein kinase, p56icP.Nalure 350, 62-66.
Albernla-Ha, J., Forbush, K. A., Seger, R., Krebs, E. G., and Perlmutter, R. M. (1995). Selective requirement for MAP kinase activation in thymocyte differentiation. Nature 373, 630- 633.
Alberti, S., Parks, D. R., and Hcrzenberg, L. A. (1987). A single laser mcthod Sor subtraction of ce11 autofluoresccnce in flow cytometry. Cytometry 8, 1 14-1 19.
Anderson, G., Anderson, K. L., Conroy, L. A., Hallam, T. J., Moore, N. C., Owen, J. J. T., and Jenkinson, E. J. ( 1995). lntracellular signaling events during positive and negati ve sel ection of CD4CCD8Cthymocytes in vitro. J. Immunol. 154, 3636-3643.
Appleby, M. W., Gross, J. A., Cooke, M. P., Levin, S. D., Qian, X., and Perlrnutter, R. M. ( 1992).Defectivc T ce11 receptor signaling in mice lacking the thymic isoforrn of p56Y'2.Cell 70, 75 1-763.
Arpaia, E., Shahar, M., Dadi, H., Cohen, A., and Roifrnan, C. ( 19%). Defective T ce11 rcceptor signaling and CD8+ thymic seleciion in humans lacking ZAP-70 kinase. Ce11 76, 947-958.
Bell, G., Dethlofî, G., and Imboden, J. (1993). CD45-negative mutants of a rat natural killcr ccll line rail to lyse tumor target cells. J. Immunol. 151, 3646-3653.
Benveniste, P., Takahama, Y., Wiest, D. L., Nakayama, T., Sharrow, S. O., and Singcr, A. ( 1994). Engagcrncnt of the external domains of CD45 tyrosine phosphatasc can rcgulaîc ihc di fferentiation of immature CD4+CD8+ thymocytes into mature T ce1 1s. Proc. Natl. Acad. Sci. USA 91, 6933-6937.
Bergman, M., Mustelin, T., Oetken, C., Partanen, J., Flint, N. A., Amrien, K. E., Autero, M., Burn, P., and Alitalo, K. (1992). The human p~~CS%tyrosinekinasc phosphorylatcs p561"kat tyr-505 and dow n-rcgulates its catal ytic activi ty. EMBO J. 11, 29 19-2924.
Bevan, J. ( 1977). in a radiation chimacra, host H-2 an tigcns dctcrminc immunc rcsponsivcncss of donor cytotoxic cclls. Naturc 269, 417-419.
BiSfcn, M., McMichael-Phillips, D., Larson, T., Venkitaraman, A., and Alexander, D. (1994). Thc CD45 tyrosine phosphatase regulates specific pools of antigen receptor-associalcd p59Lv)1 and CDCassociated p56kk tyrosine kinasc in hurnan T-cells. EMBO 13, 1920- 1929.
Bjorkman, P., Sapcr, M., Samradi, B., Bennett, W., Strominger, J., aiid Wiley, D. (1987).Thc forcign antigen binding si te and T ceIl recognition regions of class 1 histocompatibility antigcns. Naturc 329, 5 12-518. ULIIIIICLIU, IVl., IVILIIVUII) .Ji, ULLU dYIIYLI) 111. \ A// l ,. A &.,"IV.+. U.7"""..ICI\I,* ,II VY , U.." VY Id LI, primary, rcsling CD4+ T cells. Ccll. Immunol. 175, 1-11.
Brandlc, D., Müllcr, C., Rülicke. T., Hengartner, H., and Pirchcr, H. (1993). Engagcmcnt of thc T-cc11 rcccptor during positive selection in the thymus down-rcgulates RAG-1 cspression. Proc. Natl. Acad. Sci. USA 89, 9539-9533.
Burns, C., Sakaguchi, K., Appclla, E., and Ashwell, J. (1994). CD45 rcgulation of tyrosinc phosphorylation and enzyme activity of src family kinases. J. Biol. Chem. 269, 13594-13600.
Byth, K., Conroy, L., Howlctt, S., Smith, A., May, J., Alexander, D., and Holmcs, N. ( 1996). CD45-nul1 lransgenic mice reveal apositive regulatory role for CD45 in early lhymocytc dcvclopment, in the selection of CD4+CD8+thymocytes, and in B cell maturation. J.Esp.Mcd. 183, 1707-1718.
Cantrell, D. (1996).T cell antigen receptor signal transduction pathways. Ann. Rev. Immunol. 14, 259-274.
Chan, A. C., Iwashima, M., Turck, C. W., and Weiss, A. (1992). ZAP-70: A 70 kd protein- tyrosine kinase that associates with the TCR 5 chain. Cell 71, 649-662.
Chan, A. C., Desai, D. M., and Weiss, A. (1994a). The rolc of protein tyrosine kinascs and protein tyrosine phosphatascs in T ce11 antigen receptor signal transduction. Ann. Rev. Immunol. 12, 555-592.
Chan, A. C., van Oers, N. S. C., Tran, A., Turka, L., Law, C.-L., Ryan, J. C., Clark, E. C., and Weiss, A. (1994b).Differcntial expression of ZAP-70 and Syk protein tyrosinc kinascs, and the rolc of this family of protein tyrosine kinases in T cell antigen receptor signaling. J. Immunol. 152,4758-4766.
Chan, A., Kadlecck, T., Eldcr, M., Filipovich, A., Kuo-W.-L., Iwashima, M., Parslow, M., and Weiss, A. (1994~).ZAP-70 deficiency in an autosomal rccessivc form of scverc combincd immunodeficicncy.Scicnce 264, 1599-1601.
Chan, A. C., Dallon, M., Johnson, R., Kong, G.-H., Wang, T., Thoma, R., and Kurosaki, 7'. (1995).Aclivation of ZAP-70 kinase activity by phosphorylation of tyrosinc 493 is rcquircd 101. lymphocytes anligcn rcceptor funciion. EMBO J. 14, 2499-2508.
Chang, H.-L., Lcfrincois, L., Ziroukian, M. H., and Essclinan, W. J. (1991). Dc\!clopn~cntal cxprcssion of CD45 alternale csons in murine T cells. J. Immunol. 147, l687-16%.
Charbonneau, H., Tonks, N. K., Kumar, S., DiIli, C. D., Harrylock, M., Cool, D. E., Krebs, E. G., Fishcr, E. H., and Walsh, K. A. (1989). Human placcnta protcin-tyrosine-phosphatasc: amino acid scqucncc and relaiionshi p to a family or rcccptor-l i ke protci ns. Proc. Natl. Acad. Sci. USA 86, 522-5256. Cheng, A. M.,Rowlcy, B., Pao, W., Hayday, A., Bolen, J. B., and Pawson, T. (1995). Syk tyrosinc kinasc rcquircd Sor mousc viabilily and B-ce11 developmcnl. Naturc 378, 303 -306.
Chow, L. M. L., Fourncl, M., Davidson, D., and Veillette, A. (1993). Ncgative rcgulalion of'T- ccll rcceptor signal ling by tyrosine protein kinüsc p50Cs! Naturc 365. 156-160.
Chu, D., Spits, H., Peyron, J., Rowley, R., Bolcn, J., and Weiss, A. (1996). The syk prolcin tyrosine kinasc can function independently of CD45 or lck in T ce11 antigen rcccptor signaling. EMBO 15, 6251 -6261.
Chui, D., Ong, C., Johnson, P., Teh, H., and Marth, J. (1994). Specil'ic CD45 isoforms di fferentially regulate T cell receptor signaling. EMBO 13, 798-807.
Cobbold, S., Jayasuriya, A., Nash, A., Prospero, T., and Waldmann, H. (1984). Thcrapy with monoclonal antibodies by elimination of T-cell subsets in vivo. Nature 312, 546-551.
Coffman, R., and Wcissman, 1. (1981). B220: a B cell-spccific mcmber of lhc T300 glycoprotcin farnily. Naturc 289, 681-683.
Cooke, M. P., Abraham, K. M., Forbush, K. A., and Perlmutter, R. M. (1991). Regulation of T ce11 receptor signaling by asrc farnily prolcin-tyrosine kinase (p5@). Cell 65, 281-291.
Cooper, J. A., and Howell, B. (1993). The when and how of src regulation. Cell 73, 1051- 1054.
Couture, C., Baier, G., Altman, A., and Musielin, T. (1994). -independent activation and tyrosine phosphorylalion of p7~ykby T-ceIl antigen receptorlCD3 stimulation. Proc. Nall. Acad. Sei. USA 91, 5301-5305.
Cystcr, J., Hcaly, J., Kishihara, K., Mak, T., Thomas, M., and Goodnow, C. (1996). Rcgulation of B-lyrnphocytc ncgativc and positive sclcction by Lyrosinc phosphatasc CD45 Nalurc 381, 325-328.
D'Oro, U., Sakaguchi, K., Appella, E., and Ashwell, J. (1996). Mutational analysis cif Lck in CU45-ncgalivc T cclls: Dominant rolc of tyrosinc 394 phosphorylation in kinasc acii\ity. Mol. Ccll. Biol. 16, 4996-5003.
Dcans, J. P., Kanncr, S. B., Torres, R. M., and Ledbcttcr, J. A. (1992). Intcraclion 01' CD4:lck with thc T cc11 rcccptorlCD3 complcx induces early signaling evcnts in Lhc absence of' CD45 tyrosinc phosphaiasc. Eur. J. Immunol, 22, 661-668.
Dcsai, D. M., Sap, J., Schlcssinger, J., and Weiss, A. (1993). Ligand-mcdiatcd ncglilivc rcgulalion of a chimcnc transmcmbrane receptor tyrosinc phosphatasc. Ccll 73, 541 -5%. ULùCIL, y.,JLLp, Ji, J11VG1111~J111G11,W., rJLllIL33III~b1,J., CLiIU VVbl>a, fi. \177Y). 111L LcllrllyLIL activity ol'thc CD45 mcmbrünc-prosimal phosphatase domüin is rcquircd for TCR signalling and rcgulütion. EMBO 13,4003-4010.
Dialynas, D. P., Wildc, D. B., Marrack, P., Pierres, A., Wall, K. A., Hanan, W., Otlcn, G., Lokcn, M. R., Pierres, M., Kappler, J., and Fitch, F. W. (1983). Characteriration of thc murinc antigenic determinant, designated L3T4a, recognized by monoclonal antibody GK1.5: csprcssion of UT& by functional T cell clones appears to corrclatc primarily with class II MHC anligcn- rcrictivity. Irnmunol. Rev. 74, 29-56.
Dianzani, U., Redoglia, V., Malavasi, F., Bragardo, M., Pileri, A., Janeway, C., and Bottomly, K. ( 1993). Isoform-specific associations of CD45 wi th accessory molecules in human T lymphocytes. Eur. J. Immunol. 22, 365-371.
Doyle, C., and Strominger, J. (1987). Interaction bclween CD4 and class II MHC molcculcs mcdiates ceIl adhcsion. Nature 330, 356-259.
Egerlon, M., Scollay, R., and Shortman, K. (1990). Thc kinetics of mature T cell devclopment in the thymus. Proc. Natl. Acad. Sci. USA 87, 2579-2583.
Elder, M., Lin, D., Clever, J., Chan, A., Hope, T., Weiss, A., and Parslow, T. (1994). Human severe combined immunodeficiency due to a defect in ZAP-70, a T cell tyrosine kinase. Scicncc 264, 15%- 1599.
Ettchadieh, E., Sanghera, J. S., Pelech, S. L., Hess-Bienz, D., Waits, J., Shastri, N., and Acbersold, R. (1992). Tyrosinc phosphorylation and activation of MAP kinases by p561Ck. Scicncc 255, 853-855.
Fink, P. J., and Bevan, M. J. (1978). H-3 antigens ol' the thymus dclcrminc lymphocyte spcci fici ty. J. Exp. Med. 148, 766-775.
Finkcl, T. H., McDuffic, M., Kapplcr, J. W., Marrack, P., and Cambicr, J. C. (1987). Bolh irnmalurc and malurc T cclls mobilizcd ca2' in rcsponsc to antigcn rcccptor crosslinking. Naturc 330, 179- 18 1.
Flanagan, W. M., Corthksy, B., Bram, R. J., and Crabtree, G. R. (1991). Nuclcar association of a T-cc11 transcription factor blmked by FK-506 and cyclosporin A. Naturc 352, 8903-807.
Fowlkcs, B. J., Schwartz, R. H., and Pardoll, D. M. (1988). Dcletion of sclf-rcaclivc thymocytcs occurs at a CD4W prccursor stage. Naiurc 334, 620-633.
Frcarson, J., and Alesander, D. (1996). Prolein tyrosine phosphatascs in T-cc11 dcvclopmcnt, apoptosis and signalling. Immunol. Today 17, 385-391. Fujiwara, K., Y asuno, M., and Kitagawa, 1. (1YM1). Novcl preparation 01 immunogcn ior hydrophobie haptem enzyme i mrnunoassay for daunomycin and adriamycin. J. of 1mmunol. Meth. 45, 195-203.
Furukawa, T., Itoh, M., Krueger, N., Streuli, M., and Saito, H. ( 1994). Spccific interaction 01' thc CD45 protcin-tyrosine phosphatasc with tyrosine-phosphorylated CD3 < chain. Proc. Natl. Açad. Sci. USA 91, 10938- 1.0932.
Gcrvais, F., and Veillette, A. (1995). The unique amino-terminal domain of p56'ck rcgulatcs interactions wi th tyrosine protein phosphatases in T l ymphocytcs. Mol. Ccll. Biol. 15, 2393- 240 1.
Gilliland, L. K., Teh, H.-S., Uckun, F. M., Norris, N. A., Teh, S.-J., Schievcn, G. L., and Lcdbetter, J. A. (1991). CD4 and CD8 are positive regulators of T cc11 rcccptor signal transduction in earl y T ce11 di fferentiation. J. Immunol. 146, 1759-1765.
Godfrey, D. I., Kennedy, J., Mombaerts, P., Tonegawa, S., and Zlotnik, A. (1994). Onsct of. TCRp rearrangement and role of TCRB expression during CD3-CD4-CDK thymocytc diffcrentiation. J. Immunol. 152, 4783-4793.
Groves, T., Katis, P., Madden, Z., Manickam, K., D., R., Wu, G., and Guidos, C. J. ( 1995). In vitro maturation of clonal CD4+CD8+ceIl lincs in rcsponse to TCR engagement. J. Immunol. 154, 50 1 1-5022.
Groves, T., Parsons, M., Miyamoto, N., and Guidos, C. J. (1997). TCR cngagcment of CD4+CD8+ thymocytes in vitro induccs car1y aspccts of posi tivc sclcction, but not apoptosis. J. 1inmunol. 1.58, 68-75.
Guidos, C. J., Weissman, 1. L., and Adkins, B. (1989). Intrathymic maturation of murinc T lymphocytes from CD8+ precursors. Proc. Natl. Acad. Sci. USA 86, 754-7546
Guidos, C. J., Danska, J. S., Fathman, C. G., and Weissrnan, 1. L. (1990). T ccll rcceptor- mediatcd negative selection of autoreactivc T lymphocyte prccursors occurs after commitmcnt to the CD4 or CD8 lincages. J. Esp. Med. 172, 835-845.
Guidos, C. J. ( 1996). Positivc sclcction OS CD4' and CD8+ T cclls. Curr. Opin. Immunol. 8, 225-333.
Gulbins, E., Coggcshall, M., Baicr, G., Katzav, S., Burn, P., and Altman, A. (1993). Tyrosinc kinase-stimulatcd guaninc nucleotidc cxchangc activity of vav in T cc11 activation. Scicncc 260, 832-824.
Hathcock, K. S., Laszlo, G., Diçklcr, H. B., Sharrow, S. O., Johnson, P., Trowbridgc, 1. S., and Hodcs, R. J. ( l992). Exprcssion of variable eson A-, B-, and C-specifiç CD45 dctcrminünis of pcriphcral and thymic T ce11 populations. J. Imtnunol. 148, 19-28. Hill, C. S., and Treisman, R. (1995). Transcriptional rcgulation by cxtraccllular signals: Mcchanisms and spcci ficity. Ce11 80, 199-31 1.
Hsi, E. D., Siegel, J. N., Minami, Y., Luong, E. T., Klausner, R. D., and Sarnelson, L. E. ( 1989). T ccll activation induces rapid tyrosine phosphorylation of a lirnited nurnbcr of ccllular substrates. J. Biol. Chem. 264, 10836-10842.
Huesmünn, M., Scott, B., Kisielow, P., and von Boehmer, H. (1991). Kinetics and efficicncy of positive selection in the thymus of normal and T ce11 receptor tnnsgenic mice. Ccll 66, 533- 540.
Hurlcy, T. R., Luo, K., and Sefton, B. M. (1993). Differential effects of exprcsion of thc CD45 tyrosine protein phosphatase on the tyrosine phosphorylation of the Ick, fyrr and c-src tyrosinc protein kinases. Mol. Cell. Biol. 13. 1651 - 16%.
Imbcrt, V., Peyron, J., Farahi Far, D., Mari, B., Auberger, P., and Rossi, B. ( 1994). Induction of tyrosine phosphorylation and T-ce11 activation by vanadate peroxide, an inhi bitor of protcin tyrosine phosphatases. Biochem. J. 297, 163-173.
Irving, B. A., Chan, A. C., and Weiss, A. (1993). Functional characterization of a signal tnnsducing motif present in the T cell antigen receptor 5 chai n. J. Ex p. Med. 177, 1093- 1 103.
Iwashima, M., Irving, B. A., van Oers, N. S. C., Chan, A. C., and Weiss, A. (1994). Sequential interactions of the TCR with two distinct cytoplasmiç tyrosine kinases. Science 263, 1 136- 2 139.
Jarneson, S. C., Hogquist, K. A., and Bevan, M. J. (1995). Positive selection of thymocytes. Annu. Rev. Immunol. 13, 93-126.
Johnson, N., Mcycr, C., Pingcl, J., and Thomas, M. ( 198%). Scqucncc Conservation in potcntial regulatory regions of the mouse and human leukocyte common antigcn gcnc. J. Biol. Chcm. 264,6220-6229.
Johnson, P., Greenbaum, L., Bottomly, K., and Trowbridge, 1. (1989b). Identification of thc ÜI tcrnütivçl y spliccd exons of murine CD45(T200) requircd for rcacli vi ty wiih B?-0 and othcr T-00-rcstrictcd anti bodies. J. Esp. Med. 169, 1 179- 1 184.
Johnson, P., Ostergaard, H., Wasden, C., and Trowbridgc, 1. (1992). Mutational anal ysis of CD45 a leucocyte-spccific protein tyrosine phosphatase. J. Biol. Chcm. 267, 8035-8041.
Julius, M., Maroun, C. R., and Haughn, L. (1993). Distinct rolcs Ibr CD4 and CD8 as ço- rcccpkxs in antigcn reccptor signalling. Immuiiol. Today 14, 176- 183. Junc, C., Flctchcr, M., Lcdbcttcr, J., Schievcn, G., Sicgal, J., Phillips, A., and Sainclsori, L. (1990a). lnhibition of tyrosinc phosphorylation prcvents T ccll rcccptcir-mediatcd signal transducrion. Proc. Natl. Acad. Sci. USA 87. 7727-7726.
Junc, C. H., Fletchcr, M. C., Lcdbctter, J. A., and Samelson, L. E. (1990b). Incrcascs in tyrosine phosphorylation arc detectable before phospholipase C activation aftcr T ceIl rcccptor stimulation. J. Immunol. 144. 1591- 1599.
Justement, L., Campbell, K., Chien, N., and Cambicr, J. (1991). Regulation OS B ccll antigcn rcccptor signal transduction and phosphorylation by CD45 Sciencc 252, 1839- 1842.
Kapplcr, J. W., Roehm, N., and Marrack, P. (1987).T-cell tolerance by clonal climination in lhc thymus. Cc11 49, 273-280.
Kearse, K., Takahama, Y., Punt, J., Sharrow, S., and Singcr, A. (1995). Early molccular events induced by TCR signaling in immature CD4+CD8+ thymocytes. J. Exp. Med. 181, 193- 202.
Kishihara, K., Penninger, J., Wallace, V. A., Kündig, T. M., Kawai, K., Wakcham, A,, Timms, E., Pfeffer, K., Ohashi, P. S.,Thomas, M. L., Furlonger, C., Paige, C. J., and Mak, T. W. (1993). Normal B lymphocyte development but impaired T ce11 maturation in CD45-Exon6 protcin tyrosine phosphatase-deficient mice. Ce11 74, 143-156.
Kishimoro, H., Surh, C. D., and Sprent, J. (1995). Upregulation of surface markers on dying lhymocyles. J. Exp. Med. 181, 649-655.
Kisiclow, P., Bluthmann, H., Staerz, U. D., Steinmetz, M., and von Boehmer, H. ( 1988). Tolerancc in T-cell receptor transgenic mice involves deletion of nonmaturc CD4+8' thymocytcs. Naturc 333, 742-746.
Kisielow , P., and Maizek, A. ( 1995). Positivc sclection of T cells: Rcscue from programmcd ccll dcath and differcntiation rcquirc conlinual engagcmcnt of thc T ce11 reccptor. J. Esp. Med. 181, 1975- 1984.
Korctzky, G. A., Picus, J., Thomas, M. L., and Weiss, A. (1990). Tyrosinc phosphatasc CD45 is csscntial roi- coupling T cc11 anligcn receptor io tlic phospliatidyl inositol pathway. Naturc 246, 66-68.
Korctzky, G., Picus, J., Schullz, T., and Weiss, A. (1991). Tyrosinc pliosphatasc CD45 1s rcquired for T-cell antigen rcceptor and CD2-mcdiatcd activation of a protcin tyrosinc kinasc and intcrlcukin 3 production. Proc. Natl. Acad. Sci. USA 88, 3037-7041.
Korelzky, G. A., Kohmctscher, M. A., Kadleck, T., and Weiss, A. (1993). Rcstoration of T cc11 rcccptor-incdialed signal transduction by transfcction of CD45 cDNA into a CD45-dcficicnt lfaririntof the Jurkat T ccll linc. J. Immunol. 149, 1 138-1 142. Kozieradzki, I., Kundig, T., Kishihara, K., Ong, C., Chiu, D., Wallacc, V., Kawai, K., Timms, E., Ionescii, J., Ohashi, P., Marth, J., Mak, T., and Pcnninger, J. (1997). T cc11 development in mice espressing splice variants of the protein tyrosinc phosphatasc CD45 S. Imrnunol. 158, 3 130-3139.
Kubo, R. T., Born, W., Kapplcr, J. W., Marrack, P., and Pigeon, M. (1989). Characterization ol' a monclonal antibody which detccts al1 murine ap T ce11 receptors. J. Immunol. 142, 3736- 3743.
Laemmli, U. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.
Ledbetter, J. A., and Herzenberg, L. A. (1979). Xenogeneic monoclonal antibodies to mousc lymphoid differentiation antigens. Immunol. Rcv. 47, 63-90.
Ledbctter, J. A., Norris, N. A., Grossmann, A., Grosmaire, L. S., June, C. H., Uckun, F. M., Cosand, W. L., and Rabinovitch, P. S. (1989). Enhanced transmembrane signalling activity of monoclonal antibody heteroconjugates suggests interactions between receptors at the T cc11 surfacc. Mol. Immunol. 26, 137- 145.
Lefrancois, L., and Goodman, T. (1987). Developmental sequence of T200 antigcn modifications in mutine T cells. J. Immunol. 139, 3718.
Leitenbcrg, D., Novak, T., Farber, D., Smith, B., and Bottomly, K. (1996). The cxtracellular domain of CD45 controls association with thc CD4-T ce11 reccptor complcx and thc rcsponsc to antigen-specific stimulation. J. Exp. Med. 183, 249-259.
Lcvclt, C. N., and Eichmann, K. (1995). Receptors and signals in carly thymic sclcction. Immunily 3, 667-672.
Lcvin, S., Andcrson, S., Forbush, K., and Pcrlmuttcr, R. (1993). A dominant-ncgülivc transgcnc dclincs a rolc for p561ck in thymopoiesis. EMBO J. 12. 1671- 1680.
Lcwis, S. M. (1994). The rnechanism of V(D)J joining: Lessons from rnolccular, immunological, and comparative analyses. Adv. Immunol. 56, 27- 149.
Liao, X., and Liltman, D. (1995). Altered T cell rcceptor signaling and disrupted T cc11 dcvelopmcnt in micc lacking Itk. Imrnunity 3, 757-769.
Lincttc, G. P., Grusby, M. S., Hcdrick, S. M., Hanscn, T. H., Glimclicr, L. H., and Korsmcycr, S. J. (1994). Bcl-2 is upregulated at thc CD4+CDgf siagc during positivc sclcction and promolcs lhymocytc diffcrcntiation at several conlrol points. Immunity 1, 197-205. Lucas, B., Vasseur, F., and Pcnit, C. (1994). Production, sclcction, and maturiilion of thymocytes ivith higb surfacc dcnsity of TCR. J. Immunol. 153, 53-67.
Lucas, B., Vasscur, F., and Penit: C. (1995). Stochastic coreceptor shiit-cif'f is restrictcd to thc CD4 lineagc maiuration pathway. J. Esp. Med. 181, 1623-1633.
Lundberg, K., Heath, W., Konlgen, F., Carbone, F. R:, and Shortman, K. (1995). lntermediatc steps in positive sclection: Differentiation of CD~%'"'TCR~"' thymocytes into CD4- X'TCR~~thymocytes. J. Exp. Med. 181, 1643-1651.
Luo, K., and Sefton, B. M. ( 1992). Activated Ick tyrosine protcin kinase stimulates anti gen- independent interleukin-2 production in T cells. Mol. Ce!!. Biol. 12, 4724-4732.
Malissen, B., and Malissen, M. (1996). Functions of TCR and pre-TCR subunits: lessons Srom gene ablation. Curr. Opin. Immunol. 8, 383-393.
Markowitz, D., Gol'i, S., and Bank, A. (1988). A safe packaging line for genc translèr: separating viral genes on two different plasmids. J. Virol. 62, 1120-1124.
Maroun, C., and Julius, M. (1994). Distinct involvement of CD45 in antigen receptor signalling in CD4+ and CD8+ primary T cells. Eur. J. Immunol. 24, 967-973.
McCarthy, S. A., Kruisbeek, A. M., K., U. I., Sharrow, S. O., and Singer, A. (1988). Engagement of the CD4 molcculc influences cell surfacc cxprcssion of thc T-ccll rcccptor on thymocytes. Nature 336, 76-79.
McFarland, E., Hurley, T. R.. Pingel, J. T., Scfton, B., Shaw, A., and Thomas, M. L. (1993). Corrclation beiwcen Src fami l y mcm ber rcgulation by thc protcin- tyrosine- phospharasc CD45 and trünsmcmbranc signaling through the T-ceIl rcccptor. Proc. Natl. Acad. Sci. USA 90, 1402- 1406.
McLachlan, A., Milich, D., Raney, A., Riggs, M., Hughes, J., Sorgc, J., and F., C. (1987). Expression of hepatitis B virus surface and core antigens: influences of pre-S and prccorc sequences. J. Virol. 61, 683-692.
Morirefi, I., LaFcvre-Bernt, M., Sicheri, F., Husc, M., Lcc, C., Kuriyan, J., and Millcr, W. (1997). Activation of the Src-Sarnily tyrosine kinase Hck by SH3 domain displacemcnl. Nalurc 385, 650-653.
Molina, T. J., Kishihara, K., Siderovski, D. P., van Ewijk, W., Narcndran, A., Timms, E., Wakcham, A., Paigc, C. J., Hartmann, K.-U., Veillette, A., Davidson, D., and Mak, T. W. (1%'). Prolound block in thyrnocyk devclopmcnt in micc lacking Naturc 357, 161 -1 64.
Mollo, D., Musci, M., and Korelzky, G. (1994). Surfacc expression of a hclcrologous phosphatasc complcmcnts CD45 dcficicncy in aT ce11 clonc. J. Exp. Mcd. 180, 1359- 1366. ~VIU~LCI~II,I ., LU~~L~IIQII,11. ~vI., ailu ~LILL~I~LII,n. \ 17071. LU.L~IU UGLLVULI\J~I VI LIIL 1 -GL.II ~yi~miii~ protcin kinasc pp56kP by thc CD45 phosphotyrosine phosphatase. Proc. Natl. Acad. Sci. USA 86, 6303-6336.
Mustelin, T., Coggeshall, K. M., Isakov, N., and Altman, A. ( 1990). T cc11 antigcn rcccpiur- mediatcd activation of phospholipase C requires tyrosine phosphorylütion. Scicnce 247, 1584- 1587.
Mustelin, T., Pessa-Morikawa, T., Autero, M., Gassrnan, M., Anderson, L., Gahmbcrg, O., and Burn, P. ( 1992). Regulation of the p5#>' protein tyrosine kinase by thc CD45 phosphotyrosine phosphatase. Eur. J. Immunol. 22, 1 173- 1178.
Mustelin, T., Williams, S., Tailor, P., Couture, C., Zenner, G., Burn, P., Ashwcll, J., and Altman, A. (1995). Regulation of the p70iq tyrosine protcin kinasc in T cells by thc CD45 phosphotyrosinc phosphaiasc. Eur. J. Immunol. 25, 942-946.
Nada, S., Obada, M., MacAuley, A., Cooper, J. A., and Nakayama, H. (1991). Cloning of a com plernentary DNA for a protei n-tyrosine kinase that speci fical 1y phosphory lates a negati vc regulatory site of p60C-SrC.Nature 351, 69-72.
Nakayama, T., Singer, A., Hsi, E. D., and Samelson, L. E. (1989). Intnthymic signalling in immature CD4%+ thyrnocytes results in tyrosine phosphorylation of thc T-ce11 receptor zcta chain. Naturc 341, 65 1-654.
Nakayama, T., June, C. H., Munitz, T. I., Sheard, M., McCarthy, S. A., Sharrow, S. O., Samclson, L. E., and Singcr, A. ( 1990). lnhibi tion of T cell rcceptor expression and l'unction in immature CD@ CD~'cclls by CD4. Science 249, 1558- 156 1.
Nakayama, T., Samelson, L. E., Nakayarna, Y ., Munits, T. I., Sheard, M., Junc, C. H., and Singer, A. ( 199 1). Ligand-stimulated signaling cvents in immature CD~' CD~'thymoçytcs espressing compctcnt T-ce11 receptor complexes. Proc. Natl. Acad. Sci. USA 88, 9949-9953.
Ncgishi, I., Motoyama, N., Nakayama, K.-I., Nakayama, K., Senju, S., Hatakcyüma, S., Zhang, Q., Chan, A. C., and Loh, D. Y. (1995). Esscntial rolc for ZAP-70 in both positivc and ncgativc sclcction of thymocytes. Nature 376, 435-438.
Niklinska, B., Hou, D., Junc, C., Weissman, A., and Ashwcll, J. (1994). CD45 tyrosine phosphatase and mcm branc anchoring arc rcquired for T-cc11 antigen rcccptor signal ing. Mol. Ccll. Bioi. 14, 8078-8084.
Normcnt, A., Saltcr, R. D., Parham, P., Engclhard, V. H.,and Littmün, D. ( 1988). Ccll-çcll adhcsion rncdiatcd by CD8 and MHC class 1 molccules. Nüturc 336, 79-8 1.
Nossal, G. J. V. ( 1994). Negative seleciion of lymphocytes. Cc11 76, 339-239. . .\'.".., *' M., A L'Y-., Y.,L,-I.-.L"I.e, W., V. V. *L\,..b, ""...'Y"L., I ., -..Y Y./LC\,I..I I) 1.. \*/, I,. Isof'orms of the trrinsmcmbrane tyrosine phosphatasc CD45 diîfcrcntially afi'cct T ccll rcçognition. Im~nunity1, 109- 1 19.
Ohashi, P. S., Pirchcr, H., Bürki, K., Zinkcrnagcl, R. M., and Hcngartncr, H. (1990). Distinct scquencc of' negativc or positivc selection implied by thymocytc T-cc11 rcccptor dcnsi tics. Naturc 346. 86 1-863.
Okada, M., Nada, S., Yamanashi, Y ., Yamamoto, T., and Nakagawa, H. (1991). csk : a protcin-tyrosinc kinase involved in thc regulation of src family kinascs. J. Biol. Chcm. 266, 34249-24252.
Okumura, M., and Thomas, M. (1995). Regulation of immune function by protcin tyrosinc phosphatases. Curr. Opinion in Immunol. 7, 312-319.
Ong, C., Chui, D., Teh, If., and Marth, J. (1994). Thymic CD45 tyrosine phosphahse regulaies apoptosis and MHC-restricted negativc sclection. J. Immunol. 152, 3793-3805.
Ostergaard, H. L., Shackelford, D. A.. Hurlcy, T. R., Johnson, P., Hyman, R., Sef'ton, B. M., and Trow bridge, 1. S. ( 1989). Expression of CD45 alters phospharylatinn of thc kk-cncodcd tyrosinc protein kinasc in murinc lymphoma T-cell lines. Proc. Natl. Acad. Sci. USA 86, 8959- 8963.
Ostcrgaard, H. L., and Trorvbridge, 1. S. (1990). Coclustering CD45 wi th CD4 or CD8 alters the phosphorylation and hnasc activity of p561Ck.J. Exp. Mcd. 172. 347-350.
Pao, L., and Cambicr, J. (1997). Syk, but not Lyn, recmitmcnt 10 B ccll antigen rcccptor and activation following stimulation of CD45- B cells. J. Immunol. 158, 2663-2669.
Pawson, T., and Gish, G. D. (1992). SH2 and SH3 domains: from structure to funclion. Cc11 71, 359-362.
Pawson, T. (1995).Prolcin modulcs and signalling nctworks. Naturc 373, 573-580.
Pctric, H. T., Livak, F., Scliatz, D. G., Strasser, A., Crispe, 1. N., and Shortman, K. (1993). Multiplc rcrirrangcmcnts in TCRu chain gcncs maximizc thc production 01' useful thymoçytcs. J. EXP.Mcd. 178, 615-622.
Pcyron, J., Vcrma, S., Malcfyt, R., Sancho, J., Tcrhorst, C., and Spits, H. (1991). Thc CD45 protcin tyrosine phosphatasc is requircd for the complction of the activation program Icriding to lymphokine production in thc Jurkat human T cell linc. Int. Immunol. 3, 1357-1366
Pingcl, J. T., and Thomas, M. L. (1989). Evidcncc that the Icukocytc-coininon antigcn is rcquired Sor antigen-induccd T lymphocyte proliferation. Ccll 58, 1055- 1065. ..VI., IL., Y. Y., V.VV..VCC., 1 ., UllU Y...* a... "IV, Y. \ L//d,. .UV.ICLLIVULILIII LI1 U LU.. LLllllllSl U,,," prolinc-rich SH3 binding sitc. Sciencc 259, 1157-61.
Rctli, M. ( 1989). Antigcn rcccptor tail cluc. Nalurc 338, 383-384.
Robey, E., and Allison, J. (1995). T-ce11 activalion: integrülion OS signals Srom thc antigcn rcceptor and costimulatory molecules. Immunol. Today 16, 3M-3 IO.
Romeo, C., Amiot, M., and Seed, B. (1992). Sequence requirements for induction of cylolysis by theT ceIl antigen1Fc receptor 1; chah Ce11 68, 889-897.
Rowc, W. P., and Pincus, T. (1972). Quanlitativc studies of naturally occurring murinc leukcmia virus infcctions in AKR mice. J. Esp. Med. 135, 479-436.
Rozdzial, M., Kubo, R., Turner, S., and Finkel, T. (1994). Developmental rcgulation of thc TCR t- Chain. J. Immunol. 153, 1563-1580.
Samclson, L. E., Patel, M. D., Weissman, A. M., Harford, J. B., and Klausncr, R, D. (1986). Anligen activation of murine T cells induces tyrosine phosphorylation of a polypeptide associated with the T ce11 antigen receptor. Cell46, 1083-1090.
Samelson, L. E., Phillips, A. F., Luong, E. T., and Klausner, R. D. (1990). Association of thc fyn protein-tyrosine kinase with the T-ce11 antigen receptor. Proc. Natl. Acad. Sci. USA 87, 4358-4362.
Sancho, J., Silvcrman, L. B., Castigli, E., Ahern, D., Laudano, A. P., Terhorst, C., Gcha, R. S., and Chatila, T. A. (1992). Developmental regulation of transmembrane signaling via thc T cc11 antigcn rcceptorlCD3 complex in human T lymphocytes. J. Immunol. 148, 13 15- 13-1.
Schauz, D. G., Octlingcr, M. A., and Baltimore, D. (1989). The V(D)J rccombination activating gcnc, RAG- 1. Cc11 59, 1035- 1048.
Schravcn, B., Schoenhaut, D., Bruyns, E., Koretzky, G., Eckcrskorn, C., Wallich, R., Kirchgcssncr, H., Sakorafas, P., Labkovsky, B., Ratnrifsky, S., and Mcucr, S. ( 1994). LPAP, anovcl33-IïDa phosphoprotcin Lhal intcracls with CD45 in human lymphocytes. J. Biol. Chcm. 269, 39 102-391 1 1.
Sccrisl, J., Burns, L., Karnitz, L., Koretzky, G., and Abrahams, R. (1993). Stimulalory cfl'cct ol' the prolein Lyrosinc phosphatase inhibitor, pervanadatc, on T-ccll activalion cvcnls. J. Biol. Chcm. 268, 5886-5893.
Shahinian, A., Pfcffcr, K., Lcc, K. P., Kündig, T., Kishihara, K., Wakcham, A., Kawai, K., Ohashi, P. S., Thompson, C. B., and Mak, T. W. (1993). Diff'crcnlial T cc11 costimulatory rcquircmcnts in CD28-deficient micc. Science 261, 609-6 13. Shiroo, M., Goff, L., BiffCn, M., Shivnan, E., and Alexander, D. (1993). CD45 Lyrosinc phophalÿsc-activaicd p59fYa coiiplcs ihe T cell antigcn receptor to pathways of diacylglyccrol prcxiiiction, protcin kinasc C activation and calcium flux. EMBO J. 1 1, 4887-4897.
Shortman, K., Vrcmcc, D., and Egerton, M. (1991). The kinetics of T ceIl antigcn rcccptor csprcssion by subproups of CD4'8' thymocytes: Delineation of ~~418'3~'thymocyics as post-selcction intcrmediates leading to mature T cells. J. Exp. Med. 173, 323-332.
Sicheri, F., Moarefi, I., and Kuriyan, J. (1997). Crystal structurc of thc Src Family tyrosine kinasc Hck. Nature 385, 602-609.
Skh, M., Bolen, J. B., and Weiss, A. (1993). CD45 specifically modulates binding of Lck to a phosphopeptidc encompassing the negative regulatory tyrosine of Lck. EMBO J. 12, 3 15-321.
Spangrudc, G. J., and Scollay, R. (1990). A simpliPied method for enrichment of mousc hcmatopoictic stem cclls. Exp. Hematol. 18, 920-926.
Stein, P. L., Lee, J.-M., Riçh, S., and Soriano, P. (1992). pp59fY" mutant miçe display differential signaling in thymocytes and peripheml T cells. Cell 70, 741-750.
Straus, D. B., and Weiss, A. (1992). Genctic cvidence for the involvement OS the Ick tyrosinc kinasc in signal transduction through the T ceIl antigen receptor. Ccll 70, 585-593.
Suzuki, H., Punt, J., Grangcr, L., and Singer, A. (1995). Asymmetric signaling requirerncnts for thymocyre commitment to the CD4' versus CD8+ T ce11 lineages: A ncw pcrspectivc on thymic commitment and selection. Irnmunity 2, 413-425.
Swan, K. A., Alberola-Ila, J., Gross, J. A., Appleby, M. W., Forbush, K. A., Thomas, J. F., and Perlmutter, R. M. (1995). Involvement of p21m distinguishes positive and ncgativc sclcction in thymocytes. EMBO J. 14, 276-285.
Tarakhovsky, A., Kanner, S. B., Hombach, J., Lcdbettcr, J. A., Müllcr, W., Killccn, N., and Rajewsky, K. (1995). A role for CD5 in TCR-mediated signal transduction and Lhymocytc sclcction. Scicncc 269, 535-537.
Tch, H. S., Kisiclow, P., Scott, B., Kishi, H., Uematsu, Y., Bluthmann, I-I., and von Bochmcr, H. (1988). Thymic major histocompütibility complcs antigcns and thc (x(3 T-cc11 rcccptor dctcrminc thc CD4CD8 phenolype OC T-cells, Nature 335, 229-233.
Thomc, M., Duplay, P., Guttingcr, M,, and Aculo, 0. (1995). Syk and ZAP-70 incdilitc rccruitmcnt of p~6ickl~~4to thc aciivaicd T cc11 rcceptorlCD3/~complcs. J. Esp. Mcil. 181, 1097-2006.
Tonks, N., and Ncel, B. (1996). From form to function: Signaling by protein tyi-osinc phosphatascs. Ccll, 365368. Towbin, H., T., S., and Gordon, J. (1979). Elcctrophorctic transt'cr of prolcins f'i-on1 polyacrylainidc gcls to nitroccllulosc sheets: procedurc and somc applications. Proc. Natl. Acad. Sci. USA 76, 4350-4344.
Trowbridge, 1. S., and Thomas, M. L. ( 19%). CD45 An emerging rolc as proicin lyrosinc phosphatase required for lymphocyte development and activation. Ann. Rcv. Immunol. 12, 85- 116.
Turka, L. A., Linsley, P. A., Paine, R., Schieven, G. L., Thompson, C. B., and Ledbctlcr, J. A. (1991a). Signal transduction via CD4, CD8, and CD28 in mature and immaturc thymocytcs. Implications for thymic selection. J. Immunol. 146, 1428-1436.
Turka, L. A., Schatz, D. G., Oettinger, M. A., Chun, J., Gorka, C., Lce, K., McCormack, W. T., and Thornpson, C. B. (1991b). Thymocyte expression of RAG- 1 and RAG-3: Tcrminaiion by T cell receptor cross-linking. Science 253, 778-781.
Turka, L., Kanner, S., Schieven, G., Thompson, C., and Ledbetter, J. (1992). CD45 mcidulatcs T cell rcceptorlCD3-induccd activation of human thymocytes via rcgulation of' tyrosinc phosphorylation. Eur. J. Immunol. 22, 551-557.
Turner, M., Mee, P. J., Costello, P. S., Williams, O., Price, A. A., Duddy, L. P., Furlong, M. T., Gcahlen, R. L., and Tybulewicz, V. L. J. (1995). Perinatal lethality and blockcd B-cell dcvelopment in mice lacking the tyrosine kinase Syk. Nature 378, 298-302. van Ocrs, N., Killeen, N., and Weiss, A. (1996). Lck regulates the tyrosine phosphorylation of thc T cell receptor subuniis and ZAP-70 in murine thymocytes. J. Exp. Med. 183, 10%- 1063.
Vaux, D. L., Cory, S., and Adams, J. M. (1988). Bcl-2 gene promotes hacmopoictic cc11 survival and cooperates with c-myc to immortalize pre-B cells. Nature 335, 440-442.
Vcillctlc, A., Bookman, M. A., Horak, E. M., and Bolen, J. B. (1988). The CD4 and CD8 T ccll surface antigens are associated with the inlernal membrane tyrosine-protein kinasc p561ci"'k. Ccll 55, 30 1-308.
Vcillctic, A., and Düvidson, D. (1992). Src-rclüted protcin tyrosinc kinüscs and T-cc11 rcccptur signalling. Trcnds in Gcnct. 8, 61-66.
Vcis, D. J., Scntman. C. L., Bach, E. A., and Korsmcycr, S. J. (1993). Expression ol'the 13~1-2 protcin in murinc and human thymocytes and in pcriphcral T lymphocytes. J. Immunol. 151, 546-2554.
Volarcvic, S., Niklinska, B., Burns, C., Yarnada, H., Junc, C., Dumont, F., and Ashwcll, 1. (1992). Thc CD45 tyrosine phosphatasc regulates phosphotyrosinc homcostasis and ils loss voiürevtc, a., iuiwnsr;a, a., auins, L., r arriaua, n., Juiie, L., ~~urriuri~,r., ariu nsriwcii, J. ( 1992). Thc CD45 tyrosinc phosphatase regulates phosphotyrosine homeostasis and ils loss rcvcals a novel pattern of latc T ccll rcccptor-induced ~a2+oscillations. J. Exp. Mcd. 176, 835- 844.
Volarevic, S., Niklinska, B. B., Burns, C. M., Junc, C. H., Wcissman, A. M., and Ashwcll, J. D. (1993). Regulation of TCR signaling by CD45 lacking transmcmbranc and cstraccllular domains. Sciencc 260. 541-546. von Boehmer, H. (1996). CD4KD8 lineage commitment: back to instruction? J. Exp. Med. 183, 713-715.
Wallace, V. A., Fung-Leung, W. P., Timms, E., Gray, D., Kishihara, K., Loh, D. Y., Penninger, J., and Mak, T. W. (1992). CD45R.4 and CD~~RB~~~~expression induced by th picselection events. J. Exp. Med. 176, 1657- 1663.
Walunas, T. L., Sperling, A. I., Khattri, R., Thompson, C. B., and Bluestonc, J. A. (1996). CDB espression is not essential for positive and negative selection of thymocytes or pcripheral T ce11 tolcrance. J. Immunol. 156, 1006-1013.
Wang, C.-R., Hashimoto, K., Kubo, S., Yokochi, T., Kubo, M., Suzuki, M., Suzuki, K., Tada, T., and Nakayama, T. (1995). T ceIl receptor-mediated signaling events in CD4+CD8+ thymocytes undergoing thymic selection: Requirement of calcineurin activation for thymic positive selection but negative selection. J. Exp. Med. 181, 927-941.
Wange, R., and Samelson, L. (1996). Complex complexes: signaling at the TCR. Immunity 5, 197-205.
Würd, S., June, C., and Olive, D. (1996). P13-Kinase: a pivoial pathway in T-cc11 activation? Immunol. Today 17, 187-197.
Wcaver, C., Pingcl, J., Nclson, J., and Thomas, M. (1991). CD81 T-cc11 clones deficicnl in thc expression of thc CD45 protein tyrosinc phosphatase have impaircd rcsponscs to T-cell rcceplor stimuli. Mol. Ccll. Biol. 11, 4415-4432.
Wciss, A., and Littman, D. R. (1994). Signal transduction by lymphocyte aniigcn rcccptoi-S.Ccll 76, 263-374.
Wiest, D. L., Yuan, L., Jefferson, J., Benveniste, P., Tsokos, M., Klausncr, R. D., Gliinchci, L. H., Sarnelson, E., and Singer, A. ( 1993). Regulation of T ceIl rcceptor espression in irnrnaturc CD4+CD8+th ymocyles by p56kk tyrosine kinase: Basis for di fferentiul signaling by CD4 and CD8 in imrnüiurc thyrnocytcs cxprcssing bolh corcccptor molccules. J. Esp. Med. 178, 1701-1712. ...-"., -. '., .--..-, ". -...> ..-->-.., -...-..> -. '., -..- -"'C" , . .. ,- - - ..,. a --. -- .. . -...... ZA WO is impaircd in CD4+CD8+ thymocytes as a conscquencc of intrathymic interactions Lhai dirninish wailablc p561ck.Iminuni ty 4, 495-504,
Wilkinson, R. W., Anderson, G., Owen, J. J. T., and Jenkinson, E. J. (1995). Posi~ivc sclcction of thymocytes involvcs sustaincd interactions with the thymic microenvironment. J. Immunol. 155,5234-524.
Wu, L., Scolhy, R., Egerton, M., Pearse, M., Spangrudc, G. J., and Shortman, K. (1991). CD4 espressed on earl iest T-lineage precursor cells in the adull mu rine thymus. Naiurc 349, 7 1 - 74.
Xu, W., Hamson, S., and Eck, M. (1997).Three-dimensional struclure of the tyrosine kinasc c- Src. Naturc 385, 595-603.
Yamaguchi, H., and Hendrickson, W. (1996). Structural basis for activation of hyman lymphocytc kinase Lck upon tyrosine phosphorylation. Nature 384, 484-489.
Yokoyama, W. M., Koning, F., Kehn, P. J., Pereira, G. M. B., Stingl, G., Coligan, J. E., and Shevach, E. M. (1988).Characteri7ation of a cell surface-expressed disulfide-linked dimer involved in mutine T cell activation. J. Immunol. 141, 369-376.
Zamoyska, R., and Parnes, J. (29%). A CD8 polypeptide that is lost af'ter passing the Golgi but bef'ore reaching the ccll surface: a novel sorting mechanism. EMBO 7, 2359-2367.
Zamoyska, R., Dcrharn, P., Gorman, S. D., von Hocgen, P., Bolen, J. B., Veillctlc, A., and Parncs, J. K. (1989).Iiiability of CDûa' polypeptides to associate wiih p?6lck correlates wilh irnpai rcd f'unction itr virro and lack of expression iti vivo. Nature 342, 278-28 1.
Zinkernagel, R. M., Callahan, G. N., Althage, A., Cooper, S., Klein, P. A., and Klein, J. (1978). On the thymus in the differentiation of "H-2 self-recognition" by T-cclls: Evidence for dual recognition? J. Exp. Med. 147, 882-896. 2 L ILILL IlIlIl IIIIII IIlllS
APPLIED -k IWGE. lnc -= 1653 East Main Slreel ----. - Rochester. NY 14609 USA ---- Phone: 71 61482-0300 ---- Fax: 7161288-5989
0 1993. Applied Imagc. lnc . Ali Aigh15 Resorved