XECHANESV FACTOR '3 MECHANISMS OF IMMUNE MODULATION BY TRAN SFORMING GROWTH FACT OR- B]: A ROLE FOR SMAD3 AS AN INTRACELLULAR MEDIATOR By Susan C. McKams A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSPOPHY Department of Pharmacology and Toxicology 200 l ‘1 .C I. . . ,3- ,'. . e . :t v .2; i- . . .1. ~11; . O .i. .‘2. E5: .. . . . _. VI: ' 1L. . «i? ‘. ‘,- 'r'f' lr< ,- a ;.3 ... i" ‘__:3 § .' t' O 1'. lad-I I ‘ Ll ¢ in- up ‘I or" 4.. .. ‘,,,,z' i... .t..:-. ‘ *Lfin'. _ ,,.. I o '“n ,. o .‘t'uf‘ . ' HIM ,,A.'." Iv Mv-o . “Tu-i. .» {A 1“, J an n a” u 5 o v .. o. um WNUA-h'orw uu‘lnm-C‘ “Mu-f: «gum» - “warmers? FACTOR-3 -. A ".: I T's.- . . dbl-J 'lou4.\“- . -— ,pa— .F’P‘ ‘ 5 c __,..-r.\ ‘3 0““. .-.[ «I'- ~; ’. -‘ __.‘_...._.\L O. T( -V "4 ,..‘ . v "'I gar-ottkhn .Aa-nh's.u-. -r-v- 9P ' .‘ul -3 In W 1'”?' .Ab“... x‘y . . .‘ «".. _ ‘ ._ tar..\u4 .i ..::_, .. 3'0- '0. ~- K..§«~....d-H-l§'] "h - ‘ 5‘... ‘hv- V can, \ . Q "Hui‘ - O . l ‘ am“) 3:4 ~ . ‘H x . \ “A — ~ u“ {9'1 vy. “l d. 2‘." . v “qt-:34 ' *4 “.2 ' ¥\;_:\ , 't‘ at ‘~ “.""n‘."‘.‘! I ' hA x -. 5‘ I‘ ’- ABSTRACT MECHANISMS OF IMMUNE MODULATION BY TRANSFORMING GROWTH FACT OR-Blz A ROLE FOR SMAD3 AS AN INTRACELLULAR MEDIATOR By Susan C. McKarns TGF-Bi (Iransforming growth factor-Q1) produces a diverse range of physiological immune modulatory effects. Targeted deletion of the TGF-B, gene exemplifies the importance of TGF-B, on maintaining immune homeostasis. TGF-B,"’ mice develop a multifocal inflammatory disease with a T cell-dependent autoimmune process. A role for TGF-B, in maintaining T cell tolerance and demonstration that TGF-B, signaling in T cells is critical for maintaining B cell tolerance is illustrated by spontaneous T cell differentiation and autoimmunity in transgenic mice expressing a dominant negative TGF-B receptor under the control of a T cell-specific promoter. Mice homozygous for the TGF-B, receptor-activated Smad3 (Smad3"‘) succumb to a deficiency in mucosal immunity and defective T cell responsiveness to TOP-Bl. In light of these observations, the overall objective of this research was to test the hypothesis that TGF-B, acts directly on T cells to regulate IL-2 (interleukin-g) expression through Smad-mediated signaling. RT-PCR and ELISA analyses demonstrated a concentration-dependent bifunctional augmentation and attenuation of IL-2 expression by TGF-Bl in a-CD3 + OL-CD28- activated T cells. In agreement, a concentration-dependent stimulation and suppression of immunoglobulin production by TGF-B, in antigen-stimulated B cells was also established. T cells obtained from Smad3”’ mice were refractory to inhibition of IL-2 mRNA and IL-2 protein secretion by TOP-[3,. Inhibition of a-CD3 + a-CDZS-induced T '50:” . " Aviva» n. N"! “H u .\l m" ‘ " H. l‘ I,‘ . ' \ \ \ . ‘ ha-V" I-y-X .1 ' ~~~»n‘;. \tt. I s . .‘ \‘V‘qr ‘n~'.~ .m£-, -€\JIK.\f ‘;~‘~ . I‘ V (’3';H ‘L- O ‘ Muss,“ . 'O.h. -.‘ . _ “4 ~.¢\:.. D‘\:\ H... k ... 3“..“ v - h I ‘-._\.5 I .|-\I"' ‘F‘ n. “‘ - “-9. , ~ O ~~ ‘ . '9 ‘ 2's ‘ '3‘ ‘\ .| : . I.“ 0 IL T. - . y , A’ .g ‘ I ‘U .O:: a. » “1L1 cell growth by TGF—Bl was also attenuated in Smad3”’ T cells. Collectively, these results establish a role for Smad3 in regulating IL-2 production and T cell growth in response to TOP-Bl. Inhibition of LPS (lipopolysaccharide)-activated B cell growth was unaffected in Smad34' mice suggesting a cell-type specificity for Smad3 signaling. Demonstration of a reversal of TGF-Bl-induced T cell growth arrest by exogenous IL-2 suggested that Smad3 may modulate T cell growth through a direct effect on IL-2 production. These results provide a putative mechanism whereby TGF-B, may selectively and specifically target T cell growth. Smad3-null B cells were demonstrably less sensitive to the inhibitory effects of TGF-B, on IgM production in vitro. Moreover, a role for Smad3 in mediating IgA production by TGF-B, in vitro was evident by an inability of TGF-B, to augment IgA secretion in LPS-activated Smad3"’ splenic B cells. Five putative M Smad3 response elements were identified in the IL-2 promoter. EMSA analyses demonstrated that M sequences were critical for constitutive and a-CD3 + a-CD28- induced DNA binding activity at the CD28RE and proximal AP-l (activating protein-l) site, but not the distal NFAT or NRE-A sites in the IL-2 promoter. Smad3 binding to the proximal AP-l site was CAQA-independent suggesting that Smad3 may mediate its effects through protein-protein interactions. TGF-Bl stimulated ERK MAP (extracellular regulated kinase mitogen activated protein) kinase activity in splenocytes. Blockade of ERK MAP kinase activity augmented Smad3 nuclear expression and attenuated inhibition of IL-2 by TGF-B, implicating Smad and MAP kinase regulatory cross talk. Taken together, these data substantiate a role for Smad3 signaling in the regulation of IL-2 production and provide a novel mechanism of immune homeostasis. DEDICATION To Mike, Ellen, Michael, Angie, Tabitha, and Brian iv Ft‘fff:€‘\i. 1 . 4-,. )d'." ' .‘x-‘Jh" EC m.L “ . Y' ‘ a ”‘3 h‘\'-.’ ‘ I! \ I“ ‘h v ,5, u.n\ .. a, '\ ' . l “13\‘ [5“ IL” ‘4... ”an MA 'vr‘c a: ‘F i ‘u‘ '1 ‘L.... u.\. A... k . 1A. . .1. ' '~ t.-. 1-.‘L. I 355 l 1'3““..23‘ "' '7‘". basin“. LU :...-l.. Mn . I “A“: ~ " ‘.‘:H s\. m . El“ ‘ Ra.) “ 5N . Ii'h‘gv- .' -M.L~‘:_ ‘ L ~ in I 'fi.. ,-. ‘5}: ' I - g h. (x -. V‘V v‘rTMu ~ .- ~.i UM .\.'\ ACKNOWLEDGEMENTS Foremost, I wish to express my sincere heartfelt appreciation to the members of my dissertation committee - Drs. Norbert E. Kaminski, Robert A. Roth, Michael P. Holsapple, James E. Trosko, and Lawrence J. Fischer for your insightful guidance, discussions, recommendations, support, and encouragement throughout my graduate studies. I had the distinct privilege of knowing many of you before I began my formal training and have come to respect you all. It was an honor to have shared my research with you. I have learned a great deal from each of you and sincerely thank you for your dedicated commitment to teach me. I most sincerely thank Dr. Jay I. Goodman for the contributions that you have made to my career development. You have been a role model for a very long time. One day, I hope to find a mentor who is equally as good, but never do I expect to find one who is better. I sincerely thank Dr. John J. Letterio for providing me with the opportunity to test my hypothesis. I am grateful for the tremendous effort that you have put forth towards my training. Our collaborative research efforts have been educational, productive, and gratifying. Our conversations have been intellectually stimulating. I look forward to working with you for many years to come. Dr. Gregory Fink, thank you for the expert statistical advice and guidance that you gave. Your comments were always critical and very much appreciated. Your contributions to my research have been significant. o,n" DY . ., .. r ~,1 - o '9 i ”5.53). A“ ““‘ .0 y to ‘ . ~1 n-' "" 32; [1‘15 z..J»‘ 4‘ l cxwv :' . 1', ,..1 Dan .‘ u~105~‘-‘\;:I‘\ ~.’ M- . I x ....... .r m. .:‘ ,‘_,,‘.1- A850 A ‘7’?"‘7‘l" " ...—..‘LusJ'. .L y I"""3'.31‘. --.‘s\l 55 ‘ ‘3‘ . ".3 “ ,av-v‘ in CAI.'H..... «4 W; for \. Ire-e. :0: {"I ' Lu l “h ‘ 49 “ i «up the "3*. “k"‘imVa . "g' 4'14 - 3‘ “Mk. 1 “ML W‘. '- ¢I44\e :0 5. c. §.', . ;. 3"", Dr. Stephanie Watts and the Watts’ laboratory, thank you for sharing your reagents, resources, protocols, and technology. Collectively, you have ‘turned mountains into mole hills’ and I sincerely appreciate your efforts. I express my gratitude to the faculty and staff of the Pharmacology and Toxicology Department, especially Drs. Kenneth Moore, Sue Barman, and Peter Cobbitt. Your support has been immense. It has been a pleasure to have been a part of your organization. I sincerely acknowledge and thank the many individuals, especially Drs. Peter ten Dijke, Carl-Henrik Heldin, Rik Derynk, Atsue Ochi, Roger Davis, Kathy Meek, Brad Upham, and Walt Esselman, who have graciously donated critical reagents to my research. Michael Flink, thank you for caring! To all the members of the Kaminski lab, thank you for your technical assistance and your friendship. I feel fortunate to have found individuals who believe in me and have worked hard to help me realize my dreams. I especially thank Drs. David J. Doolittle and R. Julian Preston — you have been far more than trusted mentors. The essence of this work can be traced back to my childhood. I thank my family for planting the seed of enthusiasm and determination in me. I thank God for giving me the courage and perseverance to finish, to move forward, and to give the world the things that I have to give. If I have ever doubted that we have the power to make choices, to give our lives new direction, I no longer do. vi LEI Of TABl IN 0F FlGl LEI Of ABBR [\TRODl'CIII I. S; H T( Ill. If G TABLE OF CONTENTS LIST OF TABLES ................................................................................. xi LIST OF FIGURES ............................................................................. xii LIST OF ABBREVIATIONS ................................................................... xvi INTRODUCTION .................................................................................. 1 1. Significance ......................................................................... 1 II. TGF-B Superfamily ................................................................. 2 III. TGF-B, ................................................................................. 5 A. Nomenclature ................................................................. 5 B. Transcriptional regulation of TGF-B, expression ...................... 5 C. Secretion of TGF-B, as a latent complex ................................ 8 D Activation of latent TGF-Bl ............................................. 14 1. Physiochemical activation ...................................... 14 2. Enzymatic activation and protein interactions .............. 15 3. Hormone-induced activation ................................... 16 E. TGF—Bl Receptors ........................................................ 16 1. TBRII (TGF-B Receptor Type II) .............................. 20 2. TBRI (TGF-B Receptor Type I) ................................ 2O 3. TBRIII (TGF-B Receptor Type III) ............................ 21 F. TGF-B receptor binding proteins ......................................... 21 G. Biological functions of TGF-B, ......................................... 22 H. TGF-Bl and immune modulation ....................................... 23 1. The TGF-B, null mouse model ................................ 24 2. T cells .............................................................. 24 3. B cells .............................................................. 26 4. TH l/T H2 development ........................................... 27 5. TGF-B, and IL-2 ................................................. 27 IV. Transcriptional regulation of the interluekin-2 gene following activation through the T cell antigen receptor ................................. 28 A. Control of IL-2 transcription ............................................... 28 B. Signaling cascades through the TcR that regulate of IL-2 gene expression .......................................................... 29 vii \Tll. kl. drama (:7 C. Regulatory elements of the IL-2 gene ................................. 33 1. NF-KB (nuclear factor-KB) .................................... 37 2. NFAT (nuclear factor of activated T cells) ................... 37 3. AP-l (activating protein-l) ..................................... 38 4. CREB (CAMP response element binding protein) .......... 42 5. ZEB (zinc finger/E box binding protein) ..................... 43 V. TGF-B1 signaling through Smads ............................................... 43 A. Structures and functions of Smad proteins ............................ 44 B Smad binding to sequence specific DNA motifs ..................... 51 C. Smads synergize with transcription factors ........................... 53 D Smads interact with transcriptional co-activators and co- repressors ................................................................... 54 VI. Implications for Smad regulation of immune cell function .................. 54 VII. Regulatory cross talk between Smad and MAP kinase signaling cascades kinase signaling cascades .............................................. 57 VIII. Objectives and Specific Aims ................................................... 61 IX. Relevance .......................................................................... 64 MATERIALS AND METHODS ............................................................... 66 I. Animals ............................................................................ 66 11. Cell lines ............................................................................ 66 HI. TGF—[31 ............................................................................. 67 IV. Chemicals ........................................................................... 67 V. Antibodies .......................................................................... 68 VI. Oligonucleotides .................................................................. 68 VII. Reporter gene plasmids .......................................................... 70 VIII. Isolation and in vitro activation of primary lymphocytes ................... 70 IX. In vitro proliferation assays ...................................................... 71 X. ELISA (Enzyme-Linked Immunosorbent Assay) ............................. 72 A. Mouse IL-2 ELISA ....................................................... 72 B. Mouse IL-4 ELISA ....................................................... 73 C. Mouse IFN-y ELISA ...................................................... 73 D. Human IL-2 ELISA ...................................................... 74 E. Mouse IgM ELISA ....................................................... 74 F. Mouse IgA ELISA ........................................................ 75 XI. Quantitative RT- PCR ............................................................. 76 A. Preparation of mouse IL—2 internal standard for RT-PCR .......... 76 viii E XII. P I .4 E C XIII. P XIV. X XX. E XVI I XVII. l XVIII P XIX. D XX. S NERIXIENT I. c. \l. XII. _\u \lH. E B. Quantitative Competitive RT-PCR for mouse IL-2 .................. 77 XII. Protein Extractions ................................................................ 78 A. Whole cell .................................................................. 78 B. Nuclear .................................................................... 78 C. Cytosolic ................................................................... 78 XIH. Protein determination ............................................................. 79 XIV. Western Immunoblotting ......................................................... 79 XV. Electrophoretic mobility shift assay (EMSA) ................................. 8O XVI. Transient Transfections ........................................................... 8O XVII. In vitro AFC (Antibody Forming Cell Response) ............................ 81 XVIII. Pronase Viability ................................................................... 83 XIX. Densitometry ....................................................................... 83 XX. Statistical Analysis ................................................................ 83 EXPERIMENTAL RESULTS ................................................................... 85 I. Concentration- and time-dependent effects of TGF-B, on T cell proliferation and IL-2 expression ................................................. 85 II. TGF-[3l bifunctionally modulates in vitro IgM AFC responses in a concentration-dependent manner ............................................... 119 III. Characterization of Smad proteins in mouse lymphoid tissue ............ 126 IV. Pathology of Smad3-null mice ................................................. 134 V. A role for Smad3 in the inhibition of IL-2 expression by TGF-B, ........ 137 VI. A role for Smad3 in the inhibition of lymphocyte proliferation by TGF [3, .......................................................................... 152 VH. Modulation of protein binding to the IL-2 promoter by TGF-[3].. . . . . 165 VIII. Effect of TGF-B, on NFAT and NF-KB transcriptional activity in Ot-CD3 + 0t-CD28-activated splenic T cells ................................ 181 IX. Involvement of MAPK in the regulation of Smad/TGF-B, signaling in T cells ............................................................................ 182 X. A role for Smad3 in immunoglobulin production by TGF-[3l in vitro ..... 199 ix ,._ , _ . z :4; .9' .3 . O 31" . . 3.-i Inr ‘. . l;. .:_ .,: .1 ‘ o‘ J‘- .4 . o' . C I 1 'T .. Q. i ‘ E'. . wasps. I. G x [1. S? .. VIII. A IX. X! X I( [y Xl. C lHIRtIt‘RE DISCUSSION ...................................................................................... 210 1. Concentration- and time-dependent regulation of T cell growth and IL-2 expression by TGF-Bl ................................................ 212 H. Smad3 is essential for inhibition of T cell growth and IL—2 expression by TGF-B, in vitro .................................................. 217 VIII. A Role for DNA sequence specific binding of Smad3 in the regulation of IL-2 Expression by TGF-[3l ..................................... 220 IX. MAPK and Smad signaling cross talk in regulating TGF-B, signaling in T cells ............................................................... 223 X. TGF-[31 differentially modulates humoral immune responses In Vitro ............................................................................ 225 XI. Conclusions ....................................................................... 230 LITERATURE CITED ......................................................................... 233 LIST OF TABLES Table 1. Phenotype comparison of Smad3-null and TGF-Bl-null mice ............... 55 Table 2. Oligonucleotides used for EMSA analyses .................................... 69 C can . a ' A». o.» I ‘ czar. 21.1 ' c . C '0 IL" “135mm... A.l‘:tmt- . .0..." C' "m u {nn'h-n‘url‘qm. .4 a. u- an . c .n M.- . . .- 4'1“!" - Q . -V ‘ *:.;.-;i'.:..'flu;..rt,. g‘. 4;; .- .. In. ' .‘ "0 '-"§"~ 0:" ‘ O L I». O .. I o ”1“». if" vw-u ‘ p’ I b v: 1; 3" mammmw‘ “-0 .:.,-' _. ,... 'IOF-c xi Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. LIST OF FIGURES Page Transcriptional regulation of TGF-[31 ............................................ 7 Diagram of the biologically inert latent precursor complex of TGF-BI .............................................................................. 10 Stepwise activation of latent TGF-B1 precursor to its biologically active form .......................................................... 13 Schematic representation of the type I and type II TGF-[3 receptors ....... 18 Two-signal model for T cell activation ......................................... 32 Schematic of the 5’ minimal essential regulatory region of the mouse H_.-2 promoter ............................................................. 35 Intracellular signaling pathways of TGF-I3l .................................... 44 Schematic of the functional domains of the Smad proteins .................. 46 TGF-[3l signaling through Smad proteins ...................................... 49 Putative regulatory cross talk between Smad signaling and MAPK signaling cascade .......................................................... 59 Identification of CAGA sequences in the 5’ minimal essential regulatory region of the mouse IL-2 promoter ................................ 63 Experimental design for characterizing the effects of TGF-B, on T cell activation and growth ................................................. 87 Proliferation of activated mouse splenic T cells and thymocytes ........... 89 Concentration-dependent effect of TGF-BI on proliferation of activated splenic T-cells and thymocytes ...................................... 92 Time of addition effect of TGF-Bl on a-CD3 + a-CD28-induced splenic T cell proliferation ....................................................... 94 Time of addition effect of TGF-B1 on a—CD3-induced splenic T cell proliferation ....................................................... 96 xii Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27. Figure 28. Figure 29. Figure 30. Figure 31. Figure 32. Figure 33. Time of addition effect of TGF-[3l on a-CD3 + a-CDZS-induced thymocyte proliferation ........................................................... 98 Time of addition effect of TGF-B, on a-CD3-induced thymocyte proliferation ......................................................... 101 IL-2 protein secretion in activated mouse splenic T cells and thymocytes ........................................................................ 103 Concentration-dependent effect of TGF—B, on IL-2 protein secretion in activated splenic T cells and thymocytes ...................... 105 TGF-B, does not induce IL-2 protein secretion in naive splenocytes or thymocytes in vitro ............................................ 108 Concentration-dependent effect of TGF-B, on IL-2 mRN A ............... 110 Time of addition effect of TGF-[3l on a-CD3 + a-CD28-induced IL-2 protein secretion in splenic T cells ...................................... 112 Time of addition effect of TGF-Bl on a-CD3-induced IL-2 protein secretion in splenic T cells ...................................................... 114 Time of addition effect of TGF-B, on a—CD3 + a-CD28-induced IL-2 protein secretion in thymocytes .......................................... 116 Time of addition effect of TGF-[3l on a-CD3-induced IL-2 protein secretion in thymocytes ................................................ 118 TGF-Bl differentially regulates IL—2 protein secretion and [3H]-thymidine incorporation in a-CD3 + a-CD28-activated splenic T cells in a concentration- and time-dependent manner ........... 121 Effect of TGF-[3l on the in vitro AFC response to sRBC .................. 123 Effect of TGF-[3l on the in vitro AFC response to DNP-Ficoll ............ 125 Effect of TGF-[3l on the in vitro AFC response to LPS ..................... 128 Smad2, Smad3, and Smad4 protein expression in mouse splenocytes and thymocytes ................................................................... 130 Concentration- and time-dependent effect of TGF-B, on nuclear expression of Smad3 in mouse splenocytes ......................... 133 Effects of targeted deletion of Smad3 on body and organ weights ....... 136 xiii . I i. I (L C S C S T QC PK. Pf 4.. - .\.. . .III. . . I he a «K; Q1. . . a .. . . .\ . 1% ~ . 0...). .UI . . It... ‘3 I.“ u A‘. “. a‘. . ..‘ . w -* SW ‘.*~ ‘ “. 0.. “. 11 ‘.V. J . . a: . as e as s at 2. ..C .5. .:. ..: .r .. a ...c. ,3. J .2 J .. .. C . "u “w. m.- 4. ~.m vm. v... V n. w... vi. ‘Cu kiwi» A “his. w...» . . . . c .3 ... t . . .. T... . .. C“ :H C- 2.. F1 .LI a... F s. Pr; r. . Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Figure 39. Figure 40. Figure 41. Figure 42. Figure 43. Figure 44. Figure 45. Figure 46. Figure 47. Figure 48. Figure 49. Splenic and thymic cellularity in Smad3-null mice .......................... 139 Inhibition of IL—2 protein secretion by TGF-[3l is attenuated in a-CD3 + a-CD28-activated Smad3-null splenic T cells ................. 142 Inhibition of IL-2 protein secretion by TGF-I3l is attenuated in a-CD3 + a-CD28-activated Smad3-null thymocytes .................... 144 Inhibition of steady state IL—2 mRN A by TGF-[3l is attenuated in activated Smad3-null splenic T cells ....................................... 146 Inhibition of steady state IL-2 mRNA by TGF-[3l is attenuated in activated Smad3~null thymocytes ........................................... 148 Elevated basal IL—4 and IFN-y expression in Smad3-null splenocytes ........................................................................ 15 1 TGF-Bl-induced phosphorylation of Smad2 is not disrupted in Smad3-null splenocytes ......................................................... 154 Growth inhibition by TGF-B1 is attenuated in activated Smad3-null splenic T cells ...................................................... 157 Growth inhibition by TGF-[31 is attenuated in activated Smad3-null thymocytes ......................................................... 159 TGF-Bl inhibits LPS-induced B cell proliferation in Smad3-null splenocytes ......................................................... 160 Reversal of TGF-Bl-induced inhibition of peripheral T cell proliferation by exogenous IL-2 ............................................... 164 A functional role for CAGA sequences in DNA binding activity in the mouse IL-2 promoter .......................................... 168 Smad3 is a component of the transcription factor complex that binds to the proximal AP-l site in the mouse IL-2 promoter. . . . . . 170 CAGA sequences are not essential for TGF-Brinduced binding to the proximal AP-l site in the mouse IL-2 promoter ............................. 173 Smad3 is not essential for TGF-Bl-induced binding to the proximal AP-l site in the mouse IL-2 promoter ............................. 178 Temporal response of TGF-Brinduced protein binding to the proximal AP-l site ............................................................ 180 xiv "_3 c; _- .\ . a; I. F»; vu- 7: . .3 h‘-‘- .u , \ 'u-g. . «I. (ht-v max ‘5‘.- :1 ., ‘ L35; .‘s— . “ - Figure 50. Figure 51. Figure 52. Figure 53. Figure 54. Figure 55. Figure 56. Figure 57. Figure 58. Figure 59. Figure 60. Figure 61. Figure 62. Temporal response of TGF-B,-induced protein binding to a (—____z_x_£__'=-—‘_ CAGA-mutated proximal AP-l proximal site ............................... 180 Effect of TGF-Bl on NFAT transcriptional activity in a-CD3 + a-CD28-activated T cells ........................................... 184 Effect of TGF-B. on NF-KB transcriptional activity in a-CD3 + a-CD28-activated T cells ........................................ 186 Effect of TGF-[3l on NFAT and NF-KB transcriptional activity in a-CD3 + a-CD28-activated Jurkat cells ......................... 188 Concentration-dependent effect of TGF-[3l on ERK MAPK activity in mouse splenocytes ................................................... 191 The effect of U0126 on TGF-Bl-induced Smad3 activation ............... 194 Effect of U0126 and PD98059 on TGF-Bl-induced inhibition of IL-2 protein secretion in activated splenic T cells. . . . . 196 The effect of TGF-[3l on JN K1 and JN K2 activation ....................... 198 Inhibition of the T cell-dependent sRBC AFC response by TGF—B1 is augmented in Smad3-null splenic B cells in vitro. . . . .. ....201 Inhibition of the T cell-independent LPS AFC response by TGF-B, is augmented in Smad3-null splenic B cells in vitro .......... 204 Inhibition of LPS-induced IgM secretion by TGF-B, in vitro is not suppressed in Smad3-null splenic B cells in vitro... . ........207 IgA secretion is not enhanced by TGF-B, in LPS—stimulated Smad3- null mouse splenic B cells in vitro .................................. 209 Putative model for Smad3-dependent inhibition of T cell growth by TGF-B, ......................................................... 220 XV Fr. . ‘1‘ a2M Act ActRI AFC AML2 AP- 1 APC ATF BCA BCS BMP bZIP CaMK CaMKII CaMKIV CaMKK CBFA3 CBP CC14 CD CD28RE LIST OF ABBREVIATIONS a2-macroglobulin activin activin receptor type 1 antibody forming cell acute myeloid leukemia-2 activator protein-1 antigen presenting cell ammonium persulfate activating transcription factor bicinchoninic acid bovine calf serum bone morphogenetic proteins basic/leucine zipper interacting protein calcium calmodulin kinase calcium calmodulin kinase two calcium calmodulin kinase four calcium calmodulin kinase kinase Core Binding Factor subunit a3 CREB binding protein carbon tetrachloride cluster of differentiation CD28 response element xvi CREB CSPD LII «"4 in.“ CD4’ CD8" Con A co—SMAD CRE CREB CSPD dI/dC DPC DNP EBSS ECM EGF ELISA EMSA ERK ETS FKBP GATA GST HBx helper T cells cytotoxic T cells concanavalin A common-mediator Smad CAMP response element cAMP response element binding protein Clontech SEAP Detection CAAT-binding transcription factor polydeoxyinosine/polydeoxycysteine deleted in pancreatic cancer dinitrophenyl Earles’ balanced salt solution extracellular matrix epidermal growth factor enzyme-linked immunosorbent assay electrophoretic mobility shift assay extracellular signal regulated kinase ETS-motif binding protein F K506 binding immunophilin femtomolar GATA-motif binding protein glutathione-S-transferase hepatitis B virus xvii HDAC r d (IQ {‘23 'P.F c—wU ‘ HDAC HTLV- l I-Smad 1g 1x13 iNOS 10 IS I-SMAD JNK LAP LPS LTBP It-ppase MAD MAPK MAPKK MAPKKK MEF’ZC histone deacetylase human T cell leukemia virus type 1 inhibitory Smad interferon immunoglobulin inhibitor of NF-KB interleukin inducible nitric oxide synthase ionomycin internal standard inhibitory Smad c-Jun N—terminal kinase kilodalton latency-associated peptide leucine lipopolysaccharide latent TGF-[3 binding protein lamba phosphatase mothers against decapentaplegic mitogen activated protein kinase mitogen activated protein kinase kinase mitogen activated protein kinase kinase kinase myocyte enhancer factor 2C xviii ELI (I‘d .v , ‘- U I III )in IEC I SCI-CH ICE VIP '12"; \15 N0 1“ .5": MEK MEKK mg “8 MHC MHCI MHCII MIS MKP mRNA NA ng NGF NF-AT NF-KB NK NO MAP/ERK kinase MAP/ERK kinase kinase milligram microgram Mad homology major histocompatability complex major histocompatability complex class I major histocompatability complex class H mullerian inhibitory substance MAP kinase phosphatase millimeter millimoloar messenger RNA na'ive nanogram nerve growth factor nuclear factor of activated T cells nuclear factor for immunoglobulin K chain in B cells natural killer nuclear localization signal nitric oxide negative response element xix W5"? Pal-I NI)“ " 7. Jr W Iiu PLC PIA KC OAP OCT PD98059 PAI-I PDGF PHA PLC PKA PKC pM PMA PMA/Io Pro PT K R-SMAD RASK Rb RCE RSK RT-PCR octamer-associated protein octamer [2-(2’-amino-3’-methoxyphenyl)-oxonaphthalen-4- one] plasminogen activator inhibitor-1 platelet derived growth factor phytohemagglutinin phospholipase C protein kinase A protein kinase C acid dissociation constant picomolar phorbol- 12-myristate- 1 3 acetate phorbol ester plus calcium ionophore; proline protein tyrosine kinase receptor-regulated Smad ribosome associated S6-kinase retinoblastoma Rb control element response element ribosomal S6-kinase reverse-transcriptase polymerase chain reaction XX «.‘O- )3: -14.? r? (/I (.1! (2‘3 l SAD SAPK SARA SB203580 SBE SEAP Ser SIB SIP-l SIF Ski SMAD Sno SNX Spl SP1 SPLN sRBC SRE SRF STAT TAK- l Smad4 activation domain stress activated protein kinase Smad anchor for receptor activation 4-[4-fluorophenyll-2-[4-methyl-sulfinylphenyl]~5- [4-pyridyl]- 1 -H-imidazole Smad binding element secreted alkaline phosphatase serine sis-inducible enhancer Smad interacting protein sis-inducible factor Sloan Kettering institute truncation of C. elegans Sma and Drosophilia Mad ski related novel sorting nexin specificity protein-l simian-virus-40 protein-1 splenocytes sheep red blood cells serum response element serum response factor signal transducer and activator of transcription TGF-B activated kinase-l xxi ‘V (’3‘? TAX TBR TBRI TBRII TBRIII TcR TGIF TGF-B THI TH2 THMC Thr TMB TPA TRAPl U0126 VI-I Tax transactivator protein of the HTLV-l virus transforming growth factor-beta receptor transforming growth factor—beta receptor type I transforming growth factor-beta receptor type II transforming growth factor—beta receptor type IH T cytotoxic T cell antigen receptor TG-interacting factor transforming growth factor-beta T helper T helper cell type I T helper cell type II thymocytes threonine tetramethylbenzidine tumor necrosis factor 1 2-O-tetradecanylphorbol 1 3-acetate TGF-B receptor-associated protein 1 TPA response element TGF-[3 receptor interacting protein 1 1 ,4-diamino-2,3-dicyano-l ,4 bis[2-aminnophenylthio]butadiene vehicle xxii Xbra2 ZEB Xenopus Brachury-2 zinc finger E-box binding protein xxiii an 3“ . \ 7'4 .I" ‘ _“__ d 56-44 - _.‘,. c” O I I.‘ ".4: 4.65. ' - u o 5,3‘ "._-o‘[ H ., .‘h.s\.o . .0.- .-_ I~ C . a-r""‘-g« .L . ._........’su “u: 9%; 3...... .1...‘ 5 A: I’VJ"; -o (4.9:.bg.‘ me v; i; j‘1 :1. . {what ___ '<.. ~x‘_“:LLE 5;. ‘30. ‘/j~l TL ’- - a; II \ INTRODUCTION 1. Significance The overall aim of this research is to acquire a mechanistic understanding of the signal transducing pathways underlying immune modulation by TGF-[3l (transforming growth factor—[34). This research is founded by a large clinical database correlating elevated plasma TGF-Bl levels with liver injury and concomitant immune suppression. In vivo exposure of B6C3F1 mice to modest hepatotoxic doses of CCl4 (carbon tetrachloride) induces potent immune suppression (Kaminski et al. 1990; Kaminski et al. 1989). Separation-crossover-reconstitution experiments with spleen cell subpopulations identified the T cell as the major immune cell population targeted by CCl4 (Delaney et al. 1994). Similar to hepatotoxicity, metabolism of CCl4 is essential for immunotoxicity (Kaminski et al. 1990). However, in contrast to hepatotoxicity, which is mediated through direct actions of metabolites, CCl4-induced immunotoxicity is mediated through an indirect mechanism of action (Delaney and Kaminski 1993; Kaminski et al. 1990; Kaminski and Stevens 1992). Sera isolated from CCl4-treated mice contain elevated levels of bioactive TGF-[3l as determined by the mink lung cell line bioassay (Delaney et al. 1994). These results are consistent with other experimental and clinical evidence suggesting a role by TGF-B, in immune suppression following exposure to a wide range of chemicals and pharmacological agents that induce liver damage and/or diseases, for example, chemotherapeutic agents, immunosupressive drugs, alcohol, and acetaminophen (Abuja et al. 1995; Gutierrez-Ruiz et al. 2001; He et al. 2000; Hori et al. 2000; Inoue et al. 2000; Lemmer et al. 1999; Neuman 2001; Simile et al. 2001; Szuster—Ciesielska et al. Page - 1 "1 PL: remix-5 w“ 5’. . '- "' ' ‘ 9..-". Sis 33»qu 6‘ ‘ L ~ C31) DEE-71’) c? TGF-3 1~ risprtise to IL-2. 4.». I. ’ 1‘ I' A has .uaLIIO‘n d‘:"‘k 3.5... teams a “fa-u- F‘A‘ VD. - ‘ rt.hl\) OI TC; ‘.‘.‘.}~.,I “-4.15. mech’ I an PIN" .A“;'I=‘l . J.'_n \i.‘.‘ .- o {Hg~\_d.> I .".".-3 twu "' 2000; Vodovotz et al. 2000; Yin et al. 2001). In light of these observations, it is conceivable that TGF-Brinduced systemic immune suppression may be a characteristic secondary effect of hepatotoxic compounds. The mechanisms underlying TGF-[3,- mediated immune suppression following CCl4 exposure have not been identified. CCl4 increases IL—2 (interleukin-2.) production in Con A (concanavalin A)-activated splenic T cells (Delaney et al. 1994). TGF-Bl is well recognized for its ability to influence the production of and response to IL—2, a cytokine critical for cell- and humoral-mediated immune responses (Kehrl et al. 1986a; Kehrl et al. 1986b; Kehrl et al. 1986c; Stoeck et al. 1990; Stoeck et al. 1989a; Stoeck et al. 1989b). Recently, Smad proteins have been identified as TGF-[3— specific signal transducing molecules (Massague 1998). The identification of elements that function downstream in the TGF-I3l signaling pathway may provide mechanistic insight towards a better understanding of the diverse physiological immune regulatory properties of TGF-[31. A principal focus of this research has been to elucidate the molecular mechanisms underlying modulation of IL-2 gene transcription by TGF-Bl. H. TGF-B Superfamily The TGF-B superfamily comprises more than 40 structurally related secreted proteins including TGF-BS, activins, inhibins, BMPs (bone morphogenic protein), MIS (mullerian inhibitory substance), and myostatin (Massague 1990; Wrana 1998). While each of these ligands has a broad range of biological activities, TGF-[3 and activin activate signaling through similar receptors, whereas BMP activates a distinctly different set of receptors (Massague 1992). Inhibin, MIS, and myostatin receptor signaling are not well defined. Page - 2 finMn- 12".le isofon‘. 3 6.15: approx; 7‘. soims. TGF-g 30:; as- o'h "1 C)": o ~a5“.\. 54‘“ v 440-:3’)“' ‘k'v u“ . . £ij ~¥.\..x \,;‘ .” O~~ . nammdsu J... .3- “ .~ 5,... . >gua‘l.‘ 1. . ‘A-‘\ ’. . \ K .1 IIKT'\, l -. _ . ““18 H. I" '- “I": Similar to many other growth factors, TGF-BS exists in multiple isoforms. To date, five isoforms have been identified: TGF-[3,, TGF-[32, TGF-B3, TGF-B4, and TGF- I35; each approximately 70-80% homologous to the others (Roberts 1998). Three of these isoforms, TGF-[3,, TGF-B2, and TGF-B3, are expressed in mammalian tissue and are greater than 98% conserved among species. The mammalian isoforms are localized on different chromosomes (human chromosomes 29q13, 1q41, and 14q24, respectively), encoded by distinct genes, expressed in a tissue—dependent manner, and are differentially regulated by numerous growth factors and hormones, including epidermal growth factor retinoic acid, dexamethosone, tamoxifen, phorbol esters, and TGF-BS themselves (Newfeld et al. 1999). The three mammalian TGF-B isoforms are interchangeable in most in vitro assays using primary cell cultures and established cell lines; however, the deletion of any one TGF-B isoform in transgenic animal models is not compensated for by any of the other isoforms and suggests that the TGF-B isoforms are not functionally redundant in viva (Barone et al. 1998; Diebold et al. 1995; Foitzik et al. 1999; Guenard et al. 1995; Kulkami et al. 1993; Li et al. 1999; Shull and Doetschman 1994). In addition, immunohistochemical and in situ hybridization analyses have defined differential in viva TGF-B isoform expression patterns. For example, within the nervous system, TGF-[3l mRNA expression is confined to the meninges and choroid plexus. In contrast, TGF-B2 and TGF-B3 are co-expressed in glial cells and neuronal axons (Flanders et al. 1991; Heine et al. 1987). Accordingly, TGF-B expression in vitro is also differentially regulated. For example, human lung WI-38 and normal rat kidney fibroblasts predominantly secrete TGFB,; however, monkey kidney BSC-l cells and human adenocarcinoma PC-3 cells primarily secrete TGF|32 (Danielpour et al. 1991; Page - 3 linuleu t: .2. 0' v' ‘ ' ’A C. jc‘e Jib C "VJ'TLIIV. I. ,‘u.h.A-fia - . .074 ‘.'-, ~} :3-....C-1...I.:.t P‘-I:-r- -' ,. Le-t-I‘V‘&Ie,\ ‘1‘ I “115 1mm or 5‘4" ,‘ . H .,\-_ EDI ‘8‘; 1‘. r - A $53,, .; ,1. “~‘ “at. 1‘ 3;”? 311‘... \« Iftumfln Jakowlew et al. 1992). TGF-[3I isoforms have nine characteristic cysteine residues; seven of these are conserved in a defined spacing pattern in all members of the TGF-B superfamily. In addition, the mature form of each member of the TGF-B superfamily is a disulfide-linked dimer. In contrast to the sequence of the mature peptide, the pro— segments (discussed below) are only minimally conserved between isoforms (Gray and Mason 1990; Lopez et al. 1992). The TGF-BS are characterized as ‘prototypic multi-functional signaling molecules’ as they function in autocrine, paracrine, and endocrine modes to control a wide variety of processes involving growth, differentiation, adhesion, and apoptosis in a host of cell types (Letterio and Roberts 1997, 1998; Pick et al. 1999; Roberts 1998; Roberts et al. 1988). Members of the TGF-B superfamily are also widely recognized for their regulatory role in formation of the ECM (extracellular matrix). For example, TGF-BS stimulate synthesis of collagens, fibronectin, vitronetin, tenascin, and proteoglycans (Bassols and Massague 1988; Ignotz and Massague 1986; Massague et al. 1992; Pearson et al. 1988), inhibit matrix degrading proteases which include plasminogen activators, collagenase, and stromelysin (Edwards et al. 1987; Kerr et al. 1990; Lund et al. 1987), and suppress activity of the protease inhibitors PAI-l (plasminogen activator inhibitor-1) and TIMP-l (tissue inhibitor of metalloproteinase-I) (Edwards et al. 1996; Edwards et al. 1987). TGF-B, is the predominant isoform that has been implicated in regulating physiological immune and proliferative responses as well as a host of pathological disorders, including fibroproliferation, parasitic and autoimmune diseases, chronic allograft rejection, fibrotic glomerulopathies, and the progression of carcinogenesis Page - 4 ~.‘.' . 4V;'_ ‘th Mums... 111 TGF-f» A. TGF-B is: Item 1‘4 \‘ .S\'-9..-‘, ‘9- ’\' “3 nah-A‘A‘I (Border 1994; Coupes et al. 2001; Hao et al. 1999; Jakubowiak et al. 2000; Letterio et al. 1996; Massague et al. 2000; Roberts 1998; Sporn and Roberts 1991). III. TGF-[3l A. Nomenclature TGF-B1, the prototype of the TGF-B family, was discovered in 1978 (de Larco and Todaro 1978). TGF-[3l was first isolated and purified from supematants of Moloney MuSV—transformed mouse 3T3 fibroblasts, characterized as a growth—stimulating polypeptide, and termed ‘sarcoma growth factor’ (Roberts et al. 1980). The nomenclature ‘transforming growth factor’ was adopted shortly thereafter when it was demonstrated that TGF-[3l stimulates anchorage independent growth of non—transformed rat kidney fibroblasts, a hallmark characteristic of neoplastic transformation (Roberts et al. 1980). B. Transcriptional regulation of TGF-Bl expression The major regulatory domains of the human TGF-[3l promoter are diagrammed in Figure 1. In contrast to TGF-[32 and TGF-B3, the TGF-B, gene lacks the classic TATA box and is characterized by a GC rich region containing several Spl (apecificity protein-t) binding sites immediately upstream of the transcriptional start site. Expression of TGF-B, is up-regulated in response to EGF (apidermal growth factor), jun, fos, src, abl, ras, HTLV-l (human I hell leukemia yirus type I), HMV (human gytomegaloyirus), HBx (hepatitis B yirus); tamoxifen, phorbol esters, TGF— B. and TGF-B2 (Akhurst et al. 1988; Bascom et al. 1989; Colletta et al. 1990; Danielpour et al. 1991; Falanga et al. 1991; Kim et al. 1990a; Kim et al. 1989a; Kim et al. 1989b; Kim et al. 1989c; Kim et al. Page - 5 Figure l. Transcriptional regulation of TGF-[3,. A diagram of the major regulatory domains of the human TGF-[3 promoter. The trans-acting factors that bind to cis-acting elements to regulate TGF-[3l gene transcription are illustrated. Page - 6 A AD_1 oow+ Luz .38 £323 53% £1 .3. .25 xfi F->.E._ $395828“ ES Em .93 £8 .5; .8. in 6.5 mmcmmouco owv+ o_o~+ T 2W- 2.x? owe- nwom- Page — 7 0._.< ooo+ _ :ia 1W“ t". .r r ,u. Irrz‘ .‘ ,.,,.)4 )‘fi'r‘\ Ih‘tL-av “ I A ’ . ‘ p v \On‘fi- I ' '- “Punk-J'd‘ \I- . . ”a. ~J4‘ 1" h} H L1.»Lb\-II.|1\\I‘ ($.15. :99 . I‘M lngsfli r: :zI 1‘ . .‘r..r ‘ : |. In :2"). J W. unit" uni-,2" " IiébA4‘l‘t‘t)n Fl; ' I "‘-.‘ I *1 . ‘ u III ”N... \Is "sat-~0‘ 4 \ ' "" \... .. "IMF I all | ' ‘ 1M9“ 4 4'. ' L‘45.4U“AIZ- Ii‘V‘aI 199‘ V n' I If" '3 ”III N» - , A, - I H F) :4- 0'5 .’ $2 TGF}; )v. ' ,. PM. ~“ng :x “I’- .' ~ v.11" ‘ I. 1990b). Importantly, the ability of TGF-[3l to induce its own expression allows for a sustained expression beyond the initiating stimulus, which has been demonstrated to be of particular significance in wound healing (Kim et al. 1990a). TGF-[3l mRNA is also induced in response to tissue injury, examples include myocardial infarction (Thompson et al. 1988), hypoxia (Henrich-Noack et al. 1994; Khaliq et al. 1995; Klempt et al. 1992; McNeill et al. 1994; Scheid et al. 2000; Williams et al. 1995), hyperglycemia (Daniels et al. 2000; James et al. 2000; Weigert et a1. 2000), bone repair (Joyce et al. 1990), liver regeneration (Fausto et al. 1986; Henrich-Noack et al. 1994; Khaliq et al. 1995; Klempt et al. 1992; McNeill et al. 1994; Scheid et al. 2000; Williams et al. 1995), hepatic schistosome infection (Czaja et al. 1989), carbon tetrachloride-induced liver injury (Armendariz-Borunda et al. 1990; Czaja et al. 1989; Nakatsukasa, 1990 #940; De Bleseret al. 1997; Delaney et al. 1994; Grasl-Kraupp et al. 1998; Greenwel et al. 1993; Jeon et al. 1997; Roth et al. 1998; Shimizu et al. 2001; Simile et al. 2001; Williams and Iredale 2000), acetaminophen-induced hepatotoxicity (Dalu and Mehendale 1996; Icon et al. 1997; Miwa et al. 1997; Neuman 2001; Tygstrup et al. 1996) and hepatitis (Dudas et al. 2001; Flisiak and Prokopowicz 2000; Kim et al. 2000; Liu et al. 1999; Neuman 2001; Zhang et al. 1999b). Rb (tetinohlastoma) protein binding to the RCE (tetinoblastoma control clement) attenuates TGF-I3l gene expression (Udvadia et al. 1992). C. Secretion of TGF-[3l as a latent complex TGF-Bl is synthesized as a biologically inactive 390—amino acid dimeric precursor complex containing the C-terminal mature TGF-[3 and its N-terminal pro-domain, LTBP (latent TGF-B hinding protein) (Figure 2). The LTBPs are members of the Page - 8 Figure 2. Diagram of the biologically inert latent precursor complex of TGF-B1. TGF-B, is synthesized as a biologically inactive 390-amino acid dimeric precursor protein. The mature form of TGF-BI is located within the carboxy-terminal domain. The latent TGF-B binding protein is contained within the amino-terminal domain. Activation to the mature biologically active form of TGF-I3l requires multiple sequential proteolytic cleavages. Page - 9 LAP Dimer H ,3... LAP Dimer I GHQ—TGF-B Dimer Hinge Region RN ;§ RN NS RV XV RR on r: a) '33 .2 g H L 8202 so}; am. Page - 10 Domain ITS? Iibriilin | secrets: of TI" EDI Lister... The TI? :I'Cz’ss to gcn; IfoTEL‘fs .1 N. l w . .. It." 0; LX135 {INN A n .‘ 4 . xx- A I“ 4K1“ Y‘ '~.-- 4 - ‘ . ‘ I .h 9‘. l4: ‘0 H p. LTBP/fibrillin family of ECM glycoproteins and are important for the folding and secretion of TGF-B, as well as for the sequestration of latent TGF-I31 complexes to the ECM (Lawrence 1991). The TGF-[3l precursor protein is a pre-pro-peptide and undergoes a two-step process to generate bioactive TGF-I3l which is able to associate with the TGF-B signaling receptors (Lawrence 1991; Munger et al. 1997; Nunes et al. 1998) (Figure 3). The first step of this process is a proteolytic cleavage that eliminates the N-terminal hydrophobic LTBP peptide (amino acid residues 1 to 29) to generate a pro-peptide molecule. The second step is a subsequent proteolytic cleavage that removes the LAP (latency- associated peptide) (amino acid residues 30 to 278) to generate a mature active peptide (amino acid residues 279 to 390). The 74 kD LAP must be cleaved prior to TGF-B, binding to the TBR (TGF-[3 teceptor), and thus is critical for TGF-[3l activity. LTBP (125 to 160 kD), on the other hand, permits association with the ECM and functions to regulate localized storage of the inactive complex within the ECM (Munger et al. 1998; Munger et al. 1997; Saharinen et al. 1999; Taipale et al. 1996). Four highly repetitive splice variants of LTBP (1-4) have been isolated. TGF-[3I bioavailability may be differentially regulated in a tissue—specific manner, in part, by the formation of variable pre-pro-precursor complexes. For example, LTBP-1 is predominantly expressed in the heart, placenta, lung, spleen, and kidney. LTBP-2 is preferentially expressed in the lung, skeletal muscle, liver, and placenta. LTBP-3 and LTBP-4 are found primarily in the heart, small intestine, and ovaries. Interestingly, LTBP-1 additionally functions as a chemoattractant (Saharinen et al. 1999; Taipale et al. 1996). Page-11 Figure 3. Stepwise activation of latent TGF-[3l precursor to its biologically active form. Illustrated are the multiple sequential post-translational regulatory steps associated with the activation of TGF-Bl in vivo. Intracellular / Extracellular Page - 12 cozm>5u< .m \ végsl mmmmdmm .v AVID 5.3.5034. 5me .m cozmbwm .N Lmdzzmumbxm /' 5.5%.: .838”. .o mcsmcma A 285% .F 53.88%; Page - 13 "_ .‘r \' 33.4) rer‘e ‘ are?" ‘_n‘ ‘ h. .!.:."1 “(hi ' C ‘ I ’3'! ‘h ' ‘1 Tg5io\‘l e. 4“. R3618 199 5‘ a .'\ VA. r“.;" I-A r?- ‘4 .kgu‘e L);- 293*),. TGF TGFfi-TGF-L x» ‘ ,' sing ‘ - I tJ.i\u.1)n 01 UL etlfrfifiljf‘m -' v ~ -U C. 9.. ‘I.O H- ‘\4 g 0 Kloien‘u iB' In addition to associating with LAP and LTBP, active TGF-B, can readily form complexes with (12M (oc_2-macroglobulin), biglycan, decorin, type IV collagen, fibronectin, and thrombospondin. a2M is probably the most critical carrier of circulating TGF-B, and plays an important role in rapidly clearing TGF-[3I from the bloodstream via binding to the a2M receptor (Munger et al. 1997). The most concentrated source of TGF-B, is the platelet alpha granules (Centrella et al. 1986). Osteoblasts and other bone cells represent additional major storage sites for TGF-Bl. Activated lymphocytes, macrophages, and neutrophils also secrete an abundance of TGF-[3l (Assoian et al. 1984). D. Activation of latent TGF-[5l Activation of the latent complex is tightly controlled and represents a critical regulatory step in TGF-[3l activity (Koli et al. 2001). A number of factors, including acidic pH, nitric oxide, plasrnin, cathepsin, thrombospondins, and subtilisin-like pro- protein convertases function to activate latent TGF-[3l (Barcellos-Hoff 1996; Chenevix- Trench et al. 1992; Godar et al. 1999; Harpel et al. 2001; Hugo et al. 1999; Letterio and Roberts 1998; Quan et al. 2001; Roberts et al. 1992; Vodovotz et al. 1999). The bioactive form of TGF-[31 is a 25 kD disulfide-linked polypeptide dimer (Massague 1990). TGF-B, is usually produced as a homodimer, ie., TGF-BpTGF-Bg however, TGFI3.°TGF—[32 heterodimers have been identified (Ogawa et al. 1992). A differential function of heterodimers relative to homodimers has not been established. 1. Physiochemical activation The association between TGF-Bl and LAP is via an electrostatic interaction and is readily disrupted in vitro by acid, high temperature, or chaotropic treatment (Brown et al. 1990; Lawrence 2001; Lawrence et al. 1985). Physiologically, it Page - l4 [1.5 km \pCCL .’ _ 1" . 011239.16 it, .. ~‘- 0 .. l 6' 1".\‘ Did f. 11135;» $3.“;er IL'II t. -\.Y .1 i >4 tinc‘r'ZLu ‘S. «34' to. Izbelio;\‘ \3A\ “5 R81“ '47 V3" >3"! . «Await?! ft $535,103 in $015.33th er ‘5 , “e [0 \IEQVQ 7-. (”‘21: \rurg er a! has been speculated that acidic microenvironments that occur within the bone and at sites of tissue repair may provide localized sites for TGF-B, activation in vivo (Salo et al. 1997). Interestingly, irradiation increases the concentration of active TGF-B, in mouse mammary tumors (Barcellos-Hoff et al. 1994). Although not confirmed, it has been hypothesized that reactive oxygen species generated during irradiation induce redox- mediated activation of latent TGF-B, (Barcellos-Hoff et al. 1994; Barcellos-Hoff and Dix 1996; Reiss and Barcellos-I-loff 1997; Vodovotz et al. 1999). 2. Enzymatic activation and protein interactions Proteolysis targets the degradation of the LAP propeptide and the subsequent release of active TGF-13,. Appropriate protease inhibitors impair TGF-[3l activation in vitro (Antonelli-Orlidge et al. 1989; Huber et al. 1992). Numerous proteolytic enzymes, including mast cell chymase, leukocyte elastase, and plasmin are able to cleave LTBPs and release latent TGF-B, from the ECM (Olofsson et al. 1995; Taipale et al. 1994). TGF-B, also undergoes proteolytic activation via deglycosylation of LAP (Miyazono and Heldin 1989). Additionally, in certain cells, i.e., monocytes, endothelial, hepatocytes, smooth muscle cells, and pericytes, secreted latent TGF-[3l can also be activated at the cell surface via direct binding to the mannose-6-phosphate receptor, which in turn induces cleavage of the LAP (Ariazi et al. 1999; Godar et al. 1999; Rudd et al. 2000). Importantly, the proteolytic release and activation of stored TGF-[3l can generate a rapid signal important for repair of tissue damage and localized activation of the immune system in the absence of gene transcription (Koli et al. 2001; Lyons et al. 1990; Saharinen et al. 1999; Taipale et al. 1996). Page - 15 n ,.,J ' f)? :§:x‘;~n\- it“ y r- I so-” 1"] " trail-tn Au - Afr“. .. .’ kLJl‘tL‘. (I L” . . 1'”. J} 4"41"‘.Jr- "“h.’y.\ “:1 PM ' M. ”a . x b \.. . H 'I' v ‘§”‘t".3v4 \ ¢ k‘“ V \ 3. Hormone-induced activation Steriod hormone superfamily members, including estrogens, anti- estrogens, retinoids, androgens and synthetic progestins, have been demonstrated to effectively activate TGF-BS in a number of in vitro cell models (Boulanger et al. 1995; Colletta et al. 1990; Djurovic et al. 2000; Fisher et al. 1992; Click et al. 1991; Harpel et al. 2001; Knabbe et al. 1987; Koli and Keski-Oja 1993). The biochemical mechanism(s) by which TGF-B is activated by hormones is unknown and is a current topic of intense research. E. TGF-[3l Receptors In contrast to most growth factors, which signal through tyrosine kinase receptors, TGF—B, binds to multimeric receptor complexes with intrinsic serine/threonine kinase activity. TGF-B signaling is mediated through ligand-induced hetero-oligomeric receptor complex formation, presumably between two molecules each of type I and type 11 receptors. Constitutively active transmembrane TBRII (TGF-fl receptor type II) homodimers recruit and phosphorylate transmembrane glycoprotein TBRI (TGF—fl receptor Type I) homodimers at the GS domain located upstream of the serine/threonine kinase domain (Figure 4). Activated TBRI is responsible for propagating the signaling via phosphorylating downstream cytoplasmic molecules, including Smads, MAPKs (Mitogen Activated Protein fiinases), PLC (_Ehospholipase Q), and PKA (Protein _Iginase A). TBRS (TGF-fl receptors) were initially identified via crosslinking of 125I-TGF-B to cell surface proteins. Based upon differing electrophoretic mobility of 125I-TGF-~B to crosslinked complexes on non-denaturing polyacrylamide gels, a total of nine TBRs (type Page - 16 Figure 4. Schematic representation of the type I and type II TGF-Breceptors. The conserved GS motif with its SGSGSGLP sequence is characteristic of the TBRI and structurally precedes the kinase domain. Page - 17 TBRII TBRI Page -—18 Amino-terminal tail Cysteine cluster motif Cysteine box Transmembrane domain GS motif Serine/threonine kinase domain Serine/threonine rich carboxy-terminal tail LH. 111. IV c L70“: :0 5i up): their a receptor pr: mi TERI} u l l 1! 53- f. ‘ LSI‘ .r:\ Pin; Li 3 RV.) I n \X,.K‘\.‘\‘L \. TERIH 1:. d: recent In H“. I, II, III, IV, V, VI, VII, VIII, and IX) have been identified; these receptors are now known to be cell type— and ligand-specific (Attisano et al. 1994; Wrana 1998). Based upon their abundance, wide distribution, high affinity binding, and correlation between receptor presence and TGF-[3 responsiveness, three of these receptors, i.e. TBRI, TBRH, and TBRIII were originally identified as putative signaling molecules. In vitro somatic cell hybrid complementation analyses of TBR mutant cells have confirmed that both co- expression of TBRI and TBRII is essential for TGF-[3,-induced signaling; in contrast, TBRIH is devoid of signaling by TGF-B (Letterio and Bottinger 1998). However, more recent in vivo studies employing a dominant negative TBRII transgenic mouse model have demonstrated that TGF-B, may actually mediate some biological responses through direct activation of either TBRI or TBRIH GBottinger et al. 1997). Additional studies are needed to clarify TBRIII signaling properties. Several highly homologous, yet different subtypes of TBRI and TBRII exist, and their ligand binding properties are highly dependent upon the TGF-B superfamily member to which they bind. Distinct receptor-associated signaling pathways may mediate separate TGF-B responses. In addition, the sequential activation of TBRII complexes prior to TBRI complexes allows for combinatorial diversity; i.e. different type H receptors can pair with different type I receptors, so that a given ligand may conceptually generate varied biological responses (Massague 1996). TGF-Bl forms a high affinity heteromeric complex with only one type of TBRII and seven types of TBRI and TBRIII. A brief structural and functional overview of the mammalian TBRs is provided below. Page - 19 a-‘ n"“-" i:‘ I:-.'- Hundl ~‘ “ana-3 .3 _ 5‘ SLQHJ5 r5 ‘4‘ .;:‘IV".’.‘. 4 - iii...“\u tK—‘r n- v 1.2.1:}. H {a A. , . . “Eh-IIIMIJ’V‘ \ Y‘N‘J o. I, 4 _q. “-th ‘nd'L t 5Lflmn~¢ ‘3" ,‘ F t “. "“41 LICK]; ..,-( l Aim““I)\h. V 1 .1” a" sliuug‘n] St.I “Mir.“ "‘ul. ’ in». \- ‘\N‘\e 1. TBRII (TGF-fl Receptor Type 11) TBRI and TBRII mediate TGF-B, signaling. Each of these receptors are transmembrane serine/threonine kinases with a single transmembrane domain (Figure 4). Structurally, TBRII is a 75- to 85-kD glycoprotein composed of a 136-residue hydrophobic N-glycolsylated extracellular domain, a single transmembrane domain, and a 376-amino acid cytoplasmic domain comprised of a serine/threonine kinase domain and the carboxy-terminal tail (Lawler et al. 1994; Lin et al. 1992). Multiple extracellular cysteine residues are indicative of extensive disulfide bridge formation and folding that is required for ligand binding. In the absence of TBRI, TGF-[3l binds TBRII with high affinity. Heteromeric complex formation between TB RI and TB RII does not demonstrably alter TGF-B, binding affinity for TBRII. TGFs—B are the only known ligands that bind to TBRII. Independent of ligand binding, TBRH constitutively resides as a homodimer; homomeric interactions are partly due to ability of the extra- and intra- cellular domains to interact with each other (Chen and Derynck 1994; Henis er al. 1994). Auto-phoshorylation of TBRII has been localized to Ser-549 and Set-551 (located in the carboxy-terminus) and Ser-223, Ser-226, and Ser—227 (located in the juxtamembrane domain) (Souchelnytskyi et al. 1996). TBRIIs auto-phosphorylate on serine residues in viva, but on serine and threonine residues in vitro (Lin et al. 1992). The physiological significance of this differential TBRII phosphorylation pattern is not known. 2. TBRI (TGF-fl Receptor Type 1) Similar to TBRII, TBRI is a transmembrane serine/threonine kinase receptor, and resembles TBRII in sequence homology and structure. TGF-B binds to seven subtypes of TBRI, i.e., TBRI, ALK-l, ALK-S, ActRI, TSR-l, TSR3, and TSR4. Page - 20 “It. ,UNC :Uni ‘* 3.“ ed 0 ‘1. URI n. :.c 23‘ ill; '. I-snrl "1‘ 1r: \‘s..h:l. 4......“de h‘ V'dfifié n».- o Leuw‘.“ 4‘ Lil {11:5 so. \dL h‘\e‘ \ TBRI is a 50 to 60 kD protein that is comprised of a 101-amino acid hydrophilic extracellular domain and a 355-amino acid cytosolic serine/threonine kinase containing motif. In contrast to TBRII, TBRI has a shorter extracellular domain, a truncated intracellular carboxy-terminal extension following the kinase domain, and the highly conserved GS domain immediately proceeds the serine/threonine kinase motif (Figure 4). TBRI does not bind TGF-B in the absence of TBRH hetero-dimerization; however, the presence of TBRI is essential for TGF-B responsiveness, thus reflecting its critical role in signal transduction. The three-dimensional inactive conformation of the TBRI is maintained by interactions of multiple cytoplasmic motifs involving the GS domain, the N-terminal tail associated with ATP-binding, and the activation loop. TBRI is converted to its active confirmation via phosphorylation within the GS domain by TBRII. The L45 loop of the amino terminus of TBRI protrudes out from the kinase domain allowing for direct interaction with intracellular substrates, e. g. Smads. 3. TBRIII (TGF-[3l binding proteins Type III) TBRIII is a 280 to 330 kD transmembrane proteoglycan that binds TGF-B, with high affinity. TBRIII is presumably devoid of intrinsic signaling capabilities and is thought to function primarily to recruit TGF-[3l to TBRII at the cell surface; however, as discussed earlier, these results are somewhat controversial. F. TGF-l3l receptor binding proteins The FK506-binding immunophilin, FKBP12 binds to TBRI via a Leu-Pro sequence located in the GS domain and functions to inhibit spontaneous, ligand- independent activation by TBRI. TRAP] (TGF-B receptor-associated protein 1) Page - 21 associates \k 11h nedzizor. Sma}- TRlPl on!) my dilation of TRA Signing m rep mm mm ; SNX‘S 1: :0 form hetero." er al. 2001 .1. \ ye: understmd receptors inch LEIGH, associates with inactive heteromeric TBR complexes and is released upon activation of ligand binding (Verrecchia et al. 2001a). TRAPl also interacts with the common mediator, Smad4, in a ligand-dependent manner (Wurthner et al. 2001). It appears that TRAP] only modestly stimulates TGF-B signaling in functional assays in vitro; however, deletion of TRAPl diminishes the interaction of Smad4 with Smad2 and inhibits TGF-[3 signaling in reporter gene assays (Wurthner et al. 2001). These data suggest a model in which TRAPl permits sequestration of Smad4 into the vicinity of the receptor complex and facilitates its transfer to the receptor-activated Smad proteins. SNX6 (sorting nexins Q) a member of the SNX superfamily has also been shown to form heteromeric complexes with TBRs and plays a role in receptor trafficking (Parks et al. 2001). While the functional significance of this protein°protein interaction is not yet understood, other members of the SNX superfamily play a role in trafficking a host of receptors including EGF (epidermal growth factor) and PDGF (platelet derived growth factor). G. Biological Functions of TGF-[3l TGF-[31 plays a critical regulatory role in a host of physiological and pathological events, including cell proliferation and differentiation, angiogenesis, extracellular matrix remodeling, wound repair, bone formation, inflammatory processes, and immune homeostasis (Bonewald 1999; Hata et al. 1998; Letterio and Roberts 1997; Massague et al. 2000). The precise nature of the response of a particular cell to TGF-[3, is highly dependent upon cell type, differentiation, activation status, cell cycle, and the cytokine/growth factor/hormonal milieu in the surrounding micro-environment (McCartney-Francis et al. 1998; Wahl et al. 2000). TGF-B, also plays an important role Page - 22 in a host of neur=.»dsgencrati Kept i999; C 02 P.2d 120mm 20 H. '1 TGF-B. ieuix'ocytes ‘LC Forcrimple. I hide TNF~( Adiiuonalh'. '. uric. adhesn 56mph} Che ,. '1 in C} cle. TC 131 respbnse to that of an i' :33???“ ‘ ' Mame ICII a: . 4.13 I630} :1 in a host of pathological disorders, including autoimmunity, carcinogenesis, neurodegeneration, impaired wound healing, parasitic diseases, and fibrosis (Branton and Kopp 1999; Connor et al. 1989; Dudas et al. 2001; Fausto et al. 1991; Lawrence 1996; Prud'homme 2000; Scheid et al. 2000). H. TGF-[31 and Immune Modulation TGF-B, plays an essential role in maintaining immune homeostasis (Letterio and Roberts 1998; Wahl er al. 2000). TGF-[3l is secreted by all leukocytes and TBRs have been identified on all lymphoid cells (Letterio and Roberts 1998; McCartney-Francis et al. 1998). The immune modulatory effects of TGF-Bl are multiple and target hematopoesis, proliferation, differentiation, and function of all classes of mature leukocytes (Letterio and Roberts 1998; McCartney-Francis et al. 1998; Wahl et al. 2000). For example, TGF-[3l modulates the production and activity of numerous cytokines which include TNF-a (tumor necrosis factor-g), IFN-y (interferon-x), IL-1, IL-2, and IL-6. Additionally, TGF-B, regulates macrophage-induced production of superoxide and nitric oxide, adhesion molecule expression, Ig (immunoglobulin) synthesis, monocyte and neutrophil chemotaxis, and activation of lymphoid cells and their progression through the cell cycle. TGF-B, modulates proinflammatory as well as immunosuppressive responses in response to tissue injury. Femtomolar concentrations of TGF-B, that are present at the onset of an inflammatory response, act as a chemoattractant to elicit proinflammatory responses (Wahl et al. 1987). As inflammatory responses progress, circulating levels of bioactive TGF-Bl rise. Substantially elevated plasma levels of bioactive TGF-[3l present at the resolution phase of the inflammatory response act to suppress the function of Page - 23 {IxmfiwS 1mm neutephil‘ ‘M‘ TGF-B in man CYBL‘é X S «szr‘\")'“; >34 h. we.» lki AZ\U {1253 l and II warms. 3r.- 1 . . {1 'fiuh .3...,..01d 0:24: 5 I I - [Put .~ ."fkaln u ._I\ "flak .1x I \ Ame n1]: .3 3M4 ardutln.) ‘ e the ef 3:2» numerous immune cells, including lymphocytes, macrophages, monocytes, and neutrophils (McCartney-Francis et al. 1998; Wahl 1992; Wahl et al. 1993). 1. The TGF-[31 null mouse model The TGF-[3l null mouse probably best exemplifies the importance of TGF-I3l in maintaining immunological balance. Targeted disruption of TGF-[3l on a C57BL/6 X SV129 background produces an autoimmune-like phenotype that is characterized by enhanced circulating pathogenic IgG autoantibodies to nuclear antigens and elevated glomerular deposition of immune complexes. Increased expression of MHC class I and II molecules, elevated numbers of circulating immature granulocytes, monocytes, and platelets, and a decreased number of B220+ B cells in peripheral lymphoid organs and bone marrow are also observed in TGF-l3I null mice (Christ et al. 1994; Geiser et al. 1993; Kobayashi et al. 1999; Kulkarni et al. 1993; Letterio et al. 1996). TGF-[3l null mice succumb within four weeks of age to a rapid, progressive multi- organ infiltration of lymphocytes. Neither thymic cellularity nor the relative distribution of cells within thymic cell subpopulations are altered in asymptomatic newborn mice; however, a significant reduction in double positive CD4“, CD84' T cell pregenitors is evident as systemic inflammation progresses (Letterio et al. 1996). 2. T cells The role of TGF-B, in the regulation of T cell responses has been perplexing, possibly because it is dependent on the type of T cell being regulated and the cytokine milieu in the surrounding microenvironment. A comprehensive picture entailing the effects of TGF-B, on T cells is complex because TGF-B, also indirectly targets T cells by regulating the function of antigen-presenting cells. Evidence Page — 24 stems tram the . 2.11 mace is i} r. :niizmm and z T1}‘r:;h.x_\!es ' Sregeptorx ICIa receptor '15 five 355'. Aidm receptors by (1? "ii‘vaiion (3;, CST-«mm and 65.317356 cell T?Cifi\g-; establishing a critical role for T lymphocytes in TGF-B,-induced immune homeostasis stems from the observation that the early onset of multi-organ inflammation in TGF—B, null mice is lymphocyte-mediated (Diebold et al. 1995). In addition, inflammatory cell infiltration and animal lethality are concomitantly reduced with in vivo depletion of CD4+ T lymphocytes in TGF-B, null mice. Although T cells express a lower number of TGF- Breceptors relative to most cells studied, the affinity for TGF-B. binding to the TGF-B receptor is five to ten fold higher in T cells (Kehrl et al. 1986c; Massague and Like 1985). Additionally, T cell activation increases the number and affinity of TGF-B receptors by approximately six- and three-fold, respectively. TGF-B, modulates T cell activation (discussed below), attenuates T cell apoptosis (Cerwenka et al. 1996; Cerwenka and Swain 1999; Genestier et al. 1999a; Genestier et al. 1999b), and induces G,/S phase cell cycle arrest (Saltis 1996). Transgenic mice expressing a truncated dominant negative TBRII under the control of a CD4 promoter construct lacking the CD8 silencer have been developed as a model to investigate the direct effects of TGF-Bl on T cells in viva (Gorelik and Flavell 2000). These mice survive beyond five months of age, but develop a phenotype with many features overlapping TGF-[31" mice, including a multi-organ perivascular infiltration of mononuclear cells, circulating autoantibodies, and renal glomeruli immune complex deposition. T cells from these transgenic mice are refractory to inhibition by TGF-[3, and spontaneously differentiate into type 1 or type 2 cytokine secreting cells, with CD4+ T cells capable of secreting IFN-y and/or IL—4 in vitro (Gorelik and Flavell 2000). These results demonstrate the importance of TGF-B in maintaining tolerance in T cells. In addition, these results also illustrate that the maintenance of B cell tolerance to Page - 25 molecular m; needed in 0rd ‘ I .‘.-v,. . .eh prolxer. emesszon. : sxszemic aut. are Roes In 3T ce‘l-in i~ i K. - Dur: «o') ill A “~n . f" ra- ‘d‘Tak 2‘34 ":~K. 5.; ‘\ self-antigens is dependent on normal TGF-B signaling in T cells. Investigation of the molecular mechanisms of how TGF-B, either stimulates or inhibits T cell function is needed in order to understand these modulatory effects at the cellular level. 3. B cells TGF-BI elicits a broad range of effects on B cells, including inhibition of cell proliferation, antibody secretion, antigen receptor and MHC class II molecule expression, and Ig isotype class-switching (Letterio and Roberts 1998). In addition, systemic autoreactivity, B cell hyper-responsiveness, IgA-deficiency, and elevated serum IgG levels in transgenic mice lacking a functional B cell TGF-B type II receptor (Cazac and Roes 2000) provides strong evidence that TGF-B, also controls B cell homeostasis in a T cell-independent manner. During the course of B cell differentiation, B cells undergo a process called isotype class switching, in which the initial synthesis of IgM antibody is converted to IgD, IgG, IgE, or IgA. During this isotype switching process, the immunoglobulin heavy chain constant region undergoes rearrangement, while the immunoglobulin light chain and the variable heavy chain remain unchanged, to generate a change of B cell effector function. Antigen specificity is not altered with isotype class switching. Class switching occurs via an intrachromosomal recombination event that joins variable region genes with a gene for a constant region to form a functional heavy chain gene for the IgD, IgE, IgG, or IgA immunoblogulin. This process appears to be directed by transcription of un— rearranged or germline transcript of the heavy chain constant region gene before switch recombination. Page - 26 TGF—B IgA isoryps’ CI Slech; 19921. iniuctron of ‘- immunoglobu Recently. it h. recepzor rexpm statute gem crack :(m I'- TGF-B, induces transcription of the germline Iga constant region gene and directs IgA isotype class switching (Cazac and Roes 2000; Pardali et al. 2000a; Sonoda et al. 1992). Mechanistically, TGF-B, augments binding of the transcription factor AML2 (acute _myeloid leukemia-2) to the regulatory region of the IgA germline promoter and induction of ‘sterile’ transcription of the locus that is known to play a critical role in immunoglobulin class switching (Hein et al. 1998; Lorenz and Radbruch 1997). Recently, it has been demonstrated that Smad3 and Smad4 directly bind to the TGF-B receptor response element within the IgA promoter and cooperate with AML proteins to stimulate germline IgA transcription (Pardali et al. 2000a; Park et al. 2001; Zhang and Derynck 2000). 4. THl/THZ development The exact regulatory role of TGF-Bl in THI/THZ development remains somewhat controversial. Dependent upon the experimental system, TGF-B, has been shown to favor selectively either THI or TH2 development (Gorham et al. 1998; Heath et al. 2000; King et al. 1998; Lee and Rich 1993; Ludviksson et al. 2000; Nagelkerken et al. 1993; Schrnitt et al. 1994a; Schrnitt et al. 1994b; Swain et al. 1991a; Swain et al. 1991b; Taylor et al. 2000; Thorbecke et al. 2000). In addition, a difference in TH1/TH2 development among mouse strains has been proposed to account for somewhat varying phenotypes of TGF-Bf" mice on different strain backgrounds (Gorham et al. 2001). 5. TGF and IL-2 One of the first demonstrated immune modulatory effects of TGF-[3l on isolated lymphoid tissue inhibition of T cell growth. TGF-B. attenuates the production of Page - 27 1990 s. an: On-[bc-OII". ‘\ IL- 936.1. A Cell- ‘5' [17118 its IL—2, down-regulates the high affinity IL-2 receptor (Gorham et al. 1998; Ruegemer et al. 1990), and disrupts IL-2 signaling downstream of the H.-2 receptor (Bright et al. 1997). On-the-other-hand, under the appropriate experimental conditions, TGF-Bl also enhances IL-2 production (Cerwenka er al. 1994). IV. Transcriptional regulation of the IL-2 gene following T cell activation through the T cell antigen receptor IL-2 is a 15 kD glycoprotein that functions as a critical autocrine/paracrine growth factor for a variety of immune cells including T cells, B cells, NK cells, and macrophages (Frey et al. 1987; Lozano Polo et al. 1990; Nelson et al. 1992; Smith 1992). IL-2 is produced predominantly but not exclusively by activated helper T cells. Activated CD4+ T cells is the primary source of IL-2. NK (natural killer) cells also produce modest amounts of IL-2 (Lozano Polo et al. 1990). IL-2 functions as a major growth factor for T cells, thus its regulation is a central control mechanism for T cell clonal expansion and subsequent cell-mediated and humoral immune responses (Belardelli 1995; Gomez et al. 1998; Swain et al. 1991a; Waldmann et al. 1998). A. Control of IL-2 Transcription IL-2 mRNA transcripts are undetected in naive resting T cells, but are up- regulated within 30 minutes of CD4+ T cell activation (Paul and Seder 1994; Seder et al. 1994). In vitro nuclear run-on assays, DNase I hypersensitivity assays, and in vivo footprinting analyses are consistent in implicating a transcriptional mechanism for IL-2 mRN A expression. Steady-state levels of IL-2 mRN A in activated T cells are maintained between a balance of gene transcription and mRNA degradation. Following T cell Page - 28 stimuatror. IOIIOuiig I '9 1 egzrne of . 2531:? Iran \C ¥ 1 I sezfadatlor 3“ ‘. ‘1.H‘nc ' .‘ k SYL,‘ \ stimulation, maximal mRNA levels are achieved approximately four to eight hours following T cell activation and return to background levels within 24 hours. The rate of decline of IL-2 mRNA expression is normally greater than the rate of decline of IL-2 gene transcription implicating a putative protein-dependent mechanism of IL-2 mRN A degradation. In support of such a mechanism, IL-2 mRNA accumulation in the absence of a change in IL-2 transcription is observed when activated T cells are treated with cyclohexamide to inhibit protein synthesis (Jain et al. 1995; Oldham et al. 1989). Furthermore, actinomycin D treatment prolongs IL-2 mRNA stabilization suggesting that mRN A degradation is regulated at the level of RNA synthesis (Jain et al. 1995). Continual T cell stimulation is required to maintain IL-2 transcription, and a long- standing question of how long IL-2 transcription proceeds under conditions of continual T cell stimulation remains unanswered. In viva footprinting analyses suggest protein binding at transcriptional regulatory elements for up to eleven hours following T cell stimulation (Jain et al. 1995). Based on this, it is tempting to speculate that transcription may proceed as long as antigen or appropriate stimulation is present, and that the decline in transcription may reflect a diminution in the level of activated transcription factor binding through secondary mechanisms e.g., recruitment of negative regulatory proteins and/or chromatin remodeling). B. Signaling cascades through the TcR that regulate IL-2 gene expression The intracellular signaling events that mediate IL-2 transcription have been studied extensively and implicate activation of multiple signaling cascades within Page - 29 seconds of. ‘ ll r 353‘s 3.11011 “162515 seconds of ligation at the TcR (I pell antigen receptor) complex. The majority of mature T cells express a transmembrane disulfide-linked 0t,B TcR heterodimer. Complete T cell activation requires a two-signal stimulation (Peterson and Koretzky 1999) (Figure 5). The first stimulus is provided by presentation of a processed antigenic peptide by an APC (antigen presenting cell) to a T helper cell through the TcR in context with an MHC class II molecule (Davis et al. 1989; Evavold et al. 1993; Farber 1998; Kane et al. 2000; O'Shea 2000; Saito et al. 1995; Sette et al. 1995; Shores and Love 1997; von Boehmer et al. 1989; Wange and Samelson 1996; Wilson and Garcia 1997). The most critical second costimulatory signal is provided through the interaction of members of the B7 family, B7-1 (CD80) and B7-2 (CD86), found on the surface of antigen presenting cells (i.e. B cells, dendritic cells, and macrophages) with the CD28 molecule located on the surface of the helper T cell (Chiang et al. 2000; Holdorf et al. 2000; Hombach et al. 2001; Powell et al. 1998; Seder et al. 1994). The cytoplasmic tails of the TcR are devoid of signaling motifs. Therefore T cell activation through the TcR is dependent on translocation of cytosolic tyrosine kinases to the TcR complex upon ligand binding to initiate intracellular signaling (Clavreul et al. 2000; Hennecke and Wiley 2001; Kim et al. 2001). TcR ligation activates the Src family PTKs (protein tyrosine kinase) lck and/or fyn, which associate with the CD3 and TcR§ subunits to initiate a rapid phosphorylation and activation of the Syk family PTK-ZAP-70. The end result of TcR-coupled tyrosine kinase activation is the initiation of numerous signaling pathways within the cell including, intracellular Ca2+, PKC (protein kinase C), PKA (protein kinase A). and phosphatases. These activated cascades ultimately culminate in the activation of Page - 30 Figure 5. Two-signal model for T cell activation. IL-2 protein synthesis requires a two-signal interaction between a T cell and an APC. Signal one is mediated through the interaction of the TcR/CD3 complex with the peptide antigen/MHC complex that leads to increased T cell surface expression of CD40L. CD40L interacts CD40 on the APC to induce expression of B7 on the surface of the APC. B7 interacts with CD28, a co- stimulatory receptor on the T cell surface. TcR/CD3 and CD28 signals are essential for sufficient IL-2 gene expression for complete T cell activation and prevention of T cell anergy (Modified from Weiss, 1999). Page - 31 AFC T cell IL -2 w. ._.. aovcu ovou m moU + UI<< \muh a £38m Nu: :8 ._. Un_< Page - 32 mm: gene ex t_i ‘ fiOfitC\pAR\:0n. regiictnr} regiu grrcmlitm d! espectnelj IR‘ ups'seam of t. e [L] gene tram. Tie mouse and {EgI‘II‘r l.\'0\ 4k t f AS ShO'.‘-. ”In P'UHZN-a ' r ,. " "~T8n2a* kiwi 'e ”LUIS.~ cytokine gene expression, including IL-2, to regulate T cell activation, proliferation, clonal expansion, and effector function. C. Regulatory elements of the IL-2 gene IL—2 gene expression is stringently regulated at the transcriptional level via binding of several nuclear trans-acting factors to cis-acting elements within the regulatory region of the IL-2 promoter. Post-translational modifications including glycosylation and phosphorylation also regulate IL-2 stabilization and secretion, respectively (Robb et al. 1984; Rooney er al. 1995). The first 300 bp region directly upstream of the transcription start site of the IL-2 promoter is essential for regulation of IL-2 gene transcription. (Rooney et al. 1995; Serfling et al. 1995; Serfling et al. 1989). The mouse and human IL-2 genes share close sequence homology of this promoter region (Novak er al. 1990). As shown in Figure 6, the minimal essential regulatory region of the IL-2 promoter/enhancer contains multiple DNA binding motifs for several critical TcR- inducible trans-acting factors important for transcriptional regulation of the IL-2 gene. These trans-acting factors include AP-l (activator protein-f), NFAT (puclear factor of activated I cells), Rel and NF-KB (auclear factor-EB) proteins, CREB (QAMP response alement-pinding protein), ATF (activation transcription factor), and ZEB (_z_inc finger/E; box binding protein). Protein binding to each of the cis-acting response elements is induced in response to TcR ligation and CD28 costimulation with the exception of the Oct sites. AP—l, NFAT, Rel/NF-KB, CREB, and ATF transcription factors, in general, are associated with positive regulation of IL-2 expression (Jain et al. 1995). In contrast, ZEB binding to the NRE-A (rregative response alement A) and CREB/CREM binding to Page - 33 Figure 6. Schematic of the 5’ minimal essential regulatory region of the mouse IL-2 promoter. The region of the H.-2 promoter that is essential for gene transcription is located approximately 300 bp 5’ of the transcription start site. Multiple cis—acting elements are localized within this 300 bp region as illustrated. In addition, numerous trans-acting factors, including AP-l, NFAT, NF-KB ZEB, CREB, and Oct bind to their respective response elements and function within the IL-2 promoter and function in a cooperative manner to regulate IL-2 gene transcription. The NRE-A represents the major negative regulatory cis-acting element in the IL-2 promoter. Page - 34 $2- 3 $1 82- 3 we: at- 8 B... :3. 3 8: 5.. 8 mm: 8... 3 83 Ann- 3 09¢ so. 3 NE mfi ._.<-h_z :42 biz :5. /./ We... mxnmz $2- 2 83 Page - 3S l. dastal TRE :l’mler al. l9“ ' | While {'1‘ renalmelmhel rages of the \Ig For example. k statues inelu. Slumzo and Acl Sgerrrrnal Luna era]. 1999: and Acute 19955,». I CDZSRE 01' the is pardoned I CDZSRE or the etal. 1999; Sh acm‘ation thro PTOV: des DNA the distal TRE (phorbol aster response alement) negatively regulate IL-2 expression (Yasui eta]. 1998). While the precise integrative contribution of CD28 signaling to TcR-signaling remains elusive, evidence suggest that CD28 positively cooperates with the TcR at early stages of the signaling cascade to enhance TcR signaling through numerous mechanisms. For example, CD28 signaling induces tyrosine phosphorylation of several cellular substrates including TCRC and ZAP-70 (Perez et al. 1997), enhances c-Fos expression (Tuosto and Acuto 1998), activates ERK (axtracellular regulated kinase) and JNK (Jun N—ferminal kinase) MAP (mitogen activated protein) kinases, (Jain et al. 1995; Kempiak et al. 1999) and potentiates NFAT, AP—l, and NF—KB transcriptional activity (Tuosto and Acuto 1998). Moreover, it has been demonstrated that full activation of the -164 bp CD28RE of the IL-2 promoter is dependent upon an adjacent AP-l response element that is positioned 2 bp downstream of the CD28RE. Site-direction mutagenesis of either the CD28RE or the adjacent AP-l binding motif disrupts function of the CD28RE (Iacobelli et al. 1999; Shapiro et al. 1997). Alternatively, CD28'B7 ligation also enhances T cell activation through post-transcriptional stabilization of IL-2 mRNA. The CD28RE also provides DNA binding sites for NF-AT, CREB/AFT, and NF-KB/Rel. The extensive integrative cross talk among multiple signaling pathways required for regulation of IL-2 expression is exemplified by the vast diversity of protein-DNA interactions that accompany T cell activation. A more complete description of the role of NF-KB, NFAT, AP—l , CREB, and ZEB trans-acting factors on IL-2 gene transcription is provided. Page - 36 NFAT p, 716' _ : 3 (hf/“k 1. N F-KB (nuclear factor-fl) The mammalian NF-KB proteins include NF-KBl (p50/p105), NF-KBZ (p52/p100), RelA (p65), c-Rel, and RelB. Inactive NF-KB proteins are retained in the cytoplasm of unstimulated T cells through an interaction with one or more of seven known IKB proteins (Baldwin 1996). TcR ligation results in rapid inactivation of IKBOL, and subsequent translocation of active NF-KB dimers into the nucleus where they bind to KB motifs to regulate transcription. CD28 costimulation generates a rapid, sustained Rel/NF-KB induction. The duration of this expression is regulated by selective phosphorylation and degradation of IKBOL and IKBB, which leads to a persistent nuclear expression of Rel/NF-KB because unlike IKBa, IKBB is not activated by NF-KB. c-Rel is the principal KB protein that binds to the CD28RE after CD28 ligation (Ghosh et al. 1993; Meyer et al. 1996). Overexpression of c-Rel or RelA activates KB-dependent transcription through the IL-2 CD28RE (Maggirwar et al. 1997). IL-2 expression is impaired in c-Rel knockout mice (Kontgen et al. 1995). Collectively, these results implicate a regulatory role of c-Rel in CD28-dependent IL-2 transcription. In addition to the CD28RE, a second KB binding DNA motif is positioned -208 bp upstream of the IL-2 gene transcriptional start. NF-KB proteins binding to each of these response elements is induced in response to TCR/CD28 costimulation. 2. NFAT (nuclear factor of activated I cells) The NFAT family of transcription factors includes the cytosolic NFATcl, NFATc2, NFATc3, and NFATc4 and the nuclear NFATn. The DNA binding motifs of NFAT proteins (120 to 140 kD) resemble those of Rel-family proteins; c-Rel, p50, and p65 show some overlap with NFAT in their ability to bind to the CD28RE in the IL-2 Page - 37 promOZCI. -\1 shown in Fill‘ and 45 bl‘ L SCIL’CUVC mu mRNA CXPII translocate tr manner upon in nuclear en dependent re PMA-activate Independent synergistic r: dependent an. NIH comp}; AP‘ 1 g, l roman: promoter. NFAT also cooperatively interacts with AP—l proteins (Rao et al. 1997). As shown in Figure 6, five NFAT response elements are positioned at -280, -l60, -l35, -90, and -45 bp upsteam of the mouse IL-2 transcription start site (Rooney et al. 1995). Selective mutation of any one of these five sites attenuates TcR/CD28-induced IL-2 mRNA expression. Phosphorylated NFATc proteins reside in the cytoplasm, and translocate into the nucleus in an intracellular Caz*lcalmodulin/calcineurin-dependent manner upon activation of the TcR. NFATc is dephosphorylated by calcineurin enabling its nuclear entry and subsequent transcriptional activity. However, induction of NFAT- dependent reporter gene expression and nuclear NFAT-DNA binding in CD28 plus PMA—activated T cells suggests that some NFAT proteins may be activated in a Ca“- independent manner (Rooney et al. 1995) and provides a putative mechanism for synergistic regulatory cross-signaling between TcR-mediated calcium/calcineurin- dependent and CD28-mediated calcium/calcineurin-independent pathways. The nuclear NFAT complex is comprised of an AP-l-NFATn heterodimer (Jain et al. 1992a). The AP-l component confers stability of the protein-DNA interaction (Jain et al. 1993). 3. AP-l (Activating protein-l) AP—l comprises homodimers of Jun (c-Jun, JunB, and JunD) or heterodimers of Jun and F05 (c-Fos, FosB, Fra-l, and Fra-2) that bind to the TGAlC/GlTCA TRE (Karin et al. 1997). Leucine zippers dictate the selectivity of AP-l protein-protein dimerization, and basic amino acid sequences confer specificity of DNA-protein interactions. The AP-l superfamily comprises the Jun family, the F08 family, and the maf family of transcription factors. Other families of trans-acting proteins are known to interact with AP-l, including additional bZIP (basic/leucine ripper Page - 38 interacting Pr" proteins of 13"- Wisdom 199‘? 199%. AP-l injependent. l'llEl-{IMEI- a]. 19943: [It though CD25; stabilizes the r Marsurnoto er residues of c-l Witty presun identified in th located in com Rome." 6’! al. “Onsengus Sl' Silt by 006 hp j Recent "SOAP‘I site interacting proteins), CREB/AFT, NFAT, NF-KB, Smads, the ETS proteins, and the Oct proteins of the POU family (Brodin et al. 2000; Liberati et al. 1999; Mulder 2000; Wisdom 1999; Wong et al. 1999; Xu et al. 2000b; Yingling et al. 1997; Zhang et al. 1998). AP-l transcriptional activity depends on the activation of at least two independent, converging signaling cascades, namely the p56lck/PKC/p21/Ras/Raf- 1/MEK1/MEK2/ERK1/ERK2 MAPK pathway activated through the TcR (Izquierdo et al. 1994a; Izquierdo et al. 1994b) and the JNKl/JNK2 MAPK pathways activated through CD28 ligation (Cheng et al. 2000; Su et al. 2001). IN K phosphorylation of c-Jun stabilizes the protein complex by suppressing ubiquitin-dependent degradation (Hermida- Matsumoto er al. 1996). Furthermore, JNK phosphorylation of the ser-63 and ser-73 residues of c-Jun augments c-Jun0CBP/p300 interactions to enhance AP—l transcriptional activity presumably by enhancing binding affinity. Five AP-l binding motifs have been identified in the mouse IL-2 promoter (Figure 6). Four of these DNA binding sites are located in conjunction with NFAT binding motifs (~280, -180 -l60, —135, and —90) (Rooney et al. 1995). The AP—l site at -l80 bp upstream of the transcriptional start site is a consensus site in the reverse orientation, but varies from the consensus AP-l binding site by one bp in the forward orientation. Recent evidence suggest that a complex of CREB and CAMP elements bind to the -l80 AP-l site and functions to suppress IL-2 gene transcription in anergic CD4“ T cells (Powell et al. 1999). The cooperative interaction of AP-l and NFAT proteins increases DNA recognition motif selectivity as well as DNAOprotein stabilization. NFAT transcriptional activity, as described above, is primarily Cay-dependent; therefore, AP-lONFAT interactions also provide a mechanism of transcriptional regulation through Page - 39 t} .1. intraceuhu heterodxmt frat} me gm): Fe: feitlilzes 0 preferentic hetero-dim; exelusixe p56ick'PK intracellular signaling pathway crosstalk. 4. CREB (QAMP response clement binding protein) CREB, a 43 kD member of the ATF/CREB family transcription factor, contains a bZIP DNA binding motif that contains a cluster of basic amino acids and leucine zipper structures (Busch and Sassone-Corsi 1990; Maekawa et al. 1989). CREB family members form dimers through their leucine zipper regions and bind to the octanucleotide CRE (QAMP response alement) element (TGANNT CA). CREB family members, including CREB, ATF-1, ATF-2, ATF-3, ATF-4, and CREM bind as homo- or heterodimers in combination with other CREB family members. Alternatively, CREB family members also heterodimerize with members of the Jun (i.e., c-Jun, JunB, and junD); Fos (i.e., c-Fos, FosB, Fra—l, and Fra-2); and maf (i.e., mafK, mafB, and nrl) families of trans-acting factors. This oligomerization is somewhat specific as c-Jun preferentially dimerizes with ATF-2, ATF-3, and ATF-4; whereas, c—Fos selectively heterodimerizes with ATF-4. TcR-induced CREB transcriptional activity is regulated exclusively by phosphorylation of ser-133 via a CAMP-independent p56lck/PKC/Ras/Raf—1/MEK/RASK2 (ribosome associated S6-lginase 2) pathway (Muthusamy and Leiden 1998). Dominant negative CREB mutant studies suggest that Fra-2 and F053 are the principal CREB trans-acting factors that are activated in response to TcR/CD28 signaling (Berkowitz and Gilman 1990; Berkowitz et al. 1989). The significance of Fra-2 and FosB selectivity in IL-2 expression has not yet been elucidated. ATF-2 is phosphorylated and stimulated by JNK MAPK CM kinase) and p38 MAPK at Thr-69, Thr-7l, and Ser-90 (Gupta and Terhorst 1994; Gupta et al. 1996; Gupta et al. 1995). Activated ATF-2 binds to CRE either as a homodimer or as a Page - 40 heECIL‘JImC $130305”? 1‘ leafusa r to the NRE- transcriptml regulation C target distir. seEeetitely r F05. e-Jun. c Inaddltton to Chair} en}; ZiJ'lCt ZEB'Uibl cor heterodimer with c-Jun (Hai and Curran 1991; Macgregor et al. 1990). ATF-2 oligomerizes with Smads and is directly phosphorylated by TAK-l (TGF-B activating kinase-f) suggests that ATF-2 is a direct nuclear target of TGF-B-induced signaling (Hanafusa er al. 1999; Sano er al. 1999). 5. ZEB ZEB (_z_inc finger/E-box binding protein) is a trans-acting factor that binds to the NRE-A (aegative response alement-A) located -108 bp upstream of the mouse IL-2 transcription start site (Yasui et al. 1998). ZEB-NRE-A binding confers negative regulation of IL-2 expression. ZEB contains two independent repressor regions that target distinct sets of transcription factors and regulate different tissues. Region I selectively represses hematopoietic-restricted transcription factors, including RelA, c- Fos, c-Jun, c-myb, E2F-1, and ETS family members (Postigo et al. 1999). The second region, Region II, specifically inhibits the activity of the myogenic transcription factor, MEF2C (myocyte anhancer factor E) to regulate muscle differentiation. ZEB binds to several different E box sequences, but has a higher affinity for the CACQTG sequence. In addition to repressing IL-2 expression, ZEB also silences the immunoglobulin heavy chain enhancer and GATA-3 transcriptional activity (Postigo et al. 1999). In contrast, ZEB-(x4b1 confers co-stimulatory signaling in T cell activation. V. TGF-[31 signaling through Smads The intracellular signaling mechanisms responsible for TGF-Bl-mediated immune modulation remain relatively unexplored. Therefore, this section will primarily, albeit not exclusively, describe the current understanding of Smad signaling in mammalian Page - 41 551:3. 51‘ I: A" mom :1? PKC. MAPK. 17 and Cher: like?) 17...: control on A. Sm B sup rt]:rr Smad prote contaim thr llHl Ariad- iinker regio; al. i999; SI“. domain of 1 Stquences Vl “’“mtd 1}. m» Mum lim\ ‘ cells, as little information is yet available regarding Smad signaling in lymphoid tissue. As shown in Figure 7, eight TGF-Bl-dependent signaling cascades have been proposed: (1) PKC, (2) PLC, (3) protein phosphatase 1, (4) ras, (5) p38 MAPK, (6) ERKl/ERK2 MAPK, (7) JNKl/JNK2 MAPK, and (8) Smads [(Hartsough and Mulder 1997; Massague and Chen 2000; Mulder 2000; Piek er al. 1999; Roberts 1999; ten Dijke et al. 2000). It is likely that these pathways integrate at multiple levels to elicit an intricate regulatory control on the overall biological effects of TGF-Bl. A. Structures and Functions of Smad proteins Smad proteins are intracellular signaling mediators unique to members of TGF- B superfamily. At present, ten Smad proteins (i.e., Smad1-10) have been identified. Smad proteins range in molecular weight from 45 to 65 kD and, as illustrated in Figure 8 contains three structurally distinct domains; i.e., two conserved domains, an N-terminal MHl (Mad bomology) domain and a C-terminal MH2 domain, and a variable proline rich linker region inserted between MHl and MH2 domains (Chacko et al. 2001; Dennler et al. 1999; Shi et al. 1998; Wrana 2000). One of the principal functions of the MHl domain of the receptor-activated Smads (i.e., Smad2 and Smad3) is to bind DNA sequences via its B-hairpin loop (Massague and Wotton 2000; Shi et al. 1998). A conserved lysine-rich motif in the N-terminus of Smad2 and Smad3 resemble the classic simian virus 40 large antigen NLS (puclear localization _s_ignal) and may function as a nuclear localization signal for Smad3 (Shi et al. 1998; Xiao et al. 2000). In contrast to Smad3, Smad2 nuclear import appears to be NLS-independent (Xu et al. 2000a). The precise regulatory mechanisms of Smad nuclear localization remains under investigation. The lysine rich MHl domain functions predominantly in DNA'protein Page - 42 Figure 7. Intracellular signaling pathways of TGF-B1. TGF- Bl mediates it biological effects by binding to a transmembrane TGF-B receptor complex with intrinsic serine/threonine kinase activity. Shown here are the eight known signaling pathways that are activated upon ligand binding to the TGF-B receptor complex. Page - 43 TGF-B1 TGF‘BRIoTGF'BRI TGF'BRII-TGF'BRII K .I llllllll IVA " T _ " KW m. uuuuuu vmm m m hm“. " D .mm ..|V_ m .mh " S \\‘muwu " \ .m u IIIIIIII x K D. m up u. uuuuuu YEW K n 8P r ............. vwm Page - 44 Figure 8. Schematic of the functional domains of the Smad proteins. The Smad proteins are intracellular signal transducers for TGF-B superfamily proteins. Shown here is a schematic structure of R-Smads and their functional domains. Smads are comprised of three distinct structural domains, including the MHl domain, the MH2 domain, and the linker region. The MHl domain is critical for DNA binding; whereas, the MH2 domain is essential for protein-protein interactions. The SSXS motif, located at the extreme carboxy terminus, represents the Ser phosphorylation sites for the TGF- Breceptor. Co-Smad, Smad4 does not contain a SSXS domain. The PSSP site, located in the linker region, represents the phosphorylation site for ERK MAPK. Page - 45 c2528.“ _acozntomcfl... . :28“..qu F-._.mx~.<.42.2.}. 6.3:: . ~Q 1.. :. .99.!2 :33: 12:... u v.9 wen-3. =3=u~wuu€Ez haw =-Z~L=l=~nfa .vAfiNuOCQCA- .h 9--a~t .:_.:.:_:_.a>....1 2.5.55.2... 23.232 0...... \ 2:227. _- Ala—r 00000.0 .000008 .0 3.00.6. 000 0.00.0 ..0008 0. 8.0.0000 m0..0> .0>.>..0m 0.200.000: 0:0 000.008.0000. 00300008 0. 0-0.0... 0. 20.508000 03.00.0n. .00.00.0>0 .0: 02.000. 8.0.05 00000.0. .000008 ....3 00.8 0. 00.50008 000 00.50.0008 800.0020 00000.00. 0-00.0 ... ..o........... 5380 9 3.32800 0.0.0. 0:00 m 00.03.00 ma... 00000:. .00..m0.0. 0. 2.00 m 03200.. 0 07. 00.0... 0. 080000. .0080 0.30% 0.00. 000 .00..0.0....0... 00000.00. 6030000.. 00.03.00 3000. 2.00 .H 2.000.000. .028 2.080.083 0. 5.00.0..00 0.8%.... 000000000 000 b.0000 .0208 5.3 02.08.08 0.3008»... .0887. 808020.60 0.005080 .0887. .0w0 .0 8.003 0 >0 0.00m 0-00... 0. 0.008.005 .0802 000.00 00..0w0..ww0 8.0.05 000.000.00.038 30.008 000 0.0.0.0... 000 0.2000000 00.000008 800.0020 00000.00. 0.00.0009... 220...). 00.0.0 0. 00000300090. 0.0.0. 00.50008... m 00.8.00 .mw. 80:00 00000000. 000 000008 0. 2.00 8.0000... <3 000000”. 30.008 0000 000 00000 008... 0. 2.00 0800... h.0 00020808.... 30.008 0000 000 00030 0.0008... 8.000.000 0. 2.00 m bmmm 00000.00Q 000083.000 80.0...00 00088. 0. 0:0 0:00 .....wQU .0 +30 .0 0000.000 03> .... >0 000.05% 00.200000 00.0800 0:088. 00.000.800.00 Ow. 0.003500 .....3 080.00.? 00088.0.0< 000.0... >0 00.20.00. 0.30% 0. 0>...80m 000.00. 2.28 00000.00. 0:00 ... 00.03.00 8.3 0000.. 0.0008... .0000.....0n. 00.8 00080.08? 0. 2.00 03200.. 0.w0.m +30 00000.00. 8.3 802008.... 032000 0.0000 +wQU +090 000000. 000 3.8.2.8 00000.05 02.0.. 2.0000: 8.80 02.80.08 0.3008»... .0802 0000082000 00030 2.... ... .03 ...... 8.5 x 0.8000 02...: ... 08 0.05.... 0.008.080 000 200082000? 000 0:0» 0. 0.00.00 800000000805 Dome ..-mfimem 8.... +0.00... 3.... 80-0000 0.... ......0020 .0 08......an 2.082... .. 2...... Page —55 al. 1999; Yang et al. 1999). Neutrophil and macrophoge chemotactic responsiveness to TGF-B, are also significantly impaired (Ashcroft et al. 1999; Yang et al. 1999). Smad3- null T cells display an activated in vivo phenotype and are unresponsive to TGF-B,- induced growth inhibition in vitro (Datto et al. 1999; Yang et al. 1999). In contrast, growth inhibition by TGF-[3l in vitro is unaffected in LPS (lipopolysaccharide)-activated Smad3-null B cells (Datto et al. 1999; Yang et al. 1999). Normal hematopoiesis and an apparent undisrupted marrow response to ongoing infections in these mice suggest that the breakdown of mucosal immune response is not due to defective myloid or lymphoid cell development (Yang et al. 1999). In contrast to TGF-[3l deficiency, Smad3-null mice present no evidence for autoimmunity (Yang et al. 1999). In further contrast to the TGF- Bl-null mice, B cell development and IgA production in vivo are seemingly unaffected with deletion of the Smad3 gene (Yang et al. 1999). Smad2 and Smad4 null mice die during embryogenesis (Dunker and Krieglstein 2000; Waldrip et al. 1998). A second line of evidence for a role of Smad3 in immune cells stems from functional cooperation between Smad3 and the transcription factor, AML (adult myeloid leukemia) in regulating TGF-[3l responsiveness (Kurokawa et al. 1998a; Kurokawa et al. 1998b). AMLl/Evi-l transcriptional activity regulates TGF—Brmediated myeloid cell growth inhibition and plays a profound functional role in leukemogenesis (Imai et al. 2001; Izutsu et al. 2001). Additionally, Smad3 has been implicated in the transcriptional regulation of germline Iga by TGF-[3,. Transcriptional regulation of Iga by TGF-[3l is mediated through a proximal promoter/enhancer region located approximately -130 bp upstream of the human and mouse intron (1) 0t transcription start site. This sequence contains an array Page - 56 of Int: bmizrj pfOlCll’ vvvvvv of interspersed Smad, acute myloid leukemia (AML), and CAMP-response element- binding protein (CREB) response elements that cooperatively, via protein'protein and protein-DNA interactions, mediate in vitro TGF-B, transcriptional responsiveness (Pardali et al. 20003; Zhang and Derynck 2000). In vitro EMSA analyses using GST (glutathione-s—transferase)-fusion proteins have revealed direct binding of both Smad3 and Smad4 to the M elements within this region (Zhang and Derynck 2000). Furthermore, in the presence of TGF-[3,, over-expression of Smad3 and Smad4 selectively increases both surface IgA expression and IgA production by murine B lymphoma cells in vitro (Park et al. 2001). Collectively, these results provide direct evidence that Smad3 plays a role in immune cell responsiveness to TGF-[3l in vitro and in vivo. VII. Regulatory cross talk between Smad and MAP kinase Signaling Cascades Significant progress has been made in elucidating intracellular signaling pathways and establishing the biochemical moieties that are activated by individual ligand'receptor interactions. More recently, it has become apparent that synergistic as well as antagonistic cross talk at multiple hierarchical levels among individual pathways influences the overall physiological significance of ligand binding. In support of this, TGF—Bl-induced activation of the TBR complex initiates signaling through multiple non- Smad-mediated signaling cascades (Figure 10). For example, TGF-Bl upregulates p38 MAPK and ERK MAPK signaling in multiple epithelial and fibroblast cell models (Fanger 1999; Mulder 2000; Terada et al. 1999; Yue and Mulder 2000). The putative regulatory crosstalk between TGF-Bl-activated Smad and TGF-Bl-activated MAPK Page - 57 Figure 10. Putative regulatory cross talk between Smad signaling and MAPK signaling cascades. TGF-B,-induced biological activity is likely to be regulated through cross talk between multiple signaling pathways that are upregulated with ligation of the TGF-[3 receptor complex. Illustrated here are examples of positive and negative regulation between Smad and MAPK signaling. Page - 58 TGF-[31 Wm unca— mum”: ME so 83 a . 'iumem iUmEm mumEm Tad. mvumem E fl 2: mvumeému & 25m 28$ 5. m8 ~\ 535 Page — 59 path“ a}' oli-‘ome: — b 1999'. H: ATF-2 It Xishzhar; thing 1h: Colleen» , ’1 P. ~ ' 55-:5‘9311 a 3‘3 pmic pathways is an intense focus of current research. It has been demonstrated that ATF-2 oligomerizes with Smad3 and Smad4 in a TGF-Bl-dependent manner (Hanafusa et al. 1999; Hocevar et al. 1999; Sano et al. 1999). ATF-2, a member of the CREB family that binds to the cAMP response element, is activated by TcR-induced p38 MAPK-mediated phosphorylation (Zhang et al. 1999a). Moreover, Smad3 and Smad4 also complex with CBP and p300, two co-activator accessory molecules that also hetero-oligomerize with ATF-2 to augment the stability of ATF-2 binding to the CRE (Nishihara et al. 1999; Nishihara er al. 1998). Synergism of ATF-2 transcriptional activity has been confirmed using the TGF—Bl-responsive p3TP-Lux reporter plasmid (Hanafusa er al. 1999). Collectively, these results demonstrate that Smad and p38 MAPK signaling act in a synergistic manner to enhance TGF-Bl-responsive transcriptional activity. These results are particularly interesting in light of the fact that TcR ligation also induces p38 MAPK activity and ATF-2 binding to DNA response elements in the IL-2 promoter. A second line of evidence suggesting Smad and MAPK signaling crosstalk is evident from oncogenic Ras-mediated inhibition of Smad nuclear translocation. Specifically, prolonged Ras-induced ERK MAPK activation has been demonstrated to result in phosphorylation of Smad2 and Smad3 at four Ser/Thr-Pro sites located within linker region, resulting in the inability of Smad2 and Smad3 to dimerize with Smad4 (Figure 8) (Kretzschmar et al. 1999; Ulloa et al. 1999). Mutation of these Ser/Thr-Pro ERK MAPK phoshorylation sites confers responsiveness to TGF-[31 in the presence of oncogenic Ras (Kretzschmar et al. 1999). Transcriptional modulation of CREB and AP-l trans-acting factors by Smads through direct protein-protein interactions provides a third distinct line of evidence implicating crosstalk between Smad and MAPK pathways Page - 60 «Denali Collect pncnorr PMRIOI immun: contri’m GIT cc Prof-our (Dennler et al. 1998; Dennler et al. 2000; Sano er al. 1999; Wong et al. 1999). Collectively, these data suggest that Smad regulation by MAPKS may be a general phenomenon for controlling TGF-Bl-mediated signaling. VIII. Objective and specific aims TGF-[3l is a multifunctional cytokine that plays a profound role in maintaining a physiological balance between positive and negative regulatory signals essential for immune homeostasis. Experimental and clinical data suggest that disruption of TGF-[3l contributes to aberrant pathogenic immune and immunological responses. One of the cell populations that is strongly regulated by TGF-[3, is the T lymphocyte. Direct modulation of T cell activation and clonal expansion is one mechanism by which TGF-[31 elicits profound regulatory control of T cell-dependent humoral immunity. The mechanisms underlying the effects of TGF-B, on T cells remains elusive. The overall goal of this research is to elucidate the mechanisms underlying the ability of TGF-B, modulate T cell effector function differently. Towards this end, the first objective was to establish an in vitro model to characterize the effects of TGF-[3l on T cell activation and proliferation. Smads, a novel family of transcription factors, have recently been identified as TGF-Bl-inducible intracellular signaling mediators. A role for Smads in lymphocytes has not been well characterized. (ALIA DNA sequences have been identified as Smad3 response elements. We have identified five M sites in the 5’ minimal essential regulatory region of the mouse IL-2 promoter (Figure 11). Thus, the second objective was to determine whether Smads play a role in regulation of IL-2 by TGF-[31. Page - 61 Figure 11. Identification of CAGA sequences in the 5’ minimal essential regulatory region of the mouse IL-2 promoter. Smad3 acts as a transcription factor by binding to sequence specific DNA motifs located in the promoter of TGF-Bl-responsive genes. Five CAGA sequences have been identified within the minimal essential regulatory region of the mouse IL-2 promoter. As illustrated, each M sequences is located adjacent to or overlaps a DNA binding element for another transcription factor(s) that regulate IL-2 transcription including NFAT, NF-KB, AP-l, CREB and ZEB. It is hypothesized that the CAGA “Smad boxes” are sequence specific DNA binding motifs for Smad3 in the IL-2 promoter. Page - 62 310 ._.._.._.uuu98% purity. Briefly, a pooled single cell suspension was isolated from the spleens of four animals, washed in labeling buffer, centrifuged at 300 x g for 10 min and resuspended in 90 ”L per 107 total cells. A 10 11L aliquot of MACS CD90 Microbeads (Miltenyi Biotec) were added to the cell suspension and incubated for 15 min at 12°C. The cell suspension was applied to a LSVVS+ positive selection column (Miltenyi Biotec). After the negative cells were allowed to pass through the column, the column was rinsed, removed from the separator, and placed into a collection tube. The positive cells were removed by adding 5 mL of LS‘“/VS+ buffer and gently forcing the contents of the column with a plunger. IX. In vitro Proliferation Assays Single cell splenocyte or thymocyte preparations were cultured in 96 well culture plates at 2.5 x 106 cells/mL in the presence or absence of TGF-Bl. T cells were activated with plate-bound (x-CD3 (2 ug/mL) + a-CD28 (1 tig/mL) or LPS (10 ug/mL). Cells were cultured at 37°C and 5% C02 for 48 hours (LPS) or 72 hours (a-CD3/0t-CD28) and pulsed with 1.0 uCi/well of [3H]-thymidine (NEN, Boston, MA) for the last 16 hours of culture. Cells were harvested onto glass fiber filters with a PHD Cell Harvester (Cambridge Technology, Inc., Cambridge, MA), and tritium incorporation was quantified using a Packard 460C liquid scintillation counter. Page - 71 X. ELISA (Enzyme-Linked Immunosorbent Assay) A. Mouse IL-2 ELISA Splenocytes or thymocytes were cultured in triplicate (5 x 106 cells/mL) in 48 well culture plates or in quadruplicate (5 x 106 cells/mL) in 96 well plates for 24 hours. Supematants were collected and quantified for secreted mouse IL-2 by sandwich ELISA. Immulon IV Removawell microtiter strip wells (Dynex Technologies, Inc., Chantilly, VA) were coated overnight at 4°C with 50 pl of purified rat (it-mouse IL-2 antibody (1.0 pg/mL) in 0.1 M sodium bicarbonate buffer (pH 8.2). Blocking buffer, 300 pl of BSA (bovine serum albumin) (3% v/v) in 0.01 M PBS (phosphate buffered saline) containing 0.1% (v/v) Tween 20 (BSA-PBST), was added to each well and incubated at 37°C for 30 minutes. Wells were washed four times with PBST followed by addition of IL-2 standard or sample (50 pl) and incubated at 37°C for one hour. After incubation, the plate was washed four times with PBST and once with distilled water. The biotinylated anti-mouse IL-2 (1.5 pl/mL), diluted in 3% BSA-PBST, was added to each well (50 pl) and incubated at room temperature for one hour. The plate was washed six times using the PBST solution and once with distilled water followed by addition of 50 pl streptavidin-horseradish peroxidase (1.5 pg/mL) for one hour at room temperature. Samples were then washed eight times and the bound peroxidase conjugate was detected by addition of a substrate solution (100 pl/well) containing 0.1 M citric-phosphate buffer (pH 5.5), 0.1 mg/mL TMB (tetramethylbenzidine) (Fluka Chemical Corp., Ronkonkoma, NY) and 1% H202. The reaction was terminated with an equal volume of 6N H2SO4, and absorbance was quantified at 450 nm using an EL808 automated microplate reader (Bio-Tek Instruments, Inc., Winooski, VT). A standard curve for each assay was Page - 72 generated using recombinant mouse IL-2 (PharMingen). The DeltaSoft 3-computer analysis program (BioMetallics, Princeton, NJ) was used to quantify secreted IL-2. B. Mouse IL-4 ELISA Secreted mouse IL—4 was quantified as described for mouse IL-2 with the following modifications. Immulon IV Removawell microtiter strip wells were coated overnight at 4°C with 50 pl of purified rat anti-mouse IL-4 antibody (1.0 pg/mL) in 0.1 M sodium bicarbonate buffer (pH 8.2). Wells were washed four times with PBST prior to the addition of IL-4 standard or sample (100 pl) and incubated at 37°C for one hour. Biotinylated anti-mouse IL-4 (1.5 pllmL), diluted in 3% BSA-PBST, was added to each well (50 pl) and incubated at room temperature for one hour. The reaction was terminated with an equal volume of 6N H2804, and absorbance was quantified at 450 nm using an EL808 automated microplate reader (Bio-Tek Instruments, Inc). A standard curve for each assay was generated using recombinant mouse IL-4 (PharMingen). The DeltaSoft 3—computer analysis program (BioMetallics) was used to quantify secreted IL-4. C. Mouse IF N-y (interferon gamma) ELISA Secreted mouse IFN-y was quantified as described for mouse IL-2 with the following modifications. Immulon IV Removawell microtiter strip wells were coated overnight at 4°C with 50 pl of purified rat anti-mouse IFN-y antibody (1.0 pg/mL) in 0.1 M sodium bicarbonate buffer (pH 8.2). Wells were washed four times with PBST prior to the addition of IFN-y standard or sample (50 p1) and incubated at 37°C for one hour. Biotinylated anti-mouse IFN-‘y (1.5 pl/mL), diluted in 3% BSA-PBST, was added to each well (50 pl) and incubated at room temperature for one hour. The reaction was Page - 73 ‘IJ terminated with an equal volume of 6N H2804, and absorbance was quantified at 450 nm using an EL808 automated microplate reader (Bio-Tek Instruments, Inc.). A standard curve for each assay was generated using recombinant mouse IFN-‘y (PharMingen). The DeltaSoft 3-computer analysis program (BioMetallics) was used to quantify secreted IFN—y. D. Human IL-2 ELISA Secreted human IL-2 was quantified as described for mouse IL-2 with the following modifications. Immulon IV Removawell microtiter strip wells were coated overnight at 4°C with 50 pl of purified rat a-human IL-2 antibody (1.0 pg/mL) in 0.1 M sodium bicarbonate buffer (pH 8.2). Wells were washed four times with PBST prior to the addition of human IL-2 standard or sample (50 pl) and incubated at 37°C for two hours. Biotinylated anti-human IL—2 (1.5 pl/mL), diluted in 3% BSA-PBST, was added to each well (50 pl) and incubated at room temperature for two hours. A standard curve for each assay was generated using recombinant human IL-2 (PharMingen). The DeltaSoft 3-computer analysis program (BioMetallics) was used to quantify secreted IL- 2. E. Mouse IgM ELISA Immulon IV Removawell microtiter strip wells (Dynex Technologies, Inc., Chantilly, VA) were coated overnight at 4°C with 50 pl with anti-mouse immunoglobulin (Ig) capture antibody (Boehringer Mannheim, Indianapolis, IN) in 0.1 M sodium bicarbonate buffer (pH 8.2). Blocking buffer, 300 pl of BSA (3% v/v) in 0.01 M PBS containing 0.1% (v/v) Tween 20 (BSA-PBST), was added to each well and incubated at Page - 74 Wane CUIIUR 37°C for 30 minutes. Wells were washed 4 times with PBST followed by addition of sample (100 pl) of supernatant and incubated at 37°C for 90 minutes. After incubation, the plate was washed four times with PBST and once with distilled water. Horseradish peroxidase conjugated-anti-mouse IgM detection antibody (Sigma, St. Louis, MO) was added to each well and incubated at 37°C for 90 minutes. The plate was washed 6 times using the PBST solution and once with distilled water to remove unbound antibody. ABTS (2,2’-azinobis (3-ethybenzthiazoline-§ulfonic acid) substrate (Boehringer Mannheim, Indianapolis, IN) was added and the rate of colorimetric change was monitored at 405 nm for 60 minutes using an EL808 automated microplate reader (Bio- Tek Instruments, Inc.) Secreted IgM concentrations were generated from the standard curve generated from known IgM concentrations using the DeltaSoft 3-computer analysis program (BioMetallics). The pronase viability assay was used to determine the number of viable splenocytes/well recovered following incubation. Results from quadruplicate cultures are expressed as the mean pg IgM/106 recovered viable splenocytes i SEM. F. Mouse IgA ELISA Immulon II Removawell microtiter strip wells (Dynex Technologies, Inc., Chantilly, VA) were coated overnight at 4°C with 50 pl purified rat anti-mouse IgA antibody (1.0 pg/mL) in 0.1 M sodium bicarbonate buffer (pH 8.2). Blocking buffer, 300 pl of BSA (3% v/v) in 0.01 M PBS containing 0.1% (v/v) Tween 20 (BSA-PBST), was added to each well and incubated at 37°C for 30 minutes. Wells were washed four times with PBST followed by addition of sample (100 pl) of supernatant and incubated at 37°C for 90 minutes. The standard curve was generated using mouse recombinant IgA, (Sigma, St. Louis, MO). After the incubation period, the wells were washed with PBST, Page - 75 strept. Follow substr and 50 pl biotinylated anti-mouse IgA in 3% BSA-PBST (1.5 pl/mL) was added to each well for 60 minutes at 37°C. The wells were washed with PBST solution, and 50 pl streptavidin-peroxidase (1.5 pg/mL) was added for one hour at room temperature. Following a final PBST wash, bound peroxidase conjugate was detected by addition of a substrate solution (100 pl/well) containing 0.1 M citric-phosphate buffer (pH 5.5), 0.1 mg/mL TMB, and 1% H202. The reaction was terminated with an equal volume of 6N H2804, and absorbance was measured at 450 nm using a microplate reader (Bio-Tek Instruments, Inc.). IgA concentrations were generated from the standard curve generated from known IgA concentrations using the DeltaSoft 3-computer analysis program (BioMetallics). The pronase viability assay was used to determine the number of viable splenocytes/well recovered following incubation. Results from quadruplicate cultures are expressed as the mean ng IgA/106 recovered viable splenocytes i SEM. XI. Quantitative RT-PCR (Reverse Transcriptase-Polymerase Chain Reaction) A. Preparation of mouse IL-2 Internal Standard for RT-PCR A recombinant IL-2 18 (internal standard) was generated using a rat B-globin sequence as the spacer gene as described (Condie et al. 1996). The design of the IS primers used was (5' to 3'): IS forward primer = T7 promoter (TAATACGACTCACTATAGG), IL-2 forward primer (TGCTCCTTGTCAACAGCG), and rat B-globin forward primer (GGTGCTTGGAGACAGAGGTC); and IS reverse primer = (dT)18, IL-2 reverse primer (TCATCATCGAATTGGCACTC), and rat B- globin reverse primer (TCCTGTCAACAATCCACAGG). PCR conditions for making the internal standard were performed using 100 ng of rat genomic DNA as described Page - 76 [\nnt Wits by tr Trans temp} RIP: tranxc llgCl Mn SQCOII XII. (Vanden Heuvel et al. 1993). Purification of PCR products was performed using the Wizard PCR Prep DNA purification system (Promega, Madison, WI), and was followed by transcription of the products into RNA using Promega's Gemini II In Vitro Transcription System. The IS was treated with RNase-free DNase to remove the DNA template. B. Quantitative Competitive RT-PCR for mouse IL-2 Total RNA was isolated using TriReagent (Sigma, St. Louis, M0) for competitive RT-PCR. One hundred ng total RNA and known amounts of IS rcRNA were reverse- transcribed into cDNA using oligo(dT)15 as primers. A PCR reaction buffer, (4 mM MgC12, 6 pmol of IL-2 forward and reverse primers, and 2.5 units of Taq DNA polymerase) was added to the cDNA samples. Samples were heated to 94°C for 15 seconds, 55°C for 30 seconds, and 72°C for 30 seconds followed by a single extension step at 72°C for 5 minutes. Thirty-five cycles were used to amplify the IL-2 PCR product, which were then electrophoresed in 3% NuSieve 3:1 gels (FMC Bioproducts, Rockland, ME) and visualized by ethidium bromide staining; the IL-2 product generated from the sample RNA was 391 bp and the 18 product was 474 bp. A Gel Doc 1000 video imaging system (BioRad, Hercules, CA) was used to quantify the 391 bp and 474 bp bands. The number of transcripts was calculated from a standard curve generated from the density ratio between the gene of interest (IL-2) and varying known amounts of IS. XII. Protein Extractions A. Whole Cell Page - 77 Cells were cultured in 60 mm2 tissue culture plates at a density of 5x10° cells/mL (5 mL/plate). For whole cell protein extraction, cells were lysed in 500 pl of RIPA buffer (1x PBS, 1% Nonidet P40, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with lmM DTT, 1 mM PMSF, and 1 pg/mL of aprotinin and leupeptin, homogenized with a dounce homogenizer (30 strokes), and incubated on ice for 30 minutes. Following lysis, samples were centrifuged at 17,500 x g for 20 minutes and the supernatant retained as whole cell protein and stored at -80°C until use. B. Nuclear Cells were cultured in 60 mm2 tissue culture plates at a density of 5x10° cells/mL (5 mL/plate). Cells were lysed with a hypotonic buffer (10 mM HEPES and 1.5 mM MgCl2, lmM DTT, lmM PMSF, and 1 pg/mL of aprotinin and leupeptin) and the nuclei were pelleted by centrifugation at 6,700 x g for 5 minutes. Nuclei were lysed in hypertonic buffer (30 mM HEPES, 1.5 mM MgCl2, 450 mM NaCl, 0.3 mM EDTA, and 10% glycerol) supplemented with lmM DTT, 1 mM PMSF, and 1 pg/mL of aprotinin and leupeptin for 30 minutes on ice. Following nuclear lysis, the samples were centrifuged at 17,500 x g for 15 minutes, and the supernatant was retained and stored at -80°C until use. C. Cytosolic During isolation of nuclear proteins, the supematants from the hypertonic lysis were retained for cytosolic protein extraction, whereas the pellet was used to extract nuclear proteins. Briefly, the supernatant was centrifuged for one hour at 100,000 x g at Page - 78 Xlll. 1.005. 3.5. 5. millUl: 4°C. Glycerol was added to the resultant supernatant to 10% final concentration. Samples were stored at -80°C until use. XIII. Protein Determination Protein determination was performed using a BCA protein assay kit (Sigma, St. Louis, MO). A standard curve was prepared using increasing concentrations of BSA (0, 2.5, 5, 10, 15, and 20 pg) in a total volume of 100 pl. Unknown protein samples (5 pl) in a final volume of 100 pl were added to a 2 mL volume of bicinchoninic acid plus copper (II) sulfate, and incubated for 30 minutes at 37°C. Absorbance at 562 nm as measured using a Beckman DU-600 spectrophotometer (Fullerton, CA). XIV. Western Immunoblotting Proteins (25 pg) were incubated with 4x loading dye (62.5 mM Tris, 2% SDS, 10% glycerol, 0.01% bromophenol blue, and 1% 2-mercaptoethanol) and heated for 10 minutes at 95°C. Samples were then resolved on a 10% SDS-PAGE gel and transferred overnight at 4°C to nitrocellulose in transfer buffer (24 mM Tris, 191 mM glycine, 20% methanol). The nitrocellulose was blocked for one hour with BSA-TBST (Tris buffered saline plus Tween 20) prior to incubation with the primary antibody. A corresponding Ig horseradish peroxidase-linked secondary antibody was used for protein detection using the ECL chemiluminescent system (Amersham, Arlington Heights, IL). Band intensity was quantified using a Gel Doc 1000 video imaging system (BioRad, Hercules, CA). Page - 79 IT. elect X\' (‘7 (D‘ XV. EMSA (Electrophoretic Mobility Shift Assay) Binding experiments were performed with 5 pg of nuclear extract and 20,000 cpm of [y-3ZP]-dATP-labelled oligonucleotide. The binding reactions were resolved by electrophoresis in 4% nondenaturing polyacrylamide gels in 0.5x TBE buffer (1X TBE: 89 mM Tris, 89 mM boric acid, and 2 mM EDTA). Nuclear extract were incubated in binding buffer (70-100 mM NaCl, 30 mM HEPES, 1.5 mM MgCl2, 0.3 mM EDTA, 10% glycerol, 0.05% Nonidet P-40, lmM DTT, 1 mM PMSF, and 1 pg/mL of aprotinin and leupeptin) with 0510 pg of poly (dI-dC). Following electrophoresis, the gel was dried and autoradiographed for analysis. Supershift experiments were performed by adding 1 pg of an anti-Smad3, Fos, or Jun antibody to the reaction mixture 20 minutes before electrophoresis. XVI. Transient Transfections Transient cell transfections of mouse T cell lymphoma EL—4 and human T cell leukemic Jurkat cells were performed using the Cytofectene Transfection Reagent (Bio- Rad, Melville, NY). Cells were resuspended (EL-4, 2 X 105 cells/mL; Jurkat, 5 X 10° cells/mL) in fresh medium (1X RPMI) containing 5% BCS and left overnight at 37°C. Cells were harvested, washed, resuspended (2 X 105 cells/mL) in transfection buffer (Bio-Rad) and incubated for one hour at 37°C. The pSV-B-galactosidase control vector was co-transfected in every experiment to monitor transfection efficiencies. The transfected cells were transferred to a-CD3 + a-CD28-pre-coated plates, treated with TGF-[3,, and incubated for an additional 23 hours at 37°C. Following incubation, supematants were collected and stored at -80°C prior to use. Cells from each well were Page - 80 lyz of St 1C1. lyzed for B-galactosidase activity. Briefly, cells were washed twice with PBS, and 50 pl of IX Reporter Lysis Buffer (Promega) was added to each well of a 96 well plate. Samples were mixed by pipetting and incubated at room temperature for 15 minutes while rocking. Plates were centrifuged at 200 x g for 5 minutes to remove cellular debri and cleared lysates were transferred to clean eppendorf tubes and stored at -80°C prior to being assayed. SEAP activity was quantified using a chemiluminescent detection kit (Clontech). Briefly, 60 pl supernatant from each well of a 96-well plate was pipetted into a clean eppendorf tube and incubated for 30 minutes at 65°C in a water bath. Samples were cooled by placing them on ice for 5 minutes, and then equilibrated to room temperature. A 60 pl aliquot of SEAP assay buffer (Clontech) was added to each sample and incubated at room temperature. Following 5 minutes, 60 pl of 62.5 pM CSPD (Clontech SEAP detection) substrate was added for 60 minutes at room temperature. Luminescence was quantified for 15 seconds using a Turner TD-20e luminometer. The luciferase activity was normalized using the B-galactosidase activity. Briefly, 50 pl of 2X B-galactosidase assay buffer (Promega) was added to 50 pl cell lysate. Samples were mixed by pipetting, covered, and incubated overnight at room temperature on a rocker. Absorbance of the samples was read at 450 nm for 15 seconds using a Turner TD-20e luminometer. XVII. In vitro AFC (Antibody Forming Cell Response) Spleens from untreated mice were isolated aseptically and made into a single cell suspension. The splenocyte suspension was adjusted to 1 X 107 cells/mL (sRBC and DNP-Ficoll) or 5 X 10° cells/mL (LPS) in RPMI supplemented with 10% BCS Page - 81 1H) {-3— H u (Hyclone), 50 pM 2-ME, 100-units/mL penicillin, and 100 pg/mL streptomycin. Cell aliquots (500 pL) were transferred to a 48 well Costar culture plate (Cambridge, MA). Quadruplicate cultures were assayed for each treatment group. Five pl of vehicle (0.02 % PBS, pH 3.5, containing 0.1% BSA) or TGF-[3l was added directly to the respective wells of the 48 well plates just prior to antigen sensitization. Respective wells were sensitized with 6.5 x 106/well sRBC, 50 pg/well LPS (Sigma), or 50 ng/mL DNP-Ficoll. Cells were subsequently cultured in a Bellco stainless steel tissue culture chamber pressurized to 6.0 psi with a gas mixture consisting of 10% 02’ 7% C02, and 83% N2 for 5 days (sRBC and DNP-Ficoll) or 3 days (LPS). The culture chamber was placed on a rocking platform for the duration of the culture period. Antibody producing cells were enumerated by their ability to hemolyze intact sRBC, when sRBC were used as the sensitizing antigen, or TNP-haptenated sRBC, when the sensitizing antigen was DNP-FICOLL or LPS as previously described (Kaminski and Stevens 1992). Sheep erythrocytes were haptenated with trinitrophenol (TNP) using picryl sulfonic acid (Sigma) as the source of TNP as previously described (Kaminski and Stevens 1992). Briefly, 5 mL of sRBC were centrifuged at 400 x g for 10 minutes and supernatant was removed. The sRBC were resuspended in 20 mL cacodylate buffer containing 4.0% picryl sulfonic acid and incubated at 37°C with gentle rocking for 10 minutes. The sRBC were centrifuged at 400 x g for 10 minutes, supernatant was removed, and the erythrocytes were resuspended in 50 mL cacodylate buffer containing 0.8% glycylglycine. Following three washes with EBSS (Earles’ balanced salt solution), the TNP-haptenated sRBC were stored under sterile conditions at 4°C for up to one week. Page - 82 Results from quadruplicate cultures were expressed as the mean AFC/ 10° recovered viable splenocytes i SE. The pronase viability assay (as described below) was used to determine the number of viable splenocytes/well recovered following incubation. XVIII. Pronase Viability Determination Resuspended cells (100 pl) were added to an equal volume of pronase (225 proteolytic units/mL) (Calbiochem—Behring Corp., San Diego, CA) and incubated for 10 minutes at 37°C. Following incubation, the cell suspension was diluted with 10 mL of Isoton (Coulter, Addison, NJ) and counted on a Coulter counter. The percent viability was calculated by the following equation: (cell counts with pronase/cell counts without pronase) x 100 = viable cells. XIX. Densitometry The optical density of each treatment group was obtained using the Multi-Analyst program and a G8—700 imaging densitometer (BioRad, Hercules, CA). Using the density values, the ratio between the control and treated samples was calculated. The control group was designated with the value of 1.0 in order to assess qualitative changes between treatments. Alternatively, density was reported as optical value/mmz. XX. Statistical Analysis The mean i SE (standard arror) was determined for each treatment group in the individual experiments. Homogeneous data were evaluated by a parametric analysis of variance, and Dunnett's two-tailed t-test was used to compare treatment groups to the vehicle control when significant differences were observed (Dunnett 1955). Page - 83 For three times experiment experiment experiments experiment control. (31 experiment calculated 1- The result. determined ell‘eument 6 k4 For the temporal relationship studies, IL-2 ELISA experiments were replicated three times using triplicate wells for each individual experiment. The cell proliferation experiments were replicated five times using quadruplicate wells for each individual experiment. The values for both control and experimental groups were pooled across experiments to yield an n=3 and n=5 for the IL-2 protein secretion and cell proliferation experiments, respectively. Specifically, (1) the data were transformed to a percent of the control, (2) the mean i SE of the replicates from each experimental group with each experiment was calculated, and (3) the overall mean t SE across experiments was calculated from the individual means. The data are expressed as a percent of the control. The results are reported as the overall mean -t_- SE. Statistical significance was determined at the p < 0.05 level using the Dunnett's two-tailed t-test to compare each experimental group with the untreated control. Page - 84 EXPERIMENTAL RESULTS 1. Concentration- and time-dependent effects of TGF-B, on T cell proliferation and IL-2 expression TGF-B, has been reported to attenuate as well as stimulate T cell growth (Cerwenka et al. 1994; Kehrl et al. 1986c; Rich et al. 1996; Swain et al. 1991b). IL-2 is a cytokine that is principally recognized for its ability to function as an autocrine growth factor in promoting T cell activation, differentiation, and clonal expansion (Gomez et al. 1998; Waldmann et al. 1998). Attenuated IL-2 production has been associated with impaired T cell growth (Gomez et al. 1998; Waldmann et al. 1998). TGF-B, reportedly augments as well as attenuates IL-2 expression (Cerwenka et al. 1994; D'Angeac et al. 1991). The mechanisms for these seemingly disparate effects by TGF-[3l are unknown. The objective of this series of experiments was to test the hypothesis that the mode of T cell activation, the concentration of TGF—[3,, and the time of addition of TGF-[3l relative to T cell activation influence the effects of TGF-[3l on T cell growth and IL-2 expression. A. Effects of TGF-[3, on [3H]-thymidine incorporation in mouse splenic T cells and thymocytes As diagrammed in Figure 12, splenocytes and thymocytes were isolated from na’i‘ve B6C3F1 mice and activated in culture with 0t-CD3 + Ot-CD28 or or-CD3 alone. TGF-[3I was added directly to the cell cultures concurrently with T cell activation or at various intervals either prior to or after T cell activation. [3H]-thymidine incorporation was quantified following 72 hours of T cell activation. Interestingly, CD28 co- stimulation augmented a-CD3-induced splenic T cell growth (Figure 13a), but Page - 85 Figure 12. Experimental design for characterizing the effects of TGF-[3l on T cell activation and growth. The following variables were investigated: (1) target cell population, (2) mode of T cell activation, (3) TGF-[3l concentration, and (4) temporal relationship between T cell activation and addition of TGF-Bl to the cell cultures. Page -86 m cozahoabooc. o:.u.E>..._.....m new 5.39.98 «.4. 3.23.5 p 4 4 4 4 4 4 4 4 Homewfia summabfi wmmwr p E m I NF ..0 .m .0 .0 ... .o ..On 05930.... I n ..O .N ... .md .o .3 30930:. ...... Tau"... ......E d ...... w _ _ . _ _ E 8.5. Eflncfioéwvfl . ‘ Qo§es®flmmuwgoosgbfl a _ _ coza>=o< =00... c03a>zo< :00... «0892... 232.5 Pgnaw... 252.0“. 2:3an 5&0... Page - 87 Figure 13. Proliferation of activated mouse splenic T cells and thymocytes. Naive B6C3F1 splenocytes (a) or thymocytes (b) were cultured (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized (I-CD3 (2 pg/mL), immobilized a-CD3 (2 pg/mL) + soluble a-CD28 (l pg/mL), or soluble a-CD28 (1 pg/mL), as indicated. Following a 56 hour incubation at 37°C, [3H]-thymidine (1 pCi/mL) was added for a final 16 hour incubation. Cells were harvested and [3H]-thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as mean [3H]-thymidine incorporation (cpm) 1': SE for five experiments in quadruplicate. #,p<0.05 (determined by Dunnett’s t- test) as compared to the Na'i've group. *,p<0.05 (determined by Dunnett’s t-test) as compared to the immobilized (it-CD3 group. Page - 88 E [3H]-Thymidine Incorporation E [3H]-Thymidine Incorporation (cpm) x 104 (cpm) x 104 OO O o c? 4:. C.) N .0 10' 7.5- 5.0- 2.5- ' Splenocytes #,* T or-CDB or-CD3+ 0t-C028 oc-CDZB Thymocytes * T #,* 0t-CD3 (X'CD3+ OL-CDZ8 0t-C028 Page - 89 attenuated (x-CD3-activated thymocyte growth (Figure 13b). TGF-[3,, when added concurrently with T cell activation, inhibited Ot-CD3 + a-CD28-induced splenic T cell and thymocyte growth in a concentration—dependent manner (Figure 14). TGF-[3l also inhibited (x-CD3-induced splenic T cell and thymocyte growth in a concentration- dependent manner (Figure 14). The observed concentration-dependent inhibition of T cell growth by TGF-Bl was further demonstrated to be dependent upon the temporal relationship between T cell activation and addition of TGF-[3l to the cell cultures. Specifically, increasing the interval between T cell activation and addition of TGF-Bl to splenocyte cultures attenuated the inhibitory effect of TGF-Bl on (it-CD3 + or-CD28-induced (Figures 15a and 15b) and or-CD3-induced (Figures 16a and 16b) T cell growth. Interestingly, a bimodal time-dependent growth inhibitory effect by TGF-[3l was observed when TGF-[3l was added to cultures of naive splenocytes at varying intervals prior to T cell activation (Figure 16a). A similar bimodal time-dependent growth inhibitory effect by TGF-Bl was also observed when TGF-B, was added to or-CD3-activated splenic T cells (Figure 16b). These bimodal temporal responses were concentration-dependent with 1 ng/mL and 10 ng/mL TGF-Bl eliciting maximal responses (Figures 16b). Increasing the interval between T cell activation and addition of TGF-[3l to thymocyte cultures also attenuated the inhibitory effects of TGF-B, on (it-CD3 + 0t-CD28- induced T cell growth (Figures 173 and 17b). A modest concentration-dependent- bimodal temporal response was observed when TGF-Bl was added to thymocytes at various intervals after a-CD3 + a-CD28-induced activation (Figure 17b). Inhibition of a-CD3-induced thymocyte growth by TGF-B, was not markedly influenced by altering Page - 90 Figure 14. Concentration-dependent effect of TGF-[3l on proliferation of activated splenic T-cells and thymocytes. Naive B6C3F1 splenocytes or thymocytes were cultured (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the presence of immobilized a-CD3 (2 pg/mL) or immobilized a-CD3 (2 pg/mL) + soluble 0t-CD28 (1 pg/mL), as indicated. TGF-[3l (0, 10", 102, 10", l, or 10 ng/mL) was immediately added directly to the cultures. Following a 56 hour incubation at 37°C, [3H]-thymidine (1 pCi/mL) was added for the final 16 hour incubation. Cells were harvested and [3H]-thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as a percentage of vehicle control [or-CD3 + 0 ng/mL TGF-B, (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture or a- CD3 + OL-CD28 + 0 ng/mL] t SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control values (mean :1; SE): a-CD3 + OL-CD28 activated splenic T—cells, 7.2 X 104 i 4.5 X 103 cpm; OL-CD3-activated splenic T-cells, 4.1 X 104 i 0.9 X 103 cpm; 0t- CD3 + or-CD28-activated thymocytes, 1.4 X 104 i 1.6 X 103 cpm; oc-CD3-activated thymocytes, 8.0 X 104 i 6 X 103 cpm. The hatched box represents the vehicle values (mean i SE). Page - 91 [3H]-Thymidine Incorporation (Percent of Control) —I—— a-CD3-activated Splenocytes ----- ew- oc-CD3 + or-CDZ8-activated Splenocytes 125 _l O--- or-CD3-activated Thymocytes "-4-" a-CD3 + a-CD28-activated Thymocytes 100 _ I l l -3 -'2 -1 o 1 Log TGF-[31 (ng/mL) Page - 92 Figure 15. Time of addition effect of TGF-B, on (l-CD3 + a-CD28-induced splenic T cell proliferation. Naive B6C3F1 splenocytes were cultured (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized a-CD3 (2 pg/mL) + a-CD28 (1 pg/mL) as indicated. TGF-[3l (0, 10“, 102, 10", 1, or 10 ng/mL) was added directly to the cultures. TGF-[3I was added 3, 2, 1, 0.5, or 0 hours prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b) as indicated. Following a 56 hour incubation at 37°C, [3H]-thymidine (1 pCi/mL) was added for the final 16 hour incubation. Cells were harvested and [3H]-thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as a percentage of vehicle control [0t- CD3 + (it-CD28 + 0 ng/mL TGF-Bl (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] i SE for five experiments in quadruplicate. *,p<0.05 (determined by Dunnett’s t—test) as compared to the appropriate vehicle control group. Vehicle control value (mean 1- SE): (it-CD3 + a-CD28 activated splenic T-cells, 7.2 X 104 i 4.5 X 103 cpm. The hatched box represents the vehicle values (mean i SE). Page - 93 £43 454...... F -..-...? £43 ._.Emc a... ....... e ....... ...-“655.22 :I: ...-“655493 55.0 5.555358... |n_.| 4.505 859.84. =8 P 3 9:29. :IE. .0 8:62 .6 as: N. o o m o ._. ~_- m- 4 -mN .. .l: ._. V.A *.\\ * ._. x... [Om ... ... ... ........ I . ...... M .. Ill 4 a . W. ............ a .L .4... w .../...“. e ........... .. .. fl ...... - mm .. ...uuuuuuv. ...... ...._. ooooooooooooooooo ._. l ...-al....auuuuuvi i .1 ll ................... i‘ iiiiiiiiiiiiiiii s\\\\\\ \\\\\\\\ \\\\\\\\\\\\\.\ \ ........ \\\s 8. I ...................... ._. ...... t. l__.|._... meF 5.5958 :8 ._. 6.5.3 cobm>zum ..mo ._. 35.3 8885-38-.. + Bo... Bonuses: .. m8-.. pots page ...-“.0... 3V 3 ._otn. BEE van-.8. Am. Page - 94 (10nu03 jo iuaaiad) uoneiodioaul autpiuMql-[HE] Figure 16. Time of addition effect of TGF-[3l on a-CD3—induced splenic T cell proliferation. Naive B6C3F1 splenocytes (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized a-CD3 (2 pg/mL) as indicated. TGF-Bl (0, 10", 102, 10", 1, or 10 ng/mL) was added directly to the cultures. TGF-Bl was added 3, 2, 1 0.5 or 0 hours prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b). Following a 56 hour incubation at 37°C, [3H]-thymidine (1 pCi/mL) was added for the final 16 hour incubation. Cells were harvested and [3H]-thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as a percentage of vehicle control [or-CD3 + 0 ng/mL TGF-B, (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] —'-" SE for five experiments in quadruplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control value (mean i SE): a-CD3-activated splenic T-cells, 4.1 X 104 i 0.9 X 103 cpm. The hatched box represents the vehicle values (mean i 8E). Page — 95 (b) TGF-[51 added after In I TGF-[31 added prior to 5.6-. 454.... . 5-55268. :i: 5.62665... ...e.... ----.13.... Eda-_- .._E\w: rod ....... ¢ ....... 5-5566: so... In] .265 8.6264. =8 F 8 9:63 £48. .6 8:63 .6 as: N. o 6 m o .- N- m- 9(- I l' 96 +14 I'll is -x- -|—-l 96* \ at 9(- I F? {I ' ' ' I ’ I II A, l . * """""" \ --- \ \S H 5.326: :8 ._. 6.83am 6836:7308 .36 82... SEE 3. . a s o 0 ~ '0 \ k a o \\\\ O ‘ | o 0v * \\\\\ ‘ | | 1 o d \\ 0 a. on ‘ |\“ c o \s eee a. co \ on. o \s 4 6“. O " ~‘ ‘ o v \ ‘ a o a o f____. l l 5% “N” l ln [\ . .... s\\\oow .0 \\ \\e\\\\\\t\ew ...-53‘ - m5 5.5956: ..ou .. 6.5.9 68365-808 2 6...... Bug £5... E (ionuog to iuaoiad) uoneJodJoouI autptuMqi-[HE] Page - 96 Figure 17. Time of addition effect of TGF-[3, on a-CD3 + a-CD28-induced thymocyte proliferation. Naive B6C3Fl thymocytes were cultured (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized Ot-CD3 (2 pg/mL) + a-CD28 (1 pg/mL) as indicated. TGF-B, (0, 10“, 102, 10", 1, or 10 ng/mL) was added directly to the cultures. TGF-[3l was added 3, 2, 1, 0.5 or 0 hours prior to T cell activation (a), or 0, l, 3, 6, 9, or 12 hours after T cell activation (b). Following a 56 hour incubation at 37°C, [3H]- thymidine (1 pCi/mL) was added for a final 16 hour incubation. Cells were harvested and [3H]~thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as a percentage of vehicle control [or-CD3 + OL-CD28 + 0 ng/mL TGF- B, (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] 1' SE for five experiments in quadruplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control value (mean 1 8E): Ot-CD3 + a-CD28-activated thymocytes, 1.4 x 104 i 1.6 x 103 cpm. The hatched box represents the vehicle values (mean i 8E). Page - 97 5-65565 ----«..-- 5.6556566 .. ..... o ....... 5-65568. --*: 5.63565 ..o 5...... 5.6556: .85 IDI .585. 8:98.. =8 5 8 9:66.. 5.3 6 8:62 .6 65.; N. o o m o F- - m- . . . . . . _ . O .L ... ... .\.\ua,w..vw .. ~ * I, :4. ....... *._. ._. *._..m« M... 6.. . mN VT 1’! * \\\\\ “\ I/ ll!!! *4- I—I "w .. .. *8. - 3%.. an‘ ........ ... i --- ....... H II I * * \sss ooooo llllll 1M“ (III! \ A ._I $33. owmna I”! 8 III r/ i—‘ * rN filer... coo 1” .. e . oo I m“ V \m \ 8.5938 3.6083. 5.626: 3.608.»... 68:65:38-6 + mow-6 6836:5308 + 80.6 65.6 .88.. 5.8 3. 3 68 88.. 5-65 .3 mm F (ionuog to tuaaied) uotieJodJoaul autptuMqi-[HE] Page - 98 the temporal relationship between T cell activation and addition TGF-Bl (Figures 18a and 18b). Noteworthy, TGF-[3l did not augment T cell growth under any T cell activation condition tested (Figures 15, 16, 17 and 18). In summary, these results establish that inhibition of a-CD3-induced splenic T cell and thymocyte growth by TGF- B, in vitro is dependent on the concentration of TGF-[3,, CD28 co—stimulation, and the temporal relationship between addition of TGF-B, and T cell activation. B. Effects of TGF-B, on IL-2 protein secretion and steady state IL-2 mRNA expression Splenocytes and thymocytes were isolated from naive B6C3F1 mice and activated in culture with (it-CD3 + (it—CD28 or 0t-CD3 alone. TGF—B, was added directly to the cell cultures concurrently with T cell activation or at various intervals either prior to or after T cell activation. Supernatants were collected 24 hours after T cell activation and analyzed for IL-2 protein secretion by ELISA. Total RNA was isolated from splenocytes activated in culture for 60 minutes. Steady state IL-2 mRNA was quantified using RT-PCR. Anti-CD3-induced IL-2 protein secretion was augmented with a—CD28 co- Stimulation in splenic T cells and thymocytes (Figures 19a and 19b, respectively). When a(ided concurrently with T cell activation, TGF-[3I either augmented or attenuated a-CD3 ‘5 a-CD28-induced IL-2 secretion from splenic T cells at M (femtomolar) and pM (picomolar) concentrations, respectively (Figure 20). TGF-B, (10'2 ng/mL) augmented OL~CD3-induced IL-2 protein secretion in splenic T cells (Figure 20). In contrast, TGF—[3l did not augment, but rather, attenuated IL-2 protein secretion in activated thymocytes in a Page - 99 Figure 18. Time of addition effect of TGF-[3l on a-CD3-induced thymocyte proliferation. Naive B6C3F1 splenocytes thymocytes were cultured (2.5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized a—CD3 (2 pg/mL) as indicated. TGF-B, (0, 10", 102, 10", 1, or 10 ng/mL) was added directly to the cultures. TGF-[3l was added 3, 2, 1, 0.5 or 0 hours prior to T cell activation (a), or 0, l, 3, 6, 9, or 12 hours after T cell activation (b). Following a 56 hour incubation at 37°C, [3H]-thymidine (1 pCi/mL) was added for a final 16 hour incubation. Cells were harvested and [3H]- thymidine incorporation was quantified using a beta scintillation counter. The data are expressed as a percentage of vehicle control [Ct-CD3 + 0 ng/mL TGF-[31 (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] i SE for five experiments in quadruplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control value (mean i 8E): Ot-CD3-activated thymocytes, 8.0 X 104 i 6 X 103 cpm. The hatched box represents the vehicle values (mean i SE). Page - 100 5.6545685 :i.-- 55.65 456: 5 ..--4.--- 5.6545655... 50.... Fnd—Pr ._....me Pod 25¢.55 5.65565 58.6 IUI .585 8:98.. =8 5 8 9:66”. 5-656 8:634 6 65.5 N.- o o m o 5.. N- m- * ._. *I :0 ._. * ..... ...-_- ._- *H ._. *1... *4: ....... . ..... P 5.58266 350E}: 6885-5-86 65.6 .626 5-...0... 3V 5.59.56: 3.62:2... 68365-808 2 6.... 66. 5.65 3 I I In C N ti 8 (ronuog to luaaiad) uoneJodJoaul auiptuMqL-[HE] 85 m~_. Page - 101 Figure 19. IL-2 protein secretion in activated mouse splenic T cells and thymocytes. Naive B6C3F1 splenocytes (a) or thymocytes (b) were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized or-CD3 (2 pg/mL), immobilized (1- CD3 (2 pg/mL) + soluble 0t-CD28 (1 pg/mL), or soluble a-CD28 (1 pg/mL), as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at -80°C until analyzed. IL—2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as mean IL-2 protein secretion (units/mL) i SE for three experiments in triplicate. #,p<0.05 (determined by Dunnett’s t-test) as compared to the Naive group. *,p<0.05 (determined by Dunnett’s t-test) as compared to the immobilized 0t-CD3 group. Page - 102 (a) 100 l 75 - 50- 25— IL-2 Secretion (U/mL) Splenocytes #, * l A 0" v _A O |L-2 Secretion (U/mL) 8 or-CD3 0VCD3+ 0t-C028 or-CDZB Thymocytes # NA (X-CD3 Ot-CD3+ oc-CD28 oc-CDZ8 Page — 103 Figure 20. Concentration-dependent effect of TGF-[3l on IL-2 protein secretion in activated splenic T cells and thymocytes. Naive B6C3F1 splenocytes or thymocytes were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized or-CD3 (2 pg/mL) or immobilized a-CD3 (2 pg/mL) + soluble Ot-CD28 (l pg/mL), as indicated. TGF-[3l (0, 10's, 10“, 10“, 102, 10", 1, or 10 ng/mL) was immediately added directly to the cultures, as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at -80°C until analyzed. IL-2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as a percentage of vehicle control [or- CD3 + 0 ng/mL TGF—[3l (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture or OL-CD3 + a-CD28 + 0 ng/mL TGF-Bl] *5 SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t—test) as compared to the appropriate vehicle control group. Vehicle control values (mean i 8E): a-CD3 + OL-CD28 activated splenic T-cells, 74.5 i’ 4.2 U/mL; a-CD3-activated splenic T-cells, 37.5 i 3.5 U/mL; a- CD3 + OL—CD28-activated thymocytes, 9.0 i 4.2 U/mL; 0t-CD3-activated thymocytes, 4.2 i 2.1 U/mL. The hatched box represents the vehicle values (mean i SE). Page - 104 200. 150- * :5 T T * 1 i] I .............. I» 100 IL-Z Secretion (Percent of Control) 8 ~21 ///. / // /.4 [Ii-...}:Zwm/f’. "5. , /////j. n ........ V.‘ ..... T _L ‘ ~. _ X,“ 9 * g .. .o, W a“. N:- * ...; ......... —-l— a-CDB-activated Splenocytes “\ >1: ------ 0""- a-CD3 + a-CDZB-activated Splenocytes 9 ----O- a-CD3-activated Thymocytes MA- a-CD3 + a-CDZB-activated Thymocytes J5 -4 -3 -2 -1 o 1' Log TGF-B1 (ng/mL) Page - 105 concentration—dependent manner (Figure 20). TGF-B, did not induce IL-2 secretion in the absence of T cell activation (Figure 21). Simultaneous addition of TGF-Bl and T cell activation also resulted in either augmentation or attenuation of a-CD3 + (It-CD28- induced steady state IL-2 mRNA expression by TGF-[3l in splenic T cells at M and pM concentrations, respectively (Figure 22). The bimodal concentration-dependent stimulatory and inhibitory effects of TGF— B, on 0t-CD3 + 0t-CD28-induced IL-2 secretion was further dependent on the temporal relationship between T cell activation and addition of TGF-[3l to cell cultures-2 Augmentation of IL-2 protein secretion by TGF-[31 was abrogated when TGF-[31 was added either 30 minutes prior to or 30 minutes after T cell activation (Figures 23a, 23b 243, and 24b). Notably TGF-[3l (1 ng/mL and/or 10 ng/mL) stimulated splenic T cell IL-2 secretion when TGF-[3l was added at 2 or 3 hours prior to T cell activation (Figure 23a and Figure 243). A similar response by TGF-[3l was not observed when added at various intervals after the T cells were activated (Figure 23b and 24b). In contrast to splenic T cells, the effect of TGF-[3l on IL-2 secretion in 0t-CD3 + Cl-CD28-activated thymocytes was relatively time-independent (Figure 25). However, as ilIustrated in Figure 26, a marked temporal response was observed in 0t-CD3-activated thYtnocytes. In summary, these data demonstrate TGF-[3l differentially regulates IL-2 Seeretion in vitro in a manner that is dependent on concentration, CD28 co-stimulation, and the activation status of the T cells. In addition, TGF-[3l differentially regulates IL-2 SecI‘etion in splenocytes and thymocytes, suggesting that T cell maturation is also a faCtor; however, secondary indirect B cell and/or macrophage-mediated effects in SPlenocyte cultures may contribute to this differential response by TGF-[3,. Collectively, Page - 106 Figure 21. TGF-B, does not induce IL-2 protein secretion in naive splenocytes or thymocytes in vitro. Naive B6C3Fl splenocytes (a) or thymocytes (b) were cultured (5 X 106 cells/mL) in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation. Following a 24 hour incubation at 37°C, supematants were collected and stored at -80°C until analyzed. IL-2 protein secretion was quantified using stande sandwich ELISA. The data are expressed as mean IL-2 protein secretion (units/mL) i SE for three experiments in triplicate. Page - 107 100- lL-2 (Units/mL) 8 6‘. N U1 I (a) Splenocytes l 0... I I I 20 _a 01 I IL-Z (Units/mL) 8 U1 I N. 0-5.2.1353. o1 LLog TGF-B1 (ng/mL)—> (b) Thymocytes * -# NA 0-5-4-5-2-1‘01 LLog TGF-B1 (ng/mL)—> Page - 108 Figure 22. Concentration-dependent effect of TGF-B, on IL-2 mRNA. Splenocytes (5 X l0° c/mL) were added to culture plates pre-coated with a-CD3 + 0t- CD28. TGF-[3I was immediately added thereafter for one hour. Total RNA was extracted for RT-PCR analysis for IL-2 mRNA expression as described in “Materials and Methods”. IL-2 expression without treatment is shown (NA). The data are reported as the mean :1; SE (n=3). *p< 0.05 with comparison to the a-CD3 + (x-CD28 + 0 ng/mL TGF-B, group (VH) . Page - 109 L‘_ 4 0 .m- i//////////////////.. a ._A _ N 3 2 1 0 225. 665 6: 8:52 x 8.5. .4sz N-.__ Figure 23. Time of addition effect of TGF-[3I on (II-CD3 + a-CD28-induced IL-2 protein secretion in splenic T cells. Naive B6C3F l splenocytes were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized 0t-CD3 (2 pg/mL) + soluble 0t- CD28 (1 pg/mL), as indicated. TGF-[3l (0, 103, 102, 10", l, or 10 ng/mL) was added directly to the cultures. TGF-Bl was added 3, 2, 1, 0.5, or 0 hours prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b) as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at -80°C until analyzed. IL-2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as a percentage of vehicle control [Ct-CD3 + a-CD28+ 0 ng/mL TGF-[3l (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] 1' SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control values (mean i SE): a-CD3 + 0t-CD28 activated splenic T-cells, 74.5 i 4.2 U/mL. The hatched box represents the vehicle values (mean i SE). Page—111 lim— 55.505 ..::mc 5 ----4.--- 55-50.. ..::mc 5.0 .-.-.256 5-65.5685 --I: 5.65.565 5... .....0: 5.65456: so... IDI .5505 56.59.64. :8 ._. 3 9.5.6.6”. £50... .6 56.5.63. .o 65:5... N5 o o m o 5- N- m- *4. .1 * .... ... - - I Om * * Jw V.“ I . ... H ._. V.A wow. +- elum'elw . . . /. ... \;v. \ 005 \ \ \ .05 I H... ,, 5 58 56.59.68 :8 5 6.6.5.6 . 56.6368 :8 5 66.3% 8685586-... 86-6 4 6885586-? 86-6 .. 65m .628 5.5.. An: B 6.6. .626 55-50... Am. fiOmN Page- 112 (101mm to iuaalad) uonaiaas z-‘n Figure 24. Time of addition effect of TGF-B, on aCD3-induced IL-2 protein secretion in splenic T cells. Naive B6C3F] splenocytes thymocytes were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized (it-CD3 (2 pg/mL) + soluble or- CD28 (1 pg/mL), as indicated. TGF-B, (0, 10", 102, 10“, l, or 10 ng/mL) was added directly to the cultures. TGF-B, was added 3, 2, l, 0.5, or 0 hours prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b) as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at —80°C until analyzed. IL-2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as a percentage of vehicle control [or-CD3 + a-CD28+ 0 ng/mL TGF-B, (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] 1 SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control values (mean 3 8E): (X-CD3 + or-CD28-activated thymocytes, 9.0 i 4.2 U/mL. The hatched box represents the vehicle values (mean i SE). Page - 113 5-50... ..ER: 5 ----4.--- Emu-.....tR: 59o ....... ¢ ....... 5.65.5685 :i.-- 5.65.56: to ...-:0 5.6556: 58.8 |D| 3505 56.59.64. :8 ._. 0. 9.52% 55-5.0... .0 56.5.63. .o 6E...- N5 o o m o 5- - m- . . . . . _ _ Id..- I_l filllll I I1 \ ae\\\ .sa..§.§\ .I.‘ r I 0 vw ax *‘III! 56.6368 :8 5 6.5.6.9. 56.59.66 :8 5 6.56.3 6885- 86.6 6885-86... 6:: 6628 5.55 An: 8 Sta cocoa 55-50.. Amv tomN ('IOJIUOZ) IO IUGDJad) UOIIGJDGS Z-'|| Page - 114 Figure 25. Time of addition effect of TGF-[3l on Ot-CD3 + a-CD28-induced IL-2 protein secretion in thymocytes. Naive B6C3Fl splenocytes were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized a-CD3 (2 pg/mL) as indicated. TGF-[3I (0, 103, 102, 10", 1, or 10 ng/mL) was added directly to the cultures. TGF-[3l was added 3, 2, 1, 0.5, or 0 hour prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b) as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at -80°C until analyzed. IL-2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as a percentage of vehicle control [Ct-CD3 + 0 ng/mL TGF-[3l (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] St SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. Vehicle control values (mean 1- SE): (X-CD3 + a-CD28-activated thymocytes, 9.0 i 4.2 U/mL The hatched box represents the vehicle values (mean i SE). Page- 115 5.65565 ----4.--- 5.65456: 6.6 -..-....-.6 5.655685 --I: 5.655565... 5.5. 5.6545658... Ifll .585 8:562 =8 5 3 9:665. 5.65 6 8:62 6 65.5 N. a e m o 5- N- __ M l III \ uuuuuuuuuuuuuuuuuuuuuu \- . \ \\ \ \ \ . exx 46.59.5566 34608....“ 683.65-35.08 + Moo-6 6% 86.6 5-65 3. 85?..qu 8460.555 68365-388 + 80.6 B 6.6 86.6 5.65 i—l-He N\\\\ .x...\\\\ “\6 ....§\ ..6\. u\\\x\\\.®\i\i§§§n\ mN 5 .8 8. (Ionuog Io iuaalad) uonaiaas 2"" Page - 116 Figure 26. Time of addition effect of TGF-B, on OL-CDB-induced IL-2 secretion in thymocytes. Naive B6C3F] thymocytes were cultured (5 X 106 cells/mL in RPMI 1640 supplemented with 5% BCS) in the absence of exogenous stimulation (Naive) or in the presence of immobilized Ot-CD3 (2 ug/mL) as indicated. TGF-Bl (0, 103, 102, 10", l, or 10 ng/mL) was added directly to the cultures. TGF—[3l was added 3, 2, l, 0.5 or 0 hours prior to T cell activation (a), or 0, 1, 3, 6, 9, or 12 hours after T cell activation (b) as indicated. Following a 24 hour incubation at 37°C, supematants were collected and stored at ~80°C until analyzed. IL-2 protein secretion was quantified using standard sandwich ELISA. The data are expressed as a percentage of vehicle control [oz-CD3 + 0 ng/mL TGF-B, (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] 1' SE for three experiments in triplicate. *,p<0.05 (determined by Dunnett’s t- test) as compared to the appropriate vehicle control group. Vehicle control values (mean i SE): oc-CD3-activated thymocytes, 4.2 i 2.1 U/mL. The hatched box represents the vehicle values (mean 1'- SE). Page - 117 5.555369 --i.-- amok 45m: F ----4:-- amoflémfid ...---.o ramp—LEE: 5.0 ..:-..:... gag-_Emfood |D| A285 85ng =8 F B 232$. 5-qu B 8:63. .6 25 NF o o m o T N- m- 5.59.58 330:5: 82.2388 5% 8% .38 5 .m M um m H .m O-Wv- w I. It :1 In n n I ..:. ......iuhh ........ ; .\ s song-sum 330255 88258.5 33.8 8%“ 5-qu A8 1 C n l m 5 Q. J u e (imiuog to iuaaiad) uonanas Z-‘II o9 Page - 118 these results support the hypothesis of c00perative signaling interactions among the TcR, the TBR, and CD28 to regulate IL-2 expression by TGF—Bl. The influence of TGF-[3l on IL—2 secretion does not parallel its modulatory effects on T cell growth under similar T cell activation conditions (Figure 27). These results reflect direct effects by TGF-[3l on other parameters associated with cell growth, for example cell cycle proteins and apoptosis. 11. TGF-[31 bifunctionally modulates in vitro IgM AFC responses in a concentration-dependent manner Picomolar concentrations of TGF-[3l inhibit in vitro IgM AFC responses (Delaney et al. 1994; Icon et al. 1997). In light of the observation that M concentrations of TGF- [31 augment IL-2 secretion (Figure 20), steady state IL-2 mRNA (Figure 22) and chemotaxis (Wahl et al. 1987), the objective of these studies was to determine whether N concentrations of TGF-B, also enhance in vitro DNP-Ficoll, sRBC, and LPS-induced AFC responses. Splenocytes were isolated from naive B6C3F1 mice and sensitized in culture with sRBC, DNP-Ficoll, or LPS in the presence of TGF-[3,. The effects of TGF- B, on T cell-dependent and T cell-independent immunoglobulin production were investigated using the hemolytic plaque assay. TGF-B, augmented and attenuated the in vitro T cell-dependent anti-sRBC AFC response at M (0.4 — 2.4 pg/mL) and pM (0.001 — 0.005 ng/mL) concentrations, respectively (Figure 28). Concentration-dependent augmentation as well as attenuation by TGF-B, was also observed in the in vitro T cell-independent DNP-Ficoll AFC response (Figure 29). In contrast, TGF-[3I attenuated the in vitro T cell-independent B Page-119 Figure 27. TGF-[il differentially regulates IL-2 secretion and [3Hl-thymidine incorporation in a-CDB + a-CDZS-activated splenic T cells in a concentration- and time-dependent manner. Data are abstracted and summarized from previously described results to illustrate the differential response by TGF-[3l on T cell activation, as determined by IL-2 protein secretion, and T cell proliferation, as determined by [3H]- thymidine incorporation. The data are expressed as a percentage of vehicle control [a- CD3 + 0 ng/mL TGF-[3l (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture] i SE for five experiments in quadruplicate. *,p<0.05 (determined by Dunnett’s t-test) as compared to the appropriate vehicle control group. The hatched box represents the control values expressed as a percentage (mean -_t-_ SE). Page - 120 :38 ER: 55 £25985 mEEEfi-Efl ...-.0 :IE ._Emc Ed ”comma-cm N2 lul E85 Sea-5:2. :8 F 8 952mm 5-qu .6 8:62 so 2:: NF o o m o T N- m- \\ $- H“ ..r ............................ w ....... ._- I cozmzuum :8 .r conmzuum :8 ._. umUDUcTwNQU-d + moo-6 nmusucTwNQUS + mouse Eta totem van-Eh 3v - 8 Sta umnum EEO-_- Amy rlom \\\\\-\\W\\\fl\\\\\\\\\\\\\\.\\\\\\\\\ o2 -mNF -omF F: ]OJ1UO'_) to iuaDJad Page - 121 Figure 28. Effect of TGF-Bl on the in vitro AFC response to sRBC. Spleens from naive (NA) female B6C3F1 mice were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 1 X 107 c/mL and transferred in SOO-uL to the wells of a 48-well culture plate. Quadruplicate cultures were prepared with NA, vehicle (VH; 0.02 % PBS final concentration in culture, pH 3.5, containing 0.1% BSA) or TGF-[3l (0.0003, 0.0006, 0.001, 0.002, 0.005, 0.01, 0.02, 0.04, 0.08, 0.2, 0.3, 0.6, 1.25, or 2.5 ng/mL final concentration in culture), and sensitized with sRBC. Cultures were subsequently analyzed for a day 5 antibody response by enumerating the number of antibody forming cells (AFC); spleen cell viability and total recovered cells/culture were also determined. The results from quadruplicate determinations are expressed as the mean AFC per 106 recovered viable cells _-I_- SE (n=4). *p < 0.05 as determined by Dunnett’s t test as compared to the vehicle control. The results are representative of three independent experiments. Page - 122 m N _r * m.~ Ar m~.— *_ mNé 30.0 32 * 20 o 33.0 * _ 356 58.0 ) * Moe-o .| 38d um. 38.0 W. * 58 o 38.0 nm .. .l 385 m. . F / _ 38° m .. l. 356 m. _ 38.0 l | ( I :80 T wooed Rw- l . F. T 88.0 w. * _III 9.8 o rTu T 88.0 c * .| £86 1 o * T . l $8.0 .lrlbx/////////Ti l- Warn-pp... _| 88.0% cor x mwiuocmam was; nmum>oumm m .l 886 ..| o as from _l rvif/////// Pangaea oz _ _ _ w w w 0 2 8300283 8550mm cor CHE 10001 . 1 .- l C C A... D- .” oo- R M. .W in. «U.- ..n 0: .10. or“ H.- «J i. .rtit nU. O 0 u Page —123 Figure 29. Effect of TGF-[3l on the in vitro AFC response to DNP-Ficoll. Spleens from naive (NA) female B6C3F1 mice were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 1 X 107 c/mL and transferred in 500-uL to the wells of a 48-well culture plate. Quadruplicate cultures were prepared with NA, vehicle (VH; 0.02 % PBS final concentration in culture, pH 3.5, containing 0.1% BSA) or TGF—[3I (0.0003, 0.0006, 0.001, 0.002, 0.005, 0.01, 0.02, 0.04, 0.08, 0.2, 0.3, 0.6, 1.25, or 2.5 ng/mL final concentration in culture), and sensitized with DNP- Ficoll (50 ng/mL). Cultttres were subsequently analyzed for a day 3 antibody response by enumerating the number of AFC; spleen cell viability and total recovered cells/culture were also determined. The results from quadruplicate determinations are expressed as the mean AFC per 10‘5 recovered viable cells :t SE (n=4). *p < 0.05 as determined by Dunnett’s t test as compared to the vehicle control. The results are representative of three independent experiments. Page - 124 .I: :— . ‘4 _ q .2 0 mNé mNod mNFm-o Moe-0 000.0 08.0 more-o (O i E TGF-[31(ng/mL) v~oo.0 N 50.0 0000.0 800.0 0 P-Ficoll lk/////////Tllm 0 6 5 4.. 3 2 1 09 x mmiuocwfim $00; 080500”. oll. Splttllj 0 wk cl transfer?” re prepred rm . tom-ill l. 0.08.0 1000} 750- *w k... *w *— *~ *0 8.520500 02 0 O 5 2 mA////////v w 0350033 020500”. 03 B“? with D)?- espmxc “1 clls/cullu’f (pressed 4‘ :rmlned b} ire of [hm ON a; 20.0 250 $20 58.0 ) .L 38.0 m / g .350 m. 0000.0 Rw a $00.0 T :80 E000 n0000.0 800.0 NP-Ficoll C Page - 125 cell polyclonal mitogen AFC response in a concentration- dependent manner (Figure 30). Augmentation of AFC responses by TGF-Bl, as described above, was not due to an increase the number of viable cells recovered (see inserts for Figures 28, 29, and 30). In summary, these results demonstrate that TGF-[3l modulates T cell-dependent and T cell- independent AFC responses in a bifunctional concentration manner. The observed profile of impaired immunoglobulin production is highly suggestive of a cell-type specific response. 111. Characterization of Smad proteins in mouse lymphoid tissue A. Smad2, Smad3, and Smad4 protein expression in mouse splenocytes and thymocytes Whole cell lysates from splenocytes and thymocytes isolated from naive B6C3F1 mice were used to investigate Smad protein expression in lymphoid tissue by Western immunoblot analyses. As illustrated in Figure 31, Smad2 Smad3, and Smad4 are expressed in mouse splenocytes and thymocytes. Importantly, the splenocyte preparation represents a mixed population of B cells, T cell, macrophages, and dendritic cells. Thymocytes are devoid of B cells, macrophages, and dendritic cells, and represent a relatively pure preparation of immature T cells. Thymocytes were employed in this set of experiments specifically to determine the expression pattern of Smad proteins in T cells. Polyclonal antibodies recognizing proline-rich linker regions revealed a 62 kD band corresponding to Smad2 (Figure 31a), 56 kD band corresponding to Smad3 (Figure 31b), and a 68 kD band corresponding to Smad4 (Figure 31c). A second antibody against Smad4 demonstrated less cross-reactivity with the other Smads (Figure 31d). Page - 126 Figure 30. Effect of TGF-B, on the in vitro AFC response to LPS. Spleens from naive (NA) female B6C3F1 mice were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 5 X 106 c/mL and transferred in 500-111. to the wells of a 48-well culture plate. Quadruplicate cultures were prepared with NA, vehicle (VH; 0.02 % PBS, final concentration in culture, pH 3.5, containing 0.1% BSA) or TGF-[3l (0.0003, 0.0006, 0.001, 0.002, 0.005, 0.01, 0.02, 0.04, 0.08, 0.2, 0.3, 0.6, 1.25, or 2.5 ng/mL final concentration in culture), and activated with LPS (10 ug/mL LPS). Cultures were subsequently analyzed for a day 3 antibody response by enumerating the number of AFC; spleen cell viability and total recovered cells/culture were also determined. The results from quadruplicate determinations are expressed as the mean AFC per 106 recovered viable cells 1- SE (n=4). *p < 0.05 as determined by Dunnett’s t test as compared to the vehicle control. The results are representative of three independent experiments. Page - 127 Tl 2 _r * .Ir 2.— i 03.0 05.0 * .I 32.0 * .I 08.0 Tl 08.0 m * _ 005.0 m. * . 0000.0 nw * T . 400.0 .| 200.0 T $00.0 TI *TI ..| 800.0 TI. m000.0 .l 0 1 figggga _i 3 2 000 x 8:60:83 $00.5 0205000 Page - 128 _I. a/////////// 00.020500 oz LPS _ O 0 0 5 5 2 8:60:93 080>oumm 000 \0.._< 1000 - 750 - Figure 31. Smad2, Smad3, and Smad4 protein expression in mouse splenocytes and thymocytes. Whole cell protein lysates obtained from splenocytes, leLu cells, COS-1 cells transfected to overexpress Smad4, or COS-1 transfected with the vector control for the Smad4 gene. Protein (25 1.1g) was loaded in each lane, resolved on a 10% non-denaturing SDS-PAGE gel, transferred to nitrocellulose, and incubated with 2 ug/mL anti-Smad2 (a), anti-Smad3, H-2 (b), anti-Smad4, B-8 (c), or anti-Smad4, S-20 ((1). lg horseradish peroxidase-linked secondary antibodies were used for protein detection using the ECL system Page — 129 " v c \ c,Q\ {06‘ ‘gX (a) Smad2 —> <62kD 1 2 3 c \9 \’~‘ 6‘ "x ‘09 4K“ ‘0 (b) Smad3 —> ‘40“) 1 2 3 (C) Smad4 —>» ..., Page - 130 Smad4 specificity was further characterized by comparing whole cell splenic and thymic extracts with whole cell extracts from Smad4 transfected COS-1 cells (Lane 4 in Figure 31c and Lane 1 of Figure 31d). The Smad3 antibody modestly cross-reacts with Smad2, as suggested by a second slower migrating band (Figure 31b, lanes 1, 2, and 3). Cell lysates from leLu mink epithelial cells, commonly used as a positive control in the in vitro TGF-[31 bioassay, were incorporated as a comparative TGF-B,-responsive control in this set of experiments. B. Concentration-dependent effects of TGF-[3l on nuclear expression of Smad3 in mouse splenocytes Having confirmed constitutive Smad protein expression in splenocytes, the next series of experiments investigated the responsiveness of splenic Smad3 to TGF-[3,. Splenocytes were isolated from naive B6C3F1 mice, cultured in the presence of TGF-[3l for 60 minutes, and nuclear lysates were isolated for Western blot analyses. TGF-B, enhanced Smad3 nuclear protein expression in a bimodal, concentration-dependent manner (Figure 32a). Maximal nuclear expression of Smad3 was demonstrated in response to a 10‘2 - 10'3 ng/mL TGF-[3l treatment. A mechanistic understanding of how Smads are regulated once they enter the nucleus is required to determine whether these results reflect increased Smad3 nuclear translocation or enhanced nuclear stabilization. C. Temporal response of TGF-Brinduced activation of Smad3 in mouse splenocytes The temporal response of Smad3 activation by TGF-B, was investigated. Splenocytes were isolated from naive B6C3F1 mice, treated with TGF-I3, (10'3 ng/mL or Page- 131 Figure 32. Concentration- and time-dependent effect of TGF-B, on nuclear expression of Smad3 in mouse splenocytes. Splenocytes (5 X106 c/mL) were treated with (a) increasing concentrations of TGF-Bl for 60 minutes, (b) 10’3 ng/mL TGF-[3l for 0, 5, 15, 30, 60, 90, 120, and 240 minutes, or (c) 10 ng/mL TGF—[3l for 0, 5, 15, 30, 60, and 90 minutes. Nuclear proteins (25 pg) were isolated and resolved on a 10% denaturing polyacrylamide gel. Proteins were transferred to nitrocellulose and incubated with a mouse polyclonal Smad3. An anti-mouse Ig horseradish peroxidase-linked secondary antibody were used for protein detection using the ECL system. The fold induction of Smad expression is shown as intensity relative to vehicle control (0 ng/mL TGF-[3,; 0.02 % PBS, pH 3.5, containing 0.1% BSA). Page - 132 (a) 60 min TGF-B1 (ng/mL) .4. 0..1,9'§...1..9;:19.-.319.'..2...19711,10, .: ‘ ""FTO’ '.'.’“‘.“.'.' tr‘r *xr'qufm'c u .\ 5.61‘1214‘133 {3'7 179 353" 145 S4 (b) TGF-B1 (10'3nglmL1 1mm) 0 5 15 30 60 90 120 240 Smad3-Ir» ...“... ., ...-I hid u M tum-d ...... ..: 1.00 0.94 1.21 1.58 2.24 1.12 1.05 1.11 (C) TGF-[11 (10 nglmL) (min) 0 5 15 30 60 90 Smad3-- G... hunt in... ...-1&1 .... . \J~ 1.00 1.11 1.14 1.46 1.03 1.54 Page —133 10 ng/mL) for 0, 5, 15, 30, 60, 90, 120, or 240 minutes, and nuclear lysates were isolated for Western blot analyses. Increased nuclear expression of Smad3 protein by TGF-[31 (10'3 ng/mL) was detected at 15 minutes (Figure 32b). Peak nuclear Smad3 expression was observed at 60 minutes and near basal expression levels were detected 240 minutes after treatment. Ten ng/mL TGF-B, only modestly increased Smad3 nuclear expression (Figure 32c). IV. Pathology of Smad3-null mice Smad3-null mice were used to investigate a role for Smad3 in modulation of IL-2 expression by TGF-[3,. Mice homozygous for a null mutation in the gene encoding Smad3 (Smad3"’) are phenotypically indistinguishable from their litterrnate controls at birth, except 70% of Smad3"' are smaller prior to weaning (Yang et al. 1999). Usually within 3 months of age, these Smad3"’ mice develop a wasting syndrome that is characteristically associated with multifocal formation of pyogenic abscesses, kyphotic posture, enlarged lymph nodes, reduced spleen weight, and an involuted thymus (Yang et al. 1999). A. Body and organ weights Grossly asymptomatic Smad3"’ mice were used in these studies as demonstrated by an absence of periorbital and peridontal pyogenic abscesses, absence of kyphotic posture, and similar body weight relative to littermate controls. Collectively, the Smad3"' mice used in these studies were of slightly lower body weight than the Smad3“+ littermates (Figure 33). The Smad3"‘ mice were further examined to ascertain any differences in gross organ weights. Organ weights are reported as tissue weight (Figure Page - 134 Figure 33. Effects of targeted deletion of Smad3 on body and organ weights. Animal body, spleen, thymus, kidney, heart, and liver weights were recorded at necropsy from the Smad3-null mice and age-matched wild type littermates used in these studies. All mice were aged 6-9 wks. The data are reported as (a) tissue weight (g) or (b) tissue weight (g)/body weight (g) and are expressed as the mean i SE (n = l2/group). Page - 135 (a) Weight (g) (b) Tissue (g)/ Body Weight (g) 307 20— I Wild Type Smad3 Null T a: T 10- 0 _ L Whole Body 0.0151 0.01- 0.005- 0.47 0.3 - 0.2 - 0.1- o _ Spleen Thymus Kidney Heart I Wild Type Smad3 Null L Spleen Thymus Kidney Heart Page - 136 0.06 7 0.05 _ 0.04- 0.03 — 0.02 — 0.01 - 00 33a) or as tissue weight per total body weight (Figure 33b) to account for the difference in body weights between the Smad3"‘ and Smad3“+ mice. When compared on a tissue per body weight basis, spleen, kidney, and liver weights were similar; however, thymus and heart weights were elevated in the Smad3"' mice (Figure 33b). B. Thymic and splenic cellularity A more extensive analysis was conducted on the immune target organs of interest by assessing spleen and thymic cellularity. Single cell suspensions of splenocytes and thymocytes were obtained and counted on a per individual animal basis, and reported as either cells per mouse or cells per total body weight in order to compensate for the smaller size of the Smad3"' mice (Figure 34). Interestingly, thymic cellularity was increased in the Smad3"’ mice relative to littermates suggesting a putative role for Smad3 in thymic T cell development. V. A role for Smad3 in the inhibition of IL-2 expression by TGF-B, A. Inhibition of IL-2 protein secretion by TGF-[3, is attenuated in a-CD3 + oc-CD28-activated Smad3-null splenic T cells and thymocytes The objective of these studies was to investigate a role for Smad3 in the inhibition of IL-2 protein secretion by TGF-Bl in activated T cells. Splenocytes and thymocytes were isolated from Smad3"' mice and activated in vitro with a-CD3 + Ot-CD28. Following 24 hours in culture, supematants were collected and IL-2 protein secretion was quantified by ELISA. TGF-Bl attenuated IL-2 protein secretion in 0t-CD3 + 0t-CD28- activated splenic T cells and thymocytes from wild type littermate control mice Page - 137 Figure 34. Splenic and thymic cellularity in Smad3-null mice. Cell counts were obtained from single cell isolations of splenocytes and thymocytes from asymptomatic Smad3-null mice (aged 6-9 wks of age) and age-matched wild type littermates. The data are reported as (a) cells X 108/animal or (b) cells X 106/ body weight (g) and are expressed as the mean -_l-_ SE (n = 12/group). Page - 138 I Wild Type I /Smad3 Null 2.0- T 1.5- 1.0- 0.5— 0- Splenocytes Thymocytes 251/ Cells x 108/Animal (b) 10' 0 I Wild Type ASmad3 Null % O- Splenocytes Thymocytes SJ! >1 O U1 1 1" U1 1 Cells X106/Body Weight Page - 139 (Smad3*’*) in a concentration-dependent manner (Figure 353 and 36a, respectively). Inhibition of IL-2 protein secretion by TGF-Bl was markedly attenuated in a-CD3 + 0t- CD28-activated Smad3"’ splenic T cells and thymocytes (Figure 35a and 36a, respectively). Notably, the magnitude of 0t-CD3 + 0t-CD28-induced IL-2 secretion was markedly reduced in SMAD3"’ thymocytes relative to SMAD3”+ thymocytes (Figure 36b). IL-2 protein secretion was detected in supernatants from untreated SMAD3” splenocytes (Figure 35b). The magnitude of a-CD3 + a-CD28-induced IL-2 secretion was also elevated in splenocytes from SMAD3"’ relative to SMAD3“ mice (Figure 35b). In contrast, IL-2 protein secretion was not elevated in untreated SMAD3"’ thymocytes. Collectively these results suggest a role for Smad3 in regulating IL-2 protein secretion by TGF-B, in vitro. These results also suggest that Smad3 may play a role in regulating IL-2 protein expression in vivo. A. Inhibition of steady state IL-2 mRNA by TGF-B, is attenuated in activated Smad3-null splenic T cells and thymocytes Having established a role for Smad3 in IL-2 protein secretion, the next objective was to investigate a putative role for Smad3 in regulating IL-2 transcription. Splenocytes and thymocytes were isolated from Smad3"' mice and activated in vitro with 0t-CD3 + 0t— CD28. Following one hour in culture, total RNA was isolated for quantitative RT-PCR. TGF-B, inhibited steady state IL-2 mRNA expression in a-CD3 + a-CD28-activated splenocytes and thymocytes from Smad3“ mice in a concentration—dependent manner (Figure 37a and 38a, respectively). Inhibition of oc-CD3 + a-CD28-induced steady state IL-2 mRNA by TGF-[3l was abrogated in Smad3"‘ splenic T cells and thymocytes (Figures 37a and 38a, respectively). Steady state IL-2 mRNA expression was detected Page - 140 Figure 35. Inhibition of IL-2 protein secretion by TGF-B, is attenuated in a-CD3 + a-CD28-activated Smad3-null splenic T cells. Splenocytes were activated with immobilized oc-CD3 (2 ug/mL) + a-CD28 (1 ug/mL) in the presence (a) or absence (b) of TGF-Bl, as indicated for 24 hours at 37°C. IL-2 was determined by ELISA. The data are expressed as the mean :1; SE of three separate experiments of triplicate cultures. “a” denotes p < 0.05, SMAD3"’ significantly different from the SMAD3”. “b” denotes p < 0.05, SMAD3” significantly different from vehicle. “c” denotes p < 0.05, SMAD3"' significantly different from vehicle. N .D., IL-2 protein below the level of detection. Page - 141 SPLN SMA03+I+ (a) 30° 1 + SPLN SMAD3"- *T— -* N N 01 O 0‘ O O O l 1 I § 50.. lL-2 Protein (Units/mL) O I I I I I I I I I I I’I’II o 10“ 10'3 10'2 10'1 1.02.5 TGF-[31 (ng/mL) L— a—CDB + a-CD28 » (b) 2501 —E]— SPLN SMA03+/+ —I— SPLN SMAD3"- A T -J 200 — T E . . T ._.-"3 .L C 150 — :3 V E 100 (D _. 4—0 0 h D- 50 - Background ‘1‘ a —' N.D. T _ 0 _ None No Vehicle [ Treatment oc-CD3 + oc-CD28-> Page - 142 Figure 36. Inhibition of IL-2 protein secretion by TGF-B, is attenuated in a-CD3 + a-CDZS-activated Smad3-null thymocytes. Thymocytes were activated with immobilized 0t-CD3 (2 ug/mL) + a-CD28 (1 ug/mL) in the presence (a) or absence (b) of TGF-[3,, as indicated for 24 hours at 37°C. IL-2 was determined by ELISA. The data are expressed as the mean t SE of three separate experiments of triplicate cultures. “a” denotes p < 0.05, SMAD3"' significantly different from the SMAD3“. “b” denotes p < 0.05, SMAD3“+ significantly different from vehicle. “c” denotes p < 0.05, SMAD34‘ significantly different from vehicle. N.D., IL-2 protein below the level of detection. Page - 143 IL-2 Protein (Units/mL) E ’0 v IL-2 Protein (Units/mL) —O— THMC SMA03+/+ I I I l+ THMC SMAps-I- 20 15- 10- I I I I I I I I I I I’I’II o 104 10'3 10'2 10'1 1.0 2.5 TGF-B1 (ng/mL) L— a-CD3 + a—CDZB p —Cl— THMC SMA03+/+ —I— THMC SMAD3-F l T l l Background None Vehicle - oc-CD3 + oc-CD28—> Page - 144 Figure 37. Inhibition of steady state IL-2 mRNA by TGF-[3l is attenuated in activated Smad3-null splenic T cells. Splenocytes were activated with immobilized a- CD3 + a-CD28 in the presence (a) or absence (b) of TGF-Bl, as indicated for 2 hours at 37°C. IL-2 mRNA was analyzed by quantitative RT-PCR. The data are expressed as the mean -_I-_ SE of triplicate cultures from two separate experiments. “a” denotes p < 0.05, Smad3"' significantly different from Smad3”. “b” denotes p < 0.05, Smad3“+ significantly different from vehicle. “c” denotes p < 0.05, Smad3"' significantly different from vehicle. Page - 145 (b) lL-2 mRNA (Molecules x105/ 100 ng Total RNA) lL-2 mRNA (Molecules x102/ 100 ng Total RNA) [:1 SPLN SMAD3 +/+ I SPLN SMAD3 -/- b 0 2.5 5.0 l—TGF-B1 (ng/mL)» a—CDS + a-CD28___> Untreated SPLN DJ T i SMAD3+/+ SMAD3'/' Page - 146 Figure 38. Inhibition of steady state IL-2 mRNA by TGF-[3l is attenuated in activated Smad3-null thymocytes. Thymocytes were activated with immobilized a- CD3 + 0t-CD28 in the presence (a) or absence (b) of TGF-[3,, as indicated for 2 hours at 37°C. IL-2 mRN A was analyzed by quantitative RT-PCR. The data are expressed as the mean i SE of two triplicate cultures from two separate experiments. “a” denotes p < 0.05, Smad3”+ significantly different from vehicle. “b” denotes p < 0.05, Smad3"‘ significantly different from Smad3“. “c” denotes p < 0.05, Smad3"' significantly different from vehicle. Page - 147 (a) (b) IL-2 mRNA (Molecules x105/ 100 ng Total RNA) lL-2 mRNA (Molecules x102/ 100 ng Total RNA) 12— El THMC SMAD3 +/+ ‘I THMC SMAD3 -/- l o 2.5 5.0 l—TGF-Bt (fig/m0» a—CD3 + a—CD28___> Untreated THMC SMAD3+/+ SMADB'l' Page - 148 in untreated Smad3"‘ splenocytes and Smad3”’thymocytes, (Figure 37b and 38b, respectively). The magnitude of (x-CD3 + 0t-CD28-induced steady state IL-2 mRNA expression was greater in splenocytes and thymocytes from Smad3"' mice relative to Smad3”+ mice (Figure 37a and 38a, respectively). Collectively, these results suggest that autocrine TGF-[3l may regulate both the baseline and activation-induced IL-2 expression in a manner that is dependent on Smad 3 expression. B. Elevated IL-4 and IFN-y expression in Smad3-null splenocytes Elevated IL-2 expression in untreated Smad3"' splenocytes relative to Smad3”+ splenocytes (Figure 37 and Figure 38) suggest an activated in vivo T cell phenotype in Smad3"' mice. In light of this observation, the objective of this series of experiments was to investigate T cell differentiation in Smad3"‘ mice as determined by cytokine production. Splenocytes were isolated from Smad3"‘ mice, cultured in vitro in the presence or absence of plate-bound OL-CD3 + a-CD28 for 24 hours, and supematants were collected for ELISA analyses. IFN-y protein was quantified as a marker of THI differentiation and IL-4 protein was quantified to assess TH2 differentiation. The magnitude of IL-4 and especially IFN-y protein expression in untreated splenocytes was greater in Smad3”' mice than Smad3“ mice (Figure 39). Moreover, the magnitude of on- CD3 + (x-CD28-induced IFN-y and IL-4 protein was increased in Smad3"' relative to Smad”+ splenocytes. These results indicate that Smad3 may play a regulatory role in T cell differentiation and further establish that Smad3”’ splenic T cells can be induced to secrete IFN-‘y as well as IL-4 in vitro. Page — 149 Figure 39. Elevated basal IL-4 and IF N-y expression in Smad3-null splenocytes. Splenocytes were cultured in uncoated (Untreated) or a-CD3 + a-CD28-precoated tissue culture plates, as indicated, for 24 hours at 37°C. (a) IFN-y and (b) H.-4 were determined by ELISA. The data are expressed as the mean i SE (n =8). *,p < 0.05, Smad3”' significantly different from Smad3“. Page - 150 (a) 4001 300‘ lFN-y (U/mL) N O O 100* I SMAD3'/' V (b) 4 lL-4 (U/mL) Li SMA03+/+ * V/ Untreated ot-CD3 + ct-CDZB I SMAD3‘/' / SMAps+/+ Untreated oc-CD3 + oc-CD28 Page- 151 C. TGF-Bl-induced phosphorylation of Smad2 is not disrupted in Smad3 null splenocytes It has been demonstrated that IFN-y negatively regulates Smad signaling by increasing the activity of the inhibitor Smad, Smad7 (Ulloa et al., 1999). In light of the observed elevated IFN-y expression in Smad3"‘ splenocytes and thymocytes, the objective of these experiments was to verify the ‘functionality’ of Smad2 signaling in Smad3" T cells. Splenocytes and thymocytes were isolated from Smad3"' mice and treated in vitro with TGF-[3, (10‘3 ng/mL) for 60 minutes. Western blot analysis of whole cell lysates demonstrated TGF-Bl-induced Smad2 phosphorylation in Smad34' splenocytes as well as thymocytes and thus verified Smad2 responsiveness to TGF-B, (Figure 40a). Moreover, Smad3 protein expression was not detected in whole cell lysates from Smad3"‘ splenocytes or thymocytes (Figure 40b). VI. A role for Smad3 in the inhibition of lymphocyte proliferation by TGF-B, A. Growth inhibition by TGF-B, is attenuated in activated Smad3-null splenic T cells and thymocytes Having established that Smad3 is critical for regulating IL-2 expression, this series of studies was conducted to investigate a role for Smad3 in T cell growth inhibition by TGF-Bl. Splenocytes and thymocytes were isolated from Smad3”‘ mice and activated in vitro with a-CD3 + Ot-CD28. Following 72 hours in culture, [3H]-thymidine incorporation was quantified to assess T cell proliferation. TGF-[3l attenuated [3H]- thymidine incorporation in (X-CD3 + Ot-CD28-activated splenocytes and thymocytes from Smad3”+ mice in a concentration-dependent manner (Figure 413 and 42a, respectively). Page - 152 Figure 40. TGF-B,-induced phosphorylation of Smad2 is not disrupted in Smad3 null splenocytes. (a) Phosphorylated Smad2 expression in Smad3”' and Smad3“+ splenocytes and thymocytes. Whole cell lysates (50 ug) were separated on SDS-PAGE (10% polyacrylamide) under reducing conditions. Membranes were probed with a polyclonal rabbit Ab recognizing the phosphorylated form of Smad2. Following ECL detection, the membranes were washed, and reprobed with a monoclonal mouse antibody recognizing actin (as a loading control). (b) Smad3 expression in Smad3"‘ and Smad3”+ splenocytes and thymocytes. Whole cell lysates (25 pg) were separated on SDS-PAGE (10% polyacrylamide) under reducing conditions. Membranes were probed with a polyclonal rabbit Ab recognizing Smad3. A corresponding Ig horseradish peroxidase- linked secondary antibodies were used for protein detection using the ECL system Page - 0153 -Splenocytes Thymocytes Wild Smad3 Wild Smad3 * Type Null Type Null TGF'B1 - + - _ + - + - + Phospho» ”t ...... Smad2 actin > ~--~~~~fl Sptenocytes Thymocytes Wild Smad3 Wild Smad3 Type Null Type Null 3‘3“" > --wud‘— Page - 154 Growth inhibition by TGF-[3, was significantly diminished, but not abrogated in a-CD3 + a-CD28-activated splenic T cells (Figure 41a) and thymocytes (Figure 42a). In comparison to Smad3“ mice, [3H]-thymidine incorporation was elevated in untreated Smad3” splenocytes but not Smad3"‘ thymocytes (Figure 41 and 42, respectively). The magnitude of Ot-CD3 + a-CD28-induced [3H]-thymidine incorporation was also greater in Smad3”‘ splenocytes than Smad3“+ splenocytes (Figure 41). Collectively, these observations demonstrate that Smad3 is important for TGF-B,-induced growth arrest of (x-CD3 + a-CD28-activated T cells in vitro. These data also suggest that additional Smad3—independent mechanism(s) of growth inhibition may be activated following exposure to high concentrations of TGF-Bl. B. TGF-B, inhibits LPS-induced B cell proliferation in Smad3-null splenocytes A role for Smad3 in B cell growth inhibition by TGF-[3l was investigated. Splenocytes were isolated from Smad3"‘ mice and activated in culture with LPS for 48 hours prior to quantifying [3H]-thymidine incorporation. TGF-Bl attenuated [3H]- thymidine incorporation in LPS-stimulated Smad3“+ splenocytes in a concentration- dependent manner (Figure 43a). In contrast to T cells, inhibition of LPS-induced B cell growth by TGF-B, was not disrupted in Smad3“+ mice (Figure 43a). In a manner similar to that observed in T cells, the magnitude of LPS—induced [3H]-thymidine incorporation was elevated in Smad3"’ relative to Smad?” B cells (Figure 43b). In summary, these results suggest that TGF-B, inhibits LPS-induced B cell growth in vitro through a Smad3- Page - 155 Figure 41. Growth inhibition by TGF-B, is attenuated in activated Smad3-null splenic T cells. Splenocytes were activated with immobilized 0t-CD3 + a-CD28 in the presence (a) or absence (b) of TGF-[3,, as indicated for 72 hours at 37°C. Cell proliferation was quantified by incorporation of [3H]-thymidine. The data are expressed as the mean i SE of three separate experiments in quadruplicate. “a” denotes p < 0.05, Smad3“+ significantly different from Smad3"‘. “b” denotes p < 0.05, Smad?“ significantly different from vehicle. “c” denotes p < 0.05, Smad3" significantly different from vehicle. Page - 156 23>: [3H]-Thymidine Incorporation (b) [3H]-Thymidine Incorporation (cpm x 104) (cpm x 104) —O— SPLN SMAD3+/+ + SPLN SMAD3'/' 10 O l l I l l l l l I I I’l’ll o 10“ 10" 10'2 10'" 1.0 2.5 TGF-B1 (ng/mL) '— a—CDB + a—CDZB p [I SPLN SMAD3+/+ to I SPLN SMAD3/- 9 - $- 8 .- 7 _ 6 _ 5 .- 4 _ 3 .4 2 _ 1 _ Background 0 _ a None None Vehicle -0t-CD3 + a-CD28—> Page - 157 Figure 42. Growth inhibition by TGF-B, is attenuated in activated Smad3-null thymocytes. Thymocytes were activated with immobilized (I-CD3 + a-CD28 in the presence (a) or absence (b) of TGF-[3,, as indicated for 72 hours at 37°C. Cell proliferation was quantified by incorporation of [3H]-thymidine. The data are expressed as the mean i SE for three separate experiments in quadruplicate. “a” denotes p < 0.05, Smad3"‘ significantly different from Smad3“. “b” denotes p < 0.05, Smad3“ significantly different from vehicle. “c” denotes p < 0.05, Smad3"‘ significantly different from vehicle. Page - 158 I”: [3H]-Thymidine Incorporation A C. v [3H]-Thymidine Incorporation (cpm) x 104 (cpm) x 104 —‘O— THMC SMADB+/+ + THMC SMAD3'/' Olllllll (tr/I" o 104 10‘3 10'2 1.0 2.5 TGF-B1 (ng/mL) L— a-CD3 + a—CDZB » 8“ III THMC SMAD3+/+ 7- I THMC SMAD3-F T T 6- 1- 1- ‘ _|_ 5- 4_ 3- 2- 1- Background 0. None None Vehicle Page - 159 e a-CDB + a—CD28—> Figure 43. TGF-[3l inhibits LPS-induced B cell proliferation in Smad3-null splenocytes. Splenocytes were stimulated with LPS in the presence (a) or absence (b) of TGF-B], as indicated for 48 hours at 37°C. Cell proliferation was quantified by incorporation of [3H]-thymidine. The data are expressed as the mean i SE of three separate experiments in quadruplicate. “a” denotes p < 0.05, Smad3"' significantly different from Smad3”. “b” denotes p < 0.05, Smad3“+ significantly different from vehicle. “c” denotes p < 0.05, Smad3"' significantly different from vehicle. Page - 160 E U ( ) [3H] - Thymidine Incorporation 9H] - Thymidine Incorporation (cpm) x 104 (cpm) x 104 8' T —O— SMA03+I+ _ r _,_ Q + SMA03-/- 6- T I T C T . T L . "' b J" c - . . ‘L 0 T C 4- i E . C - . o b,C o o 2‘ \ b,c ‘ o 0llllllllrll""l o 104 10'3 10‘2 10‘1 1.0 2.5 L TGF-B1 (ng/mL) LPS (10 ug/mL) > 8 ‘ Z] SMA03+/+ _ I SMAos-l- 6 — 4 _ 2 _ 0 -4] A Back round No 9 treatment Vehicle L LPS (10 ug/mL) —-> Page—161 independent mechanism of action and suggests a cell-type specificity for Smad3- mediated regulation of lymphoid proliferation. C. Exogenous IL-2 reverses TGF-Bl-induced inhibition of T cell growth IL-2 is essential for OL-CD3 + (x-CD28-induced splenic T cell proliferation in vitro. On-the-other-hand, in vitro LPS-stimulated splenic B cell proliferation is not dependent upon IL-2. In light of these observations, a series of experiments were undertaken to determine whether inhibition of OL-CD3 + a-CD28-induced T cell growth by TGF-B, was due to a direct inhibition of IL-2. [3H]-thymidine incorporation was quantified in a—CD3 + (x-CD28-activated splenic T cells in the presence of exogenous IL-2. Exogenous recombinant IL-2 (50 U/mL) reversed the growth inhibitory effect of TGF-Bl in activated SMAD3“+ T cells (Figure 44a). These results indicate that IL-2 can overcome TGF-Bl-induced inhibition of T cell growth and support a hypothesis that inhibition of (at-CD3 + Ot-CD28-induced T cell growth by TGF-[3l in vitro is mediated through a direct effect of TGF-B| on IL-2 production. Thus, it is tempting to speculate that attenuated TGF-B,-induced inhibition of SMAD3"‘ T cell growth by TGF-B, to due to a direct effect of Smad3 on IL-2 expression. Moreover, augmentation of proliferation by exogenous IL-2 was demonstrated in activated SMAD3“+ splenocytes (Figure 44a), but not SMAD3"’ splenocytes (Figure 44b). These results are consistent with elevated IL-2 expression in a-CD3 plus a-CD28-activated SMAD3"‘splenocytes relative to SMAD3“ splenocytes, as discussed previously. Page - 162 Figure 44. Reversal of TGF-B,-induced inhibition of peripheral T cell proliferation by exogenous IL-2. Splenocytes obtained from (a) wild type or (b) Smad3-null splenocytes were activated with immobilized 0t-CD3 + Ot-CD28 with or without 50 U/mL IL-2 in the presence or absence of TGF-Bl, as indicated for 72 hours at 37°C. Cell proliferation was quantified by incorporation of [3H]-thymidine. The data are expressed as the mean 1 SE of two separate experiments in quadruplicate. *,p < 0.05, Smad3"‘ significantly different from vehicle. Page — 163 27.. I|._"_ 5 S 0...... u .. g .0 ant ...-...... 1G+ rm. f... a :1: I la. Smad3+ / + Treatment No 10.0“ vow x 0003 00000000005 00_0_E>c._.- fa (a) - Exogenous lL-2 1 I + Exogenous IL 2 T T 2.5 5 TGF-B1 (ng/mL) -CD3 + a-CD28——> Treatment No 10‘0" Smad3'/' 75 5.- 25 vow x 0008 m 002059605 0530?:- - HIE ( Page - 164 VII. Modulation of protein binding to the IL-2 promoter by TGF-[3l A. CAGA elements in the regulatory region of the mouse IL-2 promoter In further investigating the mechanism(s) by which Smad3 regulates IL-2 gene transcription in response to TGF-Bl, five CAGA sequences were identified in the minimal essential regulatory region of the mouse IL-2 promoter. The four bp CAGA sequence functions as a Smad response element in the promoter of numerous TGF-Bl-responsive genes (Chen et al. 1999; Dennler et al. 1998; Jonk et al. 1998; Wu et al. 1997; Zhang et al. 2000; Zhang and Derynck 2000). The five M sequences identified in the IL-2 promoter are located -102, -1 17, -145, -156, and -268 bp upstream (5’) of the transcriptional start site, and lie adjacent to or overlap the distal -285 distal NFAT site, the —168 CD28RE, the -153 proximal AP-l site, or the —1 13 NRE-A. The close proximity of the QAfiA sequence with transcription factor binding sites in the IL-2 promoter is consistent with the cooperative interaction of Smad proteins with other transcription factors (Brodin et al. 2000; Chen et al. 1999; Kon et al. 1999; Liberati et al. 1999; Qing et al. 2000; Wong et al. 1999). To investigate the importance of CAGA sequences for DNA binding activity in the IL-2 promoter, a pair of oligonucleotides were constructed for each of the four aforementioned DNA binding elements (i.e., the -285 NFAT site, the -168 CD28RE, the -153 proximal AP-l site, and the -113 NRE-A) with each pair consisting of a wild type oligonucleotide containing the naive CAQA sequence and a mutant oligonucleotide containing a mutated CAGA sequence. The sequences for each of these eight oligonucleotides are provided in Table 2. Splenocytes obtained from naive B6C3F1 mice, activated in culture with a-CD3 + 0t-CD28 for two hours, and nuclear lysates were Page - 165 obtained for EMSA analyses. Mutation of the CAGA sequence markedly shifted DNA binding complexes at the proximal AP-l site and the CD28RE (Figure 45; compare lanes 1 and 3, lanes 2 and 4, lanes 5 and 7, and lanes 6 and 8). A similar shift was not observed in DNA binding complexes at the distal NFAT site or NRE-A (Figure 45; compare lanes 9 and 11, lanes 10 and 12, and lanes 13 and 14). B. Presence of Smad3 in the transcription factor complex that binds to the proximal AP-l site in the mouse IL-2 promoter EMSA supershift analyses were performed to investigate whether Smad3 protein was a component of the transcription factor complex binding to the proximal AP—l site. Splenocytes obtained from naive B6C3F1 mice were activated in culture with a-CD3 + a-CD28 for 60 minutes in the presence or absence of 10 ng/mL TGF-[3,. Nuclear lysates were obtained for analyses of DNA binding activity. For supershift studies, nuclear lysates were incubated with an antibody specific for Smad3. Modest Smad3 binding was present only with nuclear extracts from TGF-Bl-treated cells as demonstrated by inhibition of DNA binding (Figure 46, compare lanes 3 and 5) and supershift (Figure 46, compare lanes 8 and 10). Importantly, intact M sequences were demonstrated to be non-essential for TGF-Bl-induced Smad3 binding (Figure 46, lane 10) and suggest that Smad3 may bind to the proximal AP-l site as a component of a protein complex, for example a Smad3-AP-1 heteromer. C. Smad3 is not essential for TGF-Bl-induced binding to the proximal AP-l proximal site in the mouse IL-2 promoter Page - 166 Figure 45. A functional role for CAGA sequences in DNA binding activity in the mouse IL-2 promoter. Splenocytes were either untreated or stimulated with a-CD3 + a-CD28 for 2 hours at 37°C. Nuclear proteins were isolated and incubated (5 ug/lane) with a 32P-labeled probe (proximal AP-l site, CD28RE, distal NFAT site, or NRE-A; as indicated) containing native CAGA (W) or mutated QACLA (M) sequences as indicated. Protein binding complexes were resolved on a 4% polyacrylamide gel. Page - 167 .1. mFNFZOFm w n 0 m wm N Foam.— - - + - + - + - + - + - + - w~008+808 2 >> 2 2 >> >> 2 2 >> >> 2 _2 >> >> mack. o J J J A A A b . \WWW \VVI 9)me ¢vaw¢mWw 0%.! A0600 Amy/co “M60 bee/co 1.0 1.0 [.7 1.7 node/0% .a ow. owes-w (by (by «.5 (by 9» Page — 168 Figure 46. Smad3 is a component of the transcription factor complex that binds to the proximal AP-l site in the mouse IL-2 promoter. Splenic T cells were isolated from B6C3F1 mice and treated as indicated. Nuclear proteins were isolated and incubated (5 jig/lane) with a 32P-labeled probe on a 4% polyacrylamide gel. The oligonucleotides used were synthesized to correspond to the native proximal AP—l site (W) or to a mutated AP-l oligonucleotide containing disrupted M Smad binding elements (M) was compared. Where indicated the DNA-binding protein complex was supershifted with anti-Smad3 (Zymed) or competed with excess unlabeled (1 pmol) probe. Page - 169 mooemv 5 0.. 0.. 0.. o; 3 3 0.. ..N S 10:280. 9:23. ,— ‘— O ‘— N s 23 3 3 coo... + - 080-3800 - - 3.50.: 0:00-000 + - - - mumEmso +++2 ++2w n+2“ '2 +++gtn NF >> + + «J .OI D. M wm0++2 Page — 170 A role for Smad3 in TGF-Brinduced binding to the proximal AP-l site was further investigated using nuclear lysates from a purified preparation of T cells. Thy1.2+ purified Smad3"’ and Smad3“’* splenic T cells were activated in culture with a-CD3 + 0t- CD28 for 60 minutes in the presence or absence of 10 ng/mL TGF-[3,. Nuclear lysates were obtained to determine the influence of Smad3 on TGF-B,-induced DNA binding activity at the proximal AP-l site. As illustrated in Figure 47, TGF-Bl-induced DNA binding activity was demonstrated in nuclear lysates from activated Smad3“+ as well as Smad3"‘ splenic T cells (Figure 47, compare lanes 3 and 4 and compare lanes 6 and 7). Moreover, basal DNA binding to the proximal AP-l site is markedly reduced in nuclear lysates from Smad3”+ mice relative to Smad3"' mice (Figure 47, compare lanes 2 and 5). These results are in consistent with differential basal expression of IL-2 mRNA and protein in untreated Smad3”+ and Smad3"' splenocytes and support a role for Smad3 in regulating IL-2 gene transcription. D. CAGA sequences are not essential for TGF-Bl-induced binding to the proximal AP-l proximal site in the mouse IL-2 promoter A role for _C_A_G_A sequences in TGF-B,-induced binding to the proximal AP-l site was also investigated using nuclear lysates from purified T cells. Thy1.2+ purified Smad3"‘ and Smad3“+ splenic T cells were activated in culture with (at-CD3 + a-CD28 for 60 minutes in the presence or absence of 10 ng/mL TGF-[3,. Nuclear lysates were obtained to determine the influence of Smad3 on TGF-B,-induced DNA binding activity to proximal AP-l site containing mutated M sequences. As illustrated in Figure 48, TGF-Bl-induced DNA binding activity was evident in nuclear lysates from activated Page - 171 Figure 47. Smad3 is not essential for TGF-B,-induced binding to the proximal AP-l site in the mouse IL-2 promoter. Thyl.2+ splenic T cells were isolated from Smad3”+ or Smad3"’ mice and activated in vitro with plate-bound Ot-CD3 + a-CD28 for 60 minutes in the presence or absence of 10 ng/mL TGF-[3,. Nuclear proteins were isolated and incubated (5 ug/lane) with a 32P-labeled probe corresponding to the naive proximal AP-l site. DNA binding complexes were resolved on a 4% polyacrylamide gel. Page - 172 Smad3+l+ ‘Smad3'/' ‘5 TGF-[31(10ng/mL) _ - ... - - “g 0t-CD3/0t-CD28 - + + - + £53 + . + AP-1> Free Probe * 1234567 Page - 173 Figure 48. QAQA sequences are not essential for TGF-B,-induced binding to the proximal AP-l site in the mouse IL-2 promoter. Thyl.2+ splenic T cells were isolated from Smad3“+ or Smad3”' mice and activated in vitro with plate-bound a-CD3 + a—CD28 for 60 minutes in the presence or absence of 10 ng/mL TGF-Bl. Nuclear proteins were isolated and incubated (5 [lg/lane) with a 32P-labeled probe corresponding to the M'— mutated proximal AP-l site. DNA binding complexes were resolved on a 4% polyacrylamide gel. Page - 174. Smad3'/' Smad3+l+ TGF-[31(1Ong/mL)_ _ + a-CD3+a-CD28— + + _ 1 21. 11 , 15.11 AP-1 —> Relative 3- .. Intensity 2.4 1.1 2.0 8.9 6.0 7.8 Page - 175 Smad3“+ and Smad3"' splenic T cells (Figure 48, compare lanes 2 and 3 and compare lanes 5 and 6). In addition, basal DNA binding is markedly suppressed in the absence of Smad3 (Figure 48, compare lanes 1 and 4). These results are consistent with the differential basal DNA binding activity by untreated Smad3”+ and Smad3"' nuclear lysates at the proximal AP-l site containing the naive QAQA sequences. Collectively, these results suggest that while neither Smad3 nor the CAGA sequences are essential for TGF-Bl-induced DNA binding activity at the proximal AP-l site, each of these factors plays a modulatory role in TGF-Bl-induced DNA binding activity. Moreover, these results also suggest that Smad3 and CAGA sequences play a role in DNA binding activity in naive resting cells. E. Temporal response of TGF-Bl-inducible DNA binding to the proximal AP-l site in the mouse IL-2 promoter. The temporal response of TGF-[3l on AP-l DNA binding activity was investigated. Splenocytes from B6C3F1 mice were activated in vitro in the presence or absence of 10 ng/mL TGF-B, for 30, 60, 120, and 240 minutes. As illustrated in Figure 49, TGF-B, (10 ng/mL) effectively impaired baseline AP-l DNA binding activity (Figure 49, compare lanes 2 and 12). However, TGF-[3l only modestly attenuated a-CD3 + 0t- CD28-induced binding to the proximal AP-l site (Figure 49). Importantly, these experiments do not address the composition of the DNA binding complex. The temporal response of TGF-[3I on DNA binding activity to the mutant proximal AP-l probe was also investigated. TGF-[3, (10 ng/mL) attenuated basal DNA binding (Figure 50, compare lane 3 and 12) and only modestly attenuated (x-CD3 + a-CD28-induced proximal AP-l binding activity at 60, 120, and 240 minutes. It has recently been Page - 176 Figure 49. Temporal response of TGF-Bl-induced protein binding to the proximal AP-l site of the IL-2 promoter. Splenocytes were isolated from B6C3F1 mice and activated in vitro with (x-CD3 + a-CD28 in the presence or absence of 10 ng/mL TGF-B, for O, 30, 60, 120, or 240 minutes. Nuclear proteins were isolated and incubated (5 ug/lane) with a 32P-labeled probe corresponding to the native proximal AP-1 site. DNA binding complexes were resolved on a 4% polyacrylamide gel. Where indicated, cold competitor studies were conducted with 1 pmol unlabeled probe. Page - 177 de m; P; m.N o.v m.m m.N 5N TN fim m.N m.m ‘Bacmus 1 . , 2.52% A P - n_< 9 NF F F or o w n o oo o OVN ovN ONF GNP co ow or or c or m N F 25.. 8 cm on 8 o EEENSSLBR. o 2 o 2 o - - :EREEiE um mwtw l mu 9.. w» ‘1 ta. w KB .1. a 1. l .m M R N .U if . d '1 ..hu l. O 0.. ..MW .m :1 l m m... .. .0 Page - 178 Figure 50. Temporal response of TGF-BI-induced protein binding to a CAGA- mutated proximal AP-l site. Splenocytes were isolated from B6C3F1 mice and activated in vitro with OL-CD3 + oc-CD28 in the presence or absence of 10 ng/mL TGF-Bl for 0, 30, 60, 120, or 240 minutes. Nuclear proteins were isolated and incubated (5 ug/lane) with a 32P-labeled probe corresponding to a M-mutated proximal AP-l site. DNA binding complexes were resolved on a 4% polyacrylamide gel. Where indicated, cold competitor studies were conducted with 1 pmol unlabeled probe. Page - 179 5N of WV m.m 5v m.m of 9m TN Tm F.m. A>tmc35o>5£mm A mack. o2... AF-n_< ‘. .. x a t . . .... , .1. .. .. . 1.. i s .1. .\ .. m F N F F F OF 0 w h o m v m N F man.— 00 o OVN OVN ONF ONF oo 00 OM Om CM 0 Arr—Ev wNoU-d + m008 fit. (.4 2 22 o 2 o 2 o 2 o . - 35355.qu Page - 180 demonstrated that AP-l-AP-l and AP-l-Smad complexes migrate as a single band and are undistinguishable on a 4% PAGE (Verrecchia et al., 2001b). In light of this observation, one interpretation of the present study is that TGF-B, modulates IL-2 gene transcription by altering the composition of the proteins in the transcription factor complex binding to the AP-l site. An alternative interpretation of these data is that the proximal AP-l site in the IL-2 promoter is not a TGF-[irresponsive element. VIII. Effect of TGF-B, on NFAT and NF-KB transcriptional activity in a-CD3 + 0t- CD28-activated splenic T cells Smad proteins cooperatively interact with other transcription factors and accessory proteins to modulate TGF-Bl-meditated transcription via multiple different mechanisms (Massague and Wotton 2000; Pick et al. 1999). Several of these mechanisms do not involve direct modulation of DNA binding by Smad3. For example, Smad3 cooperatively interacts with p300 through a proteinOprotein interaction to effectively enhance transcriptional activity through the remodeling of chromatin structure (Feng et al. 1998; Itoh et al. 2000; Shen et al. 1998). In light of the numerous protein binding—independent mechanisms of Smad- mediated transcription, the putative mechanisms whereby TGF-[3l regulates IL-2 gene transcription were further investigated using reporter gene assays. Mouse EL—4 and human Jurkat T cell lymphoma cells were transiently transfected with pNFAT or pNF-KB reporter plamsids. These reporter plasmids were selected on the basis that each is regulated by a promoter containing multiple copies of CAGA sequences upstream of the transcriptional start site. The transfected cells were activated with a-CD3 + a-CD28 in Page-181 the presence of TGF-B, (10‘3 10", and 10 ng/mL) for 24 hours. The pSV-B-gal control vector was co-transfected in each experiment to monitor transfection efficiency. SEAP activity and IL-2 protein were measured from supematants and B-gal activity was measured from cell lysates. Anti-CD3 + a-CD28-induced p-NFAT-SEAP activity in EL- 4 and Jurkat cells was attenuated by TGF-B, in a concentration-dependent manner (Figure 51a and Figure 53a respectively). In contrast, TGF-[3l stimulated OL-CD3 + Ot- CD28-induced p-NF-KB—SEAP activity in EL-4 as well as Jurkat cells in a concentration- dependent manner (Figure 523 and Figure 53b, respectively). Interestingly, TGF-[3l attenuated a-CD3 + a-CD28- induced IL-2 protein secretion in p-NF-KB-SEAP as well as p-NFAT-SEAP-transfected EL-4 cells (Figure 51b and Figure 52b, respectively). IX. Involvement of MAPK in the regulation of Smad/TGF-Bl signaling in T cells It is well established that MAPK signaling is activated upon TcR ligation (Whitehurst and Geppert 1996). TGF—[3l also activates MAPK signaling cascades through the TBR (Choi 2000; Hartsough and Mulder 1997; Hu et al. 1999; Mulder 2000; Visser and Themmen 1998). Furthermore regulatory cross talk between Smad and MAPK signaling has been established in several epithelial and fibroblast cell lines (Kretzschmar et al. 1999; Mulder 2000; Stroschein et al. 1999b; Watanabe et al. 2001; Yue and Mulder 2000). In light of these observations, the objective of this series of experiments was to investigate a putative role for MAPK signaling in regulating TGF-B,- induced Smad signaling in T cells. Page - 182 Figure 51. Effect of TGF-[3l on NF AT transcriptional activity in (l-CD3 + a- CD28-activated T cells. ISL-4 cells (2 X 105c/mL) were transiently transfected using Cytofectene transfection reagent with 8 ug pNFAT-SEAP reporter construct and pGL3 B- galactosidase reporter plasmid (1 1.1g) in RPMI for 1 hour at 37°C. The cells were then transferred to un-coated or a-CD3 + a-CD28-precoated 96-well plates and either left untreated (NA) or treated with TGF-B, (0.001 — 10 ng/mL) or VH (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture) for 24 hours at 37°C. (a) SEAP activity was normalized for transfection efficiency using B-gal activity from cell lysates. (b) Aliquots of the supematants were removed for IL-2 protein secretion (ELISA). Data are reported as the mean 1 SE of quadruplicate cultures. *,p < 0.05 as compared to the matched vehicle control. Results are representative of two experiments. Page - 183 a-CDS +0- 165191 33:; and 96155» 1]: were 3' nd emu fci PBS. p113}. .'. 1315B.” c511 Ewifi .1511. Di pared to LT (a) (b) Normalized pNFAT-SEAP Activity (RLU) lL-2 (Units/mL) 1000- 60- 50- 401 30- 20- 0 NA 0 10-310-1 10 TGF-B1 (ng/mL) ND ND ND ND ND NA 6 16-316-1 1b TGF-B1 (ng/mL) Page - 184 o 10'3 10-1 1o TGF-[31 (ng/mL) a-CD3 + a-CDZ8" 0 10-310-1 10 TGF-B1 (ng/mL) OL-CD3 + tut-C028—> Figure 52. Effect of TGF-B, on NF -KB transcriptional activity in (It-CD3 + 0t- CD28-activated T cells. EL-4 cells (2 X 10%/m1.) were transiently transfected using Cytofectene transfection reagent with 8 ug pNF-KB-SEAP reporter construct and pGL3 B-galactosidase reporter plasmid (1 1.1g) in RPMI for 1 hour at 37°C. The cells were then transferred to un-coated or oc-CD3 + a-CD28-precoated 96-well plates and either left untreated (NA) or treated with TGF-B, (0.001 — 10 ng/mL) or VH (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture) for 36 hour at 37°C. (a) SEAP activity was normalized for transfection efficiency using B-gal activity from cell lysates. (b) Aliquots of the supematants were removed for IL-2 protein secretion (ELISA). Data are reported as the mean i SE of quadruplicate cultures. *,p < 0.05 as compared to the matched vehicle control. Results are representative of a two separate experiments. Page - 185 (a) 10001 5 >1: é * I a l jg 750d 4.) U < D. < 3’“ l lfl-CDltfl' é 500. I . . 2‘ I Name; L2;- ructandJGL‘ ‘ '0 250- '6115 were in: g andeitheriei § 135.1135. c23 1 c.1511 NA 0 10-310-1 10 0 10-310-1 10 “can“. TGF-[31 (ng/mL) TGF-B1 (ng/mL) ELISA}. Da: a-CD3 + a-CDZB—> 111131611101? (b) 50" 1161115. T 40- 3 E E 30' 'E 3 N 2 20' 10 ND ND ND ND ND NA 0 10-310-1 10 0 10-310-1 10 TGF-[31 (ng/mL) TGF-B1 (ng/mL) a-CD3 + on-CDZ8"> Page - 186 Figure 53. Effect of TGF-B, on NFAT and NF-KB transcriptional activity in on- CD3 + a-CDZS-activated Jurkat cells. Jurkat cells (5 X 106 c/rnL) were transiently transfected using Cytofectene transfection reagent with (a) pNFAT-SEAP or (b) pNF- KB-SEAP reporter constructs and co-transfected with a pGL3 B—galactosidase reporter plasmid in RPMI for 1 hour at 37°C. The cells were then transferred to un-coated or a- CD3 + OL-CD28—precoated 96-well plates and either left untreated (NA) or treated with TGF-[3, (0.001 — 10 ng/mL) or VH (0.02% PBS, pH 3.5, containing 0.1% BSA, final concentration in culture) for 24 hours at 37°C. SEAP activity was normalized for transfection efficiency using B-gal activity from cell lysates. Data are reported as the mean 1 SE of quadruplicate cultures. *,p < 0.05 as compared to the matched vehicle control. Results are representative of two separate experiments. Page - 187 (a) S 600- _1 ‘5 > i; 500‘ .5 u < a 400- E) V.’ 1. ml utility in 6.5 300 Z u are transfer; _3' ZAP or [1911- E 20° 76 tondase more: E 100- ‘ ‘r un-coatcéora- ‘23 T T T T 1 or treated 17: 0 116135.11: NA 0 10-310-1 10 0 10-310-1 10 . .. TGF-[l1 (ng/mL) TGF-B1 (ng/mL) ”mum a-CD3 + a-CDZ8-> repent”? inched 13353 (b) S 600' _l 5 5. 5001 '5 '5 * U 'r < 400- o. E) ch ch 300‘ 2‘ LL Z 0- 200' .0 Q) .5 To 100 E L _ 0 Z 0.1 NA 0 10-310-1 10 0 10-310-1 10 TGF-B1 (ng/mL) TGF-[31 (ng/mL) a-CD3 + a-CD28-> Page - 188 A. Concentration-dependent effect of TGF-B, on ERK MAPK activity in mouse splenocytes ERK MAPK has been previously shown to negatively regulate Smad3 nuclear translocation by phosphorylation of serine residues within the linker region of the Smad3 protein (Kretzschmar et al. 1997; Kretzschmar et al. 1999). TGF-[3l reportedly activates ERKl MAPK and ERK2 MAPK in numerous cell lines (Axmann et al. 1998; Reimann et al. 1997). The effects of TGF-B, on ERK MAPK activity in myeloid and/or lymphoid cells has not yet been studied. The objective of the present experiments was to investigate ERKl MAPK and ERK2 MAPK activity in response to TGF-[3l in mouse splenocytes. Towards this end, splenocytes were isolated from naive B6C3F1 mice, treated with TGF-B, for 15 minutes, and nuclear lysates were isolated for Western blot analyses. Expression of phosphorylated MAPK is indicative of kinase activity (Cobb and Goldsmith 1995), therefore an antibody specifically recognizing phosphorylated ERKI MAPK and ERK2 MAPK was utilized. A concentration-dependent augmentation of ERKI MAPK and ERK2 MAPK phosphorylation by TGF-[3l was demonstrated (Figure 54). B. Effect of U0126 and PD98059 on TGF-Bl-induced inhibition of IL-2 protein secretion in activated splenic T cells Having established that TGF-[3l increases splenic ERK MAPK activity, a role for ERK MAPK in regulating TGF-Bl-induced inhibition of IL-2 secretion was investigated in Ot-CD3 + a-CDZS-activated splenic T cells. Initial experiments investigated the effects of U0126, a MEK1/2-specfic inhibitor, on TGF-B,-mediated Smad3 nuclear Page - 189 Figure 54. Concentration-dependent effect of TGF-B, on ERK MAPK activity in mouse splenocytes. Naive splenocytes were treated with TGF-B, for 15 minutes. Nuclear proteins (25 pg) were isolated and resolved on a 10% denaturing polyacrylamide gel. Proteins were transferred to nitrocellulose and incubated with antibodies for the phosphorylated ERK and ERK2. Nuclear lysates for splenocytes treated with PMA/Io (80 nM/ 1 11M) for 15 minutes was incorporated as a positive control for ERK activity. Page - 190 111111111! :3 .11 11.1 if . A gm-.. . A $5... 222.2 F 1:1: 1: 1: o <2 :53 5.9 Page - 191 expression. Splenocytes isolated from naive B6C3F1 mice were pretreated with U0126 (1.25, 2.5, 5, 10, or 20 1.1M) for 30 minutes followed by stimulation with 10 ng/mL TGF— B, for 60 minutes. As illustrated, U0126 effectively augmented TGF—Bl-induced Smad3 nuclear translocation (Figure 55). Next, splenocytes isolated from naive B6C3F1 mice were pretreated with either PD98059 (25 11M) or U0126 (10 11M) for 30 minutes prior to a—CD3 + (x-CD28-induced activation in the presence or absence of TGF-[3,. Following a 24 hour incubation, supernatants were collected for ELISA analyses. As illustrated in Figure 56, PD98059 and U0126 attenuated TGF-Bl-induced inhibition of IL-2 secretion. These results suggest that ERK MAPK signaling plays a role in Smad3-dependent TGF- B,-mediated inhibition of IL-2 protein production. C. Concentration-dependent effects of TGF-B, on JNK MAPK activity in mouse splenocytes JNK MAPK also plays a regulatory role in T cell activation, and is of particular importance in CD28-associated signaling (Kempiak et al. 1999). An inter-dependent relationship between the JNK and SMAD signaling by TGF-[31 has been established (Engel et al. 1999). Activation of JNK MAPK by TGF-[5, in T cells has not yet been investigated. The objective of these studies was to determine whether TGF-[3l enhances JNK MAPK activity in splenocytes. Splenocytes were isolated from naive B6C3F1 mice, treated with TGF—B, for 15 minutes, and nuclear lysates were isolated for Western blot analyses. In contrast to ERK MAPK, TGF-B1 (10‘s, 10“, 10", 1, and 10 ng/mL) did not induce phosphorylation of either JNK] or JNK2 MAPK at any concentration tested (Figure 57). Page - 192 Figure 55. The effect of U0126 on TGF-Bl-induced Smad3 activation Naive or a- CD3 + a-CD28-activated splenocytes were pretreated with U0126 (10 1.1M) for 30 minutes and then with TGF-B, for 15 minutes. Nuclear proteins (25 pg) were isolated and resolved on a 10% denaturing polyacrylamide gel. Proteins were transferred to nitrocellulose and incubated with an antibody specific for Smad3. The fold induction is reported as intensity relative to naive. Page - 193 35:35 w>.5£om 1' $251 oF.N Fo.F mNF woF m.N mo.F mN.F SF SF 0 I 23 £5: 35...... o: 5-qu Page -194 Figure 56. Effect of U0126 and PD98059 on TGF-BI-induced inhibition of IL-2 protein secretion in activated splenic T cells. Spleens from B6C3F1 mice were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 5 X 106 c/mL. Naive splenocytes were pretreated with U0126 (10 1.1M) or PD98059 (25 11M) for 30 minutes followed by a-CD3 + a-CD28—induced activation in the presence or absence of TGF-[3l for 24 hours. Supernatants were collected and analyzed for secreted IL-2 protein via ELISA as described in “Materials and Methods”. Data are expressed as the mean i of quadruplicate samples, and are representative of two separate experiments. *p < 0.05 as determined by Dunnett’s t test as compared to the relevant vehicle control. Page - 195 v ppppp w; . lL-2 Protein Secretion (U/mL) 200‘ $ é U1 0 1 I Ot-CD3 + a-CDZB only a-CD3 + a-CD28 + PD98059 Pretreatment [I] a-CDB + a—CD28 + U0126 Pretreatment * 1" WW3“ VH -5 -4 -3 -2 -1 0 Log TGF-B1 (ng/mL) a-CD3 + a-CD28 p Page - 196 Figure 57. The effect of TGF-B, on JNK] and JNK2 activation. Naive splenocytes were treated with TGF-[3l (10 ng/mL) for 15 minutes. Nuclear proteins (25 pg) were isolated and resolved on a 10% denaturing polyacrylamide gel. Proteins were transferred to nitrocellulose and incubated with a polyclonal rabbit anti-phospho antibody for JNK1/2 as described in “Materials and Methods”. Data are representative of two separate experiments E .3 E E E E E 2.3 . Vma 2 F 1:1: 99 o 0...... 35355.? coo/24.0% erxn/MQ/o Page - 198 X. A role for Smad3 in immunoglobulin production by TGF-B, in vitro Smad3"' mice are viable, but succumb at an early age to a deficiency in mucosal immunity (Yang et al. 1999). Serum IgA levels and the number of intestinal IgA” B cells are normal in Smad3-null mice (Datto er al. 1999; Yang et al. 1999). Having established that inhibition of LPS-induced B cell growth by TGF-l3, is unaffected in Smad3"' B cells, the present studies investigated whether Smad3 is essential for modulation of humoral immune responses by TGF-[3, in vitro. A. Inhibition of the in vitro T cell-dependent sRBC IgM AFC response by TGF-[3l is augmented in Smad3-null splenic B cells Class 11 major histocompatability complex (MHC) molecules function to present processed antigens to T helper cells and are thus essential for T cell-dependent humoral immune responses. It is well established that TGF-B, attenuates humoral immune responses, in part, through down—regulation of major MHC molecules (Kobayashi et al. 1999; Letterio et al. 1996; Nakabayashi et al. 1997). Smad3 is essential for down- regulation of MHC H expression by TGF-B, in astrocytes (Dong et al. 2001). In light of these observations, a role for Smad3 in mediating TGF-Bl-induced inhibition of the T cell-dependent sRBC IgM AFC response was investigated. Splenocytes were isolated from Smad3"' mice and sRBC-sensitized for 5 days in culture. The hemolytic plaque assay was used to quantify antibody producing cells. TGF-Bl attenuated the AFC response in Smad3“+ B cells in a concentration-dependent manner (Figure 58). In contrast, Smad3"‘ B cells were less sensitive to the inhibitory effects of TGF-B, as illustrated by a significant (p < 0.05) inhibition of antibody producing cells by TGF-B, at the 1 ng/mL concentration only. The total number of viable splenocytes Page - 199 Figure 58. Inhibition of the T cell-dependent sRBC AFC response by TGF-B, is unaffected in Smad3-null splenic B cells in vitro. Spleens from Smad3"' and Smad“ were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 1 X 107 c/mL and transferred in SOO-uL to the wells of a 48-well culture plate. Quadruplicate cultures were left untreated (no treatment) or sensitized with sRBC in the presence or absence of TGF-[3l (0.001, 0.01, 0.1, and 1 ng/mL final concentration in culture) for 5 days. Alternatively cells were treated with the vehicle (VH; 0.02% PBS final concentration in culture, pH 3.5, containing 0.1% BSA) for 3 days. Cultures were subsequently analyzed for a day 5 antibody response by enumerating the number of AFC; spleen cell viability and total recovered cells/culture were also determined. The results from quadruplicate determinations are expressed as the mean AFC per 106 recovered viable cells i SEM (n=4). *p < 0.05 as determined by Dunnett’s t test as compared to the vehicle control. “#” denotes p < 0.05, Smad3”' compared to same group Smad3“ as determined by Dunnett’s t test. The results are representative of two separate experiments. Page - 200 an A... 09 x mzmu min; US$88. 0.001 0.01 0.1 TGF-B1 (ng/ml) 1000‘ Page - 201 recovered after 5 days in culture were not markedly altered with targeted deletion of Smad3 (Figure 58 insert). Noteworthy, the magnitudes of basal and sRBC-induced primary AFC responses were greater in Smad3"‘ splenocytes than Smad3“+ splenocytes (Figure 58). B. Inhibition of the in vitro polyclonal antibody response by TGF-[i1 is augmented in Smad3 null splenic B cells In further defining a role for Smad3 in regulating humoral immunity, the polyclonal response to LPS was also investigated. Splenocytes were isolated from Smad3"’ mice and sensitized with LPS for 3 days in culture. Similar to the T cell- dependent sRBC AFC response, a decreased sensitivity to TGF-Bl-induced inhibition of the polyclonal response to LPS was demonstrated in Smad3"' B cells relative to Smad3“+ B cells (Figure 59). The total number of viable splenocytes recovered after 3 days in culture was unaffected with targeted deletion of Smad3 (Figure 59 insert) demonstrating that the AFC response was not indirectly confounded by cytotoxic effects of TGF-B1 treatment. Notably, the magnitude of LPS-induced IgM AFC response was elevated in Smad3"' splenocytes relative to Smad3”+ splenocytes (Figure 59). These results were somewhat surprising as normal numbers of splenic and lymph nodal IgM B cells in SMAD3"’ mice have been reported (Yang et al. 1999). C. Inhibition of LPS-induced IgM production in Smad3-null B cells by TGF-Bl in vitro. The results of the LPS IgM AFC response were corroborated by measuring IgM production in vitro. Splenocytes were isolated from Smad3"' mice, stimulated with LPS Page - 202 Figure 59. Inhibition of the T cell-independent LPS AFC response by TGF-B, is unaffected in Smad3-null splenic B cells in vitro. Spleens from Smad34' and SMAD3” were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 5 X 106 c/mL and transferred in SOO-uL to the wells of a 48-well culture plate. Quadruplicate cultures were left untreated (no treatment) or activated with LPS (10 ug/mL) in the presence or absence of TGF-Bl (0.001, 0.01, 0.1, and 1 ng/mL final concentration in culture) for 3 days. Alternatively cells were treated with the vehicle (VH; 0.02% PBS final concentration in culture, pH 3.5, containing 0.1% BSA) for 3 days. Cultures were subsequently analyzed for a day 3 IgM antibody response by enumerating the number of AFC; spleen cell viability and total recovered cells/culture were also determined. The results from quadruplicate determinations are expressed as the mean AFC per 106 recovered viable cells i SEM (n=4). “a” denotes p < 0.05 as determined by Dunnett’s t test and compared to the Smad?” vehicle control. “b” denotes p < 0.05 as determined by Dunnett’s t test and compared to the Smad3"' vehicle control. “#” denotes p < 0.05, Smad3”‘ compared to same group Smad?” as determined by Dunnett’s t test. The results are representative of two independent experiments. Page - 203 l Smad3'/' Smad3”+ 4 1 3- 2 1 OOF X 2de 03m; UQLm>OUmm 13' 11.18112? W b} TGF-ii ~33:an»‘:5 :: 115115013183: 0.001 0.01 No . or 3cm: 1“. 0.1. and 12;: TGF-[31 (ng/ml) Treatment ’I + .II /.# F4 dm aa mm 55 HI _ _ _ W W W W 0 0 5 0 5 2 1 1 treated 1:: '3 8:60:33 @2989. 03 B“? n13 .... m .m x: V. .1. ”x. ,. . uh. ..Mu. ' 1 K1 .1 t . \1 .H. rum 0.. n1 1 W» I 11w“ W Ix u .au 0 :1“ “C TM I an. ..I m .\ L at p . . 1». an S 11. JUN ..rCu m .9)». 10 he .5 m I l l 40 NA .. t \J U .C Page - 204 in the presence or absence of TGF-B1 for 3 days, and supematants were collected for ELISA. LPS-induced IgM secretion was attenuated by TGF-[3l in a concentration- dependent manner in SMAD3"' B cells as well as SMAD3” B cells (Figure 60). In agreement with the in vitro AFC responses, SMAD3"' B cells were less sensitive to inhibition by TGF-[3l when compared with wild type littermates. Moreover, a greater magnitude of LPS-induced IgM production in Smad3"’ B cells than SMAD3+l+ B cells corresponds to the elevated polyclonal IgM AFC response in Smad3"‘ spleen cells. D. IgA secretion is not enhanced by TGF-B, in LPS-stimulated Smad3- null mouse splenic B cells in vitro TGF-Bl is well established as a positive regulator of IgA production. Recent in vitro studies have implicated that Smad proteins physically and functionally interact with transcription factors that regulate IgA transcription in response to TGF-B..(Park et al. 2001; Shi et a1. 2001; Zhang and Derynck 2000). The objective of this set of experiments was to investigate the role of Smad3 in IgA production by TGF-[3l in vitro. Splenocytes were isolated from Smad3"' mice, stimulated with LPS and IL-10 in the presence or absence of TGF-B, for 5 days, and supematants were collected for ELISA analyses. While an expected increase (6X) in IgA production by TGF-[3l was demonstrated in LPS- stimulated B cells from Smad3”+ mice, the production of IgA in response to TGF-[3l was ablated in Smad3"‘ B cells (Figure 61a). A concomitant decrease in total IgM secretion in the LPS-stimulated Smad3+l+ B cells by TGF-B, is consistent with TGF-Bl-induced IgA class switching (Figure 61b). Interestingly, LPS—induced IgM secretion by TGF-[3l was also attenuated in Smad3"' spleen cells (Figure 61b). One interpretation of these results is that TGF-[3l induces IgA class switching in Smad3"’ B cells; however, the mature Page - 205 Figure 60. Inhibition of LPS-induced IgM secretion by TGF-B, in vitro is unaffected by Smad3-null splenic B cells in vitro. Spleens from Smad3"' and Smad3“+ were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 5 X 106 c/mL and transferred in SOO-uL to the wells of a 48-well culture plate. Quadruplicate cultures were left untreated (no treatment) or activated with LPS (10 ug/mL) in the presence or absence of TGF-Bl (0.001, 0.01, 0.1, and 1 ng/mL final concentration in culture) for 3 days. Alternatively cells were treated with the vehicle (VH; 0.02% PBS final concentration in culture, pH 3.5, containing 0.1% BSA) for 3 days. Supernatants were collected at 72 hours and analyzed for total IgM by ELISA. The results from quadruplicate determinations are expressed as the mean IgM (11g 106 recovered viable cells i SEM (n=4). “a” denotes p < 0.05 as determined by Dunnett’s t test and compared to the Smad3“ vehicle control. “b” denotes p < 0.05 as determined by Dunnett’s t test and compared to the Smad3"’ vehicle control. “#” denotes p < 0.05, Smad3“+ compared to same group Smad3”' as determined by Dunnett’s t test. The results are representative of two independent experiments. Page - 206 le ~13] in rim: 71.113" and Sat“ 1e splenoqte: "-2 1e tells of air—1i nt 1 or actuate i: 1. 012111111:- :re treated 111‘: :taim'ng .1153? d for [013111111 rd as the mean 1'1 i as determined? lenotes p < 015‘ ontrol. “if 15915 N Dunnett’silfi 1751 # T 150‘ 125- 100‘ 75-: 50- 25-1 T O-M/ A N0 0 0.001 Treatment Total IgM my 106 Recovered Splenocytes) Smad3”+ T. Smad3'/' A 0.01 0.1 1 TGF-[31 (ng/mL) LPS Page - 207 Figure 61. IgA secretion is not enhanced by TGF-[31 in LPS-stimulated Smad3- null mouse splenic B cells in vitro. Spleens from Smad3"‘ and Smad3“ were isolated aseptically and made into single cell suspensions. The splenocytes were washed and adjusted to 5 X 106 c/mL and transferred in SOO-uL to the wells of a 48-well culture plate. Quadruplicate cultures were left untreated (no treatment) or activated with LPS (10 ug/mL) in the presence or absence of TGF-B, (1 ng/mL final concentration in culture) for 5 days. Alternatively cells were treated with the vehicle (VH; 0.02% PBS final concentration in culture, pH 3.5, containing 0.1% BSA) for 5 days. Supernatants were harvested and subsequently analyzed for total (a) IgA and (b) IgM secretion by ELISA. The results from quadruplicate determinations are expressed as the mean IgA (nanogram per 106 recovered viable cells i SEM) or mean IgM (microgram per 10" recovered viable cells i SEM) (n=4). “a” denotes p < 0.05 as determined by Dunnett’s t test and compared to the SMAD3“+ vehicle control. “b” denotes p < 0.05 as determined by Dunnett’s t test and compared to the SMAD3"‘ vehicle control. “#” denotes p < 0.05, SMAD3“+ compared to same group SMAD3"’ as determined by Dunnett’s t test. The results are representative of two independent experiments. Page - 208 ~Iimulaled Sui 433°" were :12: 5 “618 “atria S-Weil enlarge lied ll'm'l LPS 1 2111011 in cult: i: 0.021 PBS 5; Superman: 1: tremor. 1)} E131. .ln IgA a." l‘ recovered. lie: zest and COL-1?”; l} Dunne-23:15 0.05. SMAD?” . 1116 11511633 (a) +/+ 50‘ @Smad3 it? '5 40- 0 ‘8: 1- 30- \ C) 5 < 20- 2’ a 1.. 0 w I.— O No lL-1O + TGF-B1 Treatment LPS b (b) 200 _I. 01 O l '10" 49’ 101 Total IgM (pg/106 cells) 0‘ ES ‘1 a: # 0W- Page - 209 protein product is not generated due to a disruption of an essential TH2 cytokine environment necessary for the translation of TGF-Bl-induced sterile mRNA transcripts with Smad3 deficiency. This interpretation is supported by elevated production of IFN-‘y, a characteristic TH] cytokine, by Smad3"’ T cells (Figure 39a). Moreover, the magnitudes of basal and LPS-induced IgA protein are greater in SMAD3"’ than SMAD3“ splenocytes (Figure 61a). These results demonstrate that SMAD3"‘ B cells can produce IgA in vitro. Page - 210 DISCUSSION It has long been established that TGF-[31 is essential for maintaining immune homeostasis. More recently an essential role for TGF-B signaling in T cells has been established. Probably the most compelling supportive evidence is exemplified by in vivo spontaneous differentiation of T cells into effector cytokine producing cells and autoimmune manifestations in transgenic mice that have impaired TGF-[3I signaling specifically and exclusively in T cells (Gorelik and Flavell 2000). Over the past several years, numerous in vitro studies have delineated a rather confusing and sometimes contradictory role for TGF-[31 in regulating T cell immune homeostasis. For example, T cell growth is impaired by high concentrations of TGF-[3I (Stoeck et al. 1989a), while low concentrations of TGF-[3l reportedly stimulate T cell growth (Kondo et al. 1993). In one report, TGF-[3l down-regulated IL-2 receptors (Kehrl et al. 1986c), while in a second report had no affect on IL-2 receptor expression (Smyth et al. 1991). TGF-[3l also paradoxically induces (Weller er al. 1994) as well as suppresses T cell apoptosis (Cerwenka et al. 1996). Moreover, TGF-Bl-null mice succumb to a CD4+ T cell-mediated multi-organ autoimmune inflammatory disease (Diebold et al. 1995; Shull et al. 1992). Transgenic mice overproducing TGF-B, also succumb to chronic inflammation (Kim et al. 1991). TGF-Bl acts on all immune cells, and elicits its effects through multiple mechanisms, thus it is not surprising that many of the reported effects of TGF-B, using completely different experimental models are complicated and disparate. The toxicological and pharmacological implications of TGF-Bl on T cell homeostasis and Page - 211 tolerance are profound and resoundingly establish a necessity for delineating the mechanisms whereby (1) TGF-B, modulates T cell function and (2) TGF-[3l signaling in T cells regulates immune homeostasis . . . hence the relevance of this research. The overall working hypothesis that was put forth to test in this dissertation research is as follows. ‘TGF-fi, acts directly on T cells in a receptor-dependent manner to regulate IL-2 expression through Smad-mediated intracellular signaling.’ The overall specific aims for this research were three-fold: (1) Develop and validate an in vitro model to characterize the effects of TGF-B, on T cell growth and IL-2 expression. (2) Determine whether Smad signaling plays a role in the regulation of IL-2 expression by TGF-[3,. (3) Determine whether cross talk between MAPK and Smad signaling plays a role in the regulation of IL-2 expression by TGF-Bl. This discussion is organized and presented in five sections to address the aforementioned working hypothesis and specific aims. Section I describes the concentration- and time-dependent regulation of T cell growth and IL-2 expression by TGF-[3,. The second section presents evidence that Smad3 is essential for inhibition of T cell growth and IL-2 production by TGF-B, in vitro. Section HI describes a putative role for DNA sequence specific binding of Smad3 in the regulation of IL-2 expression by TGF-[3,. The fourth section discusses evidence for putative MAPK and Smad signaling cross talk in the regulation of TGF-[31 signaling in T cells. And the last section, section V, discusses a role for Smad3 in the regulation of T-cell dependent and T cell- independent humoral immune responses by TGF-[3l in vitro. Page - 212 I. Concentration- and time-dependent regulation of T cell growth and IL-2 expression by TGF-[3l TGF-[3l mediates its effects through the activation of a heteromeric complex of transmembrane receptors with intrinsic serine threonine kinase activity. Despite an overwhelming amount of newly-acquired mechanistic data regarding the regulation of TGF-Bl-responsiveness through Smad-dependent processes, Smad signaling in T cells remains relatively unexplored. This is probably due, in part, to the pleotropic nature of TGF-B,-mediated immune responses. In light of these often seemingly paradoxical pleotropic responses by TGF-Bl, one of the initial focuses of this research was to establish an in vitro model to investigate a role for Smad-dependent signaling in the regulation of T cell growth and IL-2 secretion by TGF-B1. Lymphoid responsiveness to TGF-[3I depends on the type and state of the cell (e. g., activation, differentiation, and maturation) (Cerwenka et al. 1994; Ludviksson et al. 2000; Wahl et al. 2000; Weller et al. 1994; Yates et al. 2000). Therefore several experimental variables including TGF-B, concentration, CD28 co-stimulation, and the time of addition of TGF-[3l relative to T cell activation were compared in primary mouse splenocytes and thymocytes to optimize TGF-B1 responsiveness. Primary thymocytes were incorporated into the study design to investigate the effects of TGF-[31 on T cells in the absence of antigen presenting cells, i.e., B cells, macrophages, and dendritic cells. Physiological concentrations of TGF-Bl (Ahmad et al. 1997) were employed. Notably, a direct comparison between the biological activity of in vivo and in vitro TGF-[3l concentrations is limited due to the secretion and distribution of TGF-[3l as a latent, biologically inactive precursor molecule (Miyazono and Heldin 1989). Splenocytes and Page - 213 thymocytes were isolated and activated in vitro using antibodies against the TcR/CD3 complex and the CD28 co-stimulatory molecule to pharmacologically mimic antigen- induced complete T cell activation. A similar concentration-dependent inhibition of splenic T cell and thymocyte growth by TGF-[3l is suggestive of a direct effect by TGF-13l on T cells. The T cell growth inhibitory effect by TGF-B, was further demonstrated to be dependent upon the time of addition of TGF-[3l to the cell cultures relative to when the T cells were activated. This temporal relationship was interpreted to implicate regulatory cross talk between signaling cascades activated downstream of the TcR and the TBR. The markedly varied responsiveness to TGF-[3l over a relatively short interval (e. g., 30 minutes, in some cases) further suggested the involvement of an early, upstream signaling event. Surprisingly, CD28 co-stimulation did not markedly influence the inhibition of T cell growth by TGF- B, implicating a lack of regulatory cross talk between CD28-associated and TBR signaling. In contrast to T cell growth, concentration-dependent bifunctional stimulation and attenuation of steady state IL-2 mRNA expression as well as IL-2 protein secretion by TGF-[3l was demonstrated in activated splenocytes. CD28 co-stimulation augmented the stimulatory effect of low concentrations of TGF-[3l on splenic T cell IL-2 expression, but did not influence the inhibitory effects by high concentrations under conditions of simultaneous addition of TGF-[3l and T cell activation. Time of addition studies substantiated a regulatory role for CD28-associated as well as TcR signaling in the bifunctional concentration-dependent stimulatory and inhibitory effects of TGF-B, on IL- 2 secretion. Importantly, quantitative RT-PCR analyses provides evidence suggesting Page - 214 that the bifunctional modulatory effects on IL-2 secretion by TGF-[3I are mediated though direct effects on IL-2 gene transcription. Collectively, these results support a hypothesis that cross talk among TcR, TBR, and CD28-associated signaling regulates IL-2 expression by TGF-B, in activated T cells. In light of these in vitro observations, it is tempting to speculate that similar in vivo bimodal concentration-dependent effects by TGF-[3l on immune cell function may also be regulated through direct effects of TGF-[3l on gene expression. These studies are the first to describe the concentration-dependent effects of TGF- B, on IL-2 expression, at the level of protein secretion as well as mRNA expression, in Ot- CD3 + Ol-CD28-activated T cells. In deciphering the observed stimulatory effects of low concentrations of TGF-[3,, putative indirect contributory factors warrant consideration. For example, low concentrations of TGF-B, may selectively target populations of T lymphocytes expressing elevated levels of high affinity TGF-B receptors; these T cell ‘subpopulations’ may be inherently stimulated by TGF-Bl. Accordingly, it has been demonstrated that TBR expression and affinity is dependent upon T cell maturation and differentiation (W ahl 1992). Alternatively, TGF-Bl may, in a concentration-dependent manner, differentially activate intracellular signaling cascades, e.g., Smads and MAPKs (Piek et al. 1999). Cross talk among these different signaling cascades as well as between TBR, TcR, and CD28-associated signaling may provide putative mechanisms whereby TGF-B, is able to differentially regulate IL-2 expression in a concentration-dependent manner. This latter possibility is particularly interesting in light of the observation that TGF-[3l modulates the activity of several transcription factors critical for regulating IL-2 transcription including Page - 215 AP-l, CREB, and NF-KB (Brabletz et al. 1993). Notably, each of these aforementioned transcription factors has been demonstrated to bind to the CD28RE in the IL-2 promoter (Shapiro et al., 1997). Upregulation of TBR expression following T cell activation also warrants consideration as a putative contributing factor for the bifunctional T cell responsiveness to TGF—B, (Ellingsworth et al. 1989). Based upon the observation that a temporal loss of TGF-Bl-induced augmentation of IL-2 secretion occurs only under conditions of CD28 co-stimulation, it is possible that CD28-associated signaling may contribute to the upregulation of TGF-[3 receptors in activated T cells. Alternatively, CD28 signaling may modulate the effects of TGF-[3I on a-CD3-activated T cells through direct stimulation of cytokine production, growth survival signals, or stablization of IL-2 mRN A (Boise et al. 1995; Shapiro et al. 1997). Notably, a requirement for CD28-associated signaling for augmentation of IL-2 secretion by TGF-B, is consistent with previous studies (Aoki et al. 1991). Probably, the single most important biological significance gleaned from this set of experiments is the demonstration that low concentrations of TGF-B, augment steady state IL-2 mRNA and IL-2 secretion in a—CD3 + a-CD28-activated T cells and high concentrations of TGF-B, paradoxically attenuate IL-2 expression under similar conditions of T cell activation. Putative mechanisms for this phenomenon have been discussed above. In light of the observation that intracellular signaling by TGF-[3I permits interactions among multiple receptors and intracellular signaling intermediates allowing for a diversity of biological responses, it is proposed that the differential regulatory effects of TGF-[3l on IL-2 expression are mediated through the cooperative interaction of Page - 216 multiple Smad-interacting signaling cascades that are activated by TGF-[3,, in a concentration-dependent manner. 11. Smad3 is essential for inhibition of T cell growth and IL-2 expression by TGF-B, in vitro At the onset of these studies, a role for Smad signaling in TGF—Bl-responsiveness had been established in numerous cell types; however, Smad protein expression had not been established in lymphoid tissue. Western blot analyses confirmed that the TBR activated-Smads, Smad2 and Smad3, as well as the co-Smad, Smad4 are expressed in murine splenocytes and thymocytes. Moreover, these studies demonstrated that TGF-[3l activated Smad3 in splenocytes in a concentration- and time-dependent manner. These results are in agreement with other studies demonstrating rapid and transient activation of Smad proteins by TGF-[3l (Shen et al. 1998). Our studies further established a role for Smad3 in mediating the inhibitory effects of TGF-[3l on IL-2 expression. Anti-CD3 + a-CD28-activated Smad3-null splenic T cells and thymocytes were refractory to the inhibitory effects of TGF-[3, on steady state IL-2 mRNA and IL-2 secretion. These observations are consistent with reports demonstrating a role for Smad3 in maintaining T cell-dependent immune homeostasis (Datto et al. 1999; Yang et al. 1999). While it has been previously reported that T cells from Smad3-null mice are resistant to growth inhibition by TGF-B, (Datto et al. 1999; Yang et al. 1999), our studies were the first to demonstrate this refractory phenomenon in fully activated mature T cells. Moreover, our results are in agreement with others (Datto et al. 1999; Yang et al. 1999) Page - 217 demonstrating that inhibition of LPS-induced B cell growth by TGF-B, in vitro is unaffected by targeted deletion of Smad3. A cause and effect relationship between the inhibition of IL-2 and T cell growth is supported by the ability of exogenous IL-2 to reverse the inhibition of T cell growth by TGF-B, in activated T cells. It has previously been established that in vitro LPS-stimulated B cell growth does not require B cell autocrine or T cell-derived IL-2 (Hashimoto et al. 1986). Collectively, these results suggest that TGF-Bl may differentially regulate lymphoproliferation in Smad3-null mice through direct regulatory effects on IL-2 expression. In light of these observations, a model is presented that defines a novel mechanism whereby TGF-B, may differentially regulate proliferation of B cells and T cells (Figure 62). Characteristically, immune competent naive T cells in peripheral tissues predominantly secrete IL-2 upon stimulation. In contrast, differentiated effector T cells generated during an immune response secrete a battery of cytokines characteristic of a TH] or TH2 phenotype. Naive Smad3-null splenic T cells display a differentiated Tnl phenotype as demonstrated by elevated basal IL-2 and IFN-y expression. These results are consistent with an in vivo activated Smad3-null peripheral T cell phenotype, as demonstrated by increased CD62L surface expression (Yang et al. 1999). Moreover, elevated IFN-y expression in Smad3-null T cells correlates with a phenotype Activation of IFN-y/STAT pathways has been shown to antagonize TGF-B,/Smad signaling, at least in part, by upregulating activity of the inhibitory Smad7 (Ghosh et al. 2001; Ulloa et al. 1999). In light of the elevated IFN-y levels observed in naive and activated Smad3"' splenocytes and thymocytes, we verified that abrogation of the inhibitory effects of TGF-B, on SMAD3"‘ T cells was not due to an indirect IFN-y— Page - 218 Figure 62. Putative model for Smad3-dependent inhibition of T cell growth by TGF-[3,. This model proposes that TGF-[3l differentially attenuates T- and B-cell growth through Smad3-dependent and Smad3-independent mechanisms, respectively. Moreover, Smad3-dependent inhibition of T cell growth by TGF-B, is through a direct inhibition of IL-2 mRN A expression. Page - 219 .5233... Emu—Eng .mn—Szw Ann—<55 _llll fling—.14 Fad—mum. .._W...1.1....M_........,: 3 <5... ..B