The generation of cytotoxic T cell epitopes and their generation for cancer immunotherapy Kessler, J.

Hele tekst

(1)The generation of cytotoxic T cell epitopes and their generation for cancer immunotherapy Kessler, J.. Citation Kessler, J. (2009, October 27). The generation of cytotoxic T cell epitopes and their generation for cancer immunotherapy. Retrieved from https://hdl.handle.net/1887/14260 Version:. Corrected Publisher’s Version. License:. Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden. Downloaded from:. https://hdl.handle.net/1887/14260. Note: To cite this publication please use the final published version (if applicable)..

(2) CHAPTER 7.

(3) 7 RETPAHC.

(4) Discussion. Cytotoxic T cell immunity is at the basis of the majority of immunotherapeutic approaches for cancer, in particular adoptive T cell trans ! " #$$" % $$ $$&

(5) '+ "! ( $"" () prominently involved in the eradication of $$ * (& . () the main goal of therapeutic vaccinations not $" / $ Basically, the success of any T cell-based im" $" 0 " 0$ " () *% ( 1 successful therapeutic CTL response, next to an optimal immunostimulatory context of the therapy, the proper selection with respect to 0 " 0$ " 2 associated or pathogen-derived antigens, and, more precisely, the T cell epitopes contained $ 3cess rates in T cell based immunotherapy for $$ $* 45& 6 $ % 7891:8

(6) 3 /$ " / $ 4;&<5& but our growing understanding of immune activation and antigen processing will likely $ 6" " In this thesis, new insights in the processing of /" 7): $ 8 $$ $ =& immunology strategy of CTL epitope identi# $ " >?:@A D?2:D) ;& E * # >?:@A $" <. 7& $ $ cussed in the context of immunotherapy for .

(7) immunology ( /" 7): $ I molecule is the end result of various events in 2$$ *" 8 only since the early 1990s that our knowledge concerning the three predominant processing events has evolved to the level that these events can be experimentally tested, allowing a prediction of the peptides that end up on $$ 8 # " prediction, only class I binding was taken into account, but, obviously, the accuracy of such a prediction is enhanced when proteolytic $ 8 / estimated that, on average, less than one third $$ /$ 7): $ 8 / a protein are actually generated by intracel$$ $" 4E5 ( * ;& * $ ! 7):2 A0201-binding peptides at their C-terminus was used as extra selection criterion to predict CTL epitopes in tumor associated antigen (:: >?:@A ( " diction was enhanced, strongly limiting the number of peptides of which natural presenta / $ /" () $ M & # ; ( ! $ # teasomal excision, being much less laborious than CTL inductions, thereby enhanced the $$ 6 " # # 2 *%.

(8) .

(9) N & / $ & ; (:> $ & < $ 8 / & *" algorithms are available that allow the in silico $ 8 $ %* In silico assessment of class I binding and $ $ (:> $ & which is much less selective) consumes only a little amount of time, but is also much less ac ! $ Upon comparison of class I peptide binding $ & # . : !$& we calculated percentages of overlap between /" D8@:3 3PN>A8(78 $ * & / ;Q& Q QQ $$ $ >?:@A QR $ 7): $ 8 $$ ( overlap in the best 20 predicted binders may / $* ;S $ 8 $$ Table 1. Overlap between predictions by !" #$ %&'. Table 2. Accuracy of binding prediction. a. Experimental IC50 rank. best 50 ranked. best 100 ranked. 9-mer 10-mer. 40% 50%. 56% 60%. 64% 62%. HLA-A2. 9-mer 10-mer. 50% 70%. 64% 68%. 68% 68%. HLA-A3. 9-mer 10-mer. 50% 25%. 54% 34%. 59% 54%. HLA-A1. a. Peptides with identical scores of those at position 20, 50 or 100 were included.. (/$ :$& % /" silico predicted binding does not perfectly correlate with the actual binding measurements and false positive prediction of binding oc N & * <th predicted R2 >?:@A >?:EE2;) actually to $% / " 7):2:Q;Q& *. ;

(10) . c. Peptide sequence. aa.. 1. 2.5. 16. NLTHVLYPV. 435. 2 3. 2.9 3.7. 1 2. QLLALLPSL g SLLQHLIGL. 394 425. 4. 4.6. 15. TLAKFSPYL. 248. 5. 5.1. 5. 6. 5.2. 8. ALAIAALEL g VLDGLDVLL. 100. YLHARLREL. 462. 7. 5.4. 4. 8. 5.7. 9. 6.8. 34 12. 10. 9.2. 21. 11. 9.3. d. d. 12. 10.2. 245 22. 13. 11.0. 33. 53. 39. SISALQSLL. 419. GLSNLTHVL. 432. ITDDQLLAL. 390. CTWKLPTLA. 242. LLKDEALAI. 34. d. TLSFYGNSI. 410. d. HLHLETFKA. 91. d. TLSITNCRL. 326. 14. 11.1. 15. 13.2. 16. 13.4. 41 9. AVLDGLDVL. 99. 17. 14.0. 3. RLRELLCEL. 466 422. 18. 14.2. 7. ALQSLLQHL. 19. 15.7. 10. RLDQLLRHV. 312. 20. 15.8. 215. RTFYDPEPI. 493. d. e. Predicted binders (in best 20) not (efficiently) binding. a b. best 20 a ranked. b. Prediction rank. c d e g. e. 35. 35.0. 11. 43. 71.3. 13. 45. 79.4. 18. 59. >100. 14. LLERASATL. 372. e. ALELLPREL. 44. e. SLFFLRGRL. 305. e. KMILKMVQL. 224. As measured in HLA-A2 binding assay. Prediction by SYFPEITHI algorithm of 9-mer peptides derived from PRAME binding in HLA-A*0201. Start aa. position in PRAME and sequence. False negative prediction of binding. False positive prediction of binding. Proven CTL epitope.. the 16th / >?:E<2EE<) bound * 6 " (/$ ; As discussed in chapters 1 and 2, in silico prediction of proteasomal excision is inaccurate *$$ ( * $ $ * * # $ integrate prediction of class I binding, protea$ ! (:> $ & / @72>*" 45& W:>> 4=5& () 4 5& 8A

(11) D 4'5& A X 4R5 : 2 =E >?:@A * 7):2:;2/ " proteasomal liberation to identify two CTL.

(12) ; 4Q5 * (/$ < The two proven nonameric CTL epitopes were predicted in the 10 best ranked predicted /" # $ @72 >*"& * % >?:QQ2Q' epitope in the top-10, but instead correctly >?:<Q2<QR * $ M /$ $ ( / algorithm from the ‘immune epitope database $" Y 8A

(13) D 4'5 $$ 7*& # /$ / Z QS falsely predicted epitopes, mainly because the 2 * /" (/$ < ( $ ! $ # $ $ is needed to improve the selection of putative : $ *% integrative algorithms is that they mostly allow $" * !tions for certain alleles), and their coverage of $ 7): $ 8 $$$ $ [$$" () & $$ # /$ /* of work in the prediction phase when predic $ $$" $ * ! $ # $ 0$ " #$ & * * $$ $ $/ *% and failed T cell inductions) in the validation ( /$ " 0 & $ $ "\ $ protein, the better the chance that top-scoring * $$ / N $$ ] 6 " the reverse immunology approach it is important to note that the prediction of class I ligands aims to select only those peptides that are enzymatically generated, survive further. Table 3. Prediction by combined algorithms. Start. Nonamer. Predict.. Experimental. Ranka. Bind.b. 1 2 3 4 5 6 7 8 9 10. 4.6 3.7 2.5 2.9 6.8 n.t.d. SLLQHLIGL GLSNLTHVL FLRGRLDQL NLTHVLYPV e VLDGLDVLL SLFFLRGRL RLDQLLRHV QLLALLPSL ALAIAALEL VLVDLFLKE. 1 2 3 4 5 6 7 8 9 10. 3.7 6.8 16.1 2.5 5.2 79.4 15.7 2.9 5.1 n.t.. QLLALLPSL TLAKFSPYL e SLLQHLIGL NLTHVLYPV e VLDGLDVLL YLHARLREL GLSNLTHVL ALAIAALEL KMILKMVQL ALQSLLQHL. 1 2 3 4 5 6 7 8 9 10. 2.9 4.6 3.7 2.5 5.2 5.4 6.8 5.1 >100 14.2. QLLALLPSL e SLLQHLIGL LYVDSLFFL TLAKFSPYL SLQCLQALY e VLDGLDVLL YLHARLREL NLTHVLYPV ALQSLLQHL KMILKMVQL. 1 2 3 4 5 6 7 8 9 10. 2.9 3.7 6.3 4.6 n.t. 5.2 5.4 2.5 14.2 >100. QLLALLPSL e SLLQHLIGL NLTHVLYPV e VLDGLDVLL KMILKMVQL TLAKFSPYL GLSNLTHVL ELFPPLFMA RLDQLLRHV FLRGRLDQL. 1 2 3 4 5 6 7 8 9 10. 2.9 3.7 2.5 5.2 >100 4.6 6.8 >100 15.7 16.1. aa.. C-term.c. MHC-Pathway 248 425 435 394 432 284 301 410 340 353. TLAKFSPYL e SLLQHLIGL NLTHVLYPV QLLALLPSL GLSNLTHVL YIAQFTSQF e LYVDSLFFL TLSFYGNSI MHLSQSPSV VLSLSGVML. WAPP 425 432 308 435 100 305 312 394 39 177. e. 6.3 11.0 n.t. 17.4. ++. n.t.d ++. n.t. n.t. ++. n.t.. ++. n.t.. -. NetCTL 394 248 425 435 100 462 432 39 224 422. ++. ++ +. n.t.. -. IEDBg 394 425 301 248 294 100 462 435 422 224. ++ ++ ++ + n.t.. EpiJen 394 425 435 100 224 248 432 51 312 308 a b c. ++. ++ n.t. n.t n.t. Ten best scoring nonamers from PRAME. Binding score in IC50 (lower score, is higher affinity). Cleavages with immunoproteasomes after 1 h digestion. Index: (-) no cleavage behind C-term. of peptide, (+) low abundant cleavage, (++) abundant cleavage.. d. n.t.; not tested.. e. published (two) and unpublished (see text) epitopes.. g. Prediction using the ARB method (see chapter 1, table 2)..

(14) <.

(15) "$ /" 6 $ A? / * 6 " $/$ $ 8 $$ A $$" constraints of proteolysis and class I binding render only a small fraction of all 9-, 10- and 11-mer peptides in a protein available for cell N& basis of kinetic data, it has been calculated RRS $$$ " / (:> 45 0$"& $ 8 $ 6 : $* / dicted epitopes, therefore, is a sign of strength of the prediction phase, provided that these ^ ! ; 4Q&;5 4<5& ! tion procedure including proteasomal digestion analysis, which selects peptides with high $ 8 / 6 " 2 $$" liberated by an abundant proteasomal cleav & " $" () Obviously, selection of peptides according to less stringent selection criteria may result in 2! $ 8 $ N " $" $ dation of putative epitopes are the lack of 2 # ( $$ ertoire of the chosen blood donor, which may be caused by tolerance in the case of tumor . & /$" $* sensitivity of the induced CTL, and unfavour/$ * () $ ( improve success rates, several adaptations have been made in the T cell induction protocols 4E5 * $ $ ! # () /" $ cytokine-secreting or tetramer-positive T cell $ 45 : $ 6 " 0 how often peptides are predicted falsely nega & *\ * " E

(16) . ` : $ " have been chosen too stringent, falsely negative predictions can result from intrinsic weak N & * / 6ity may be missed because the algorithms are not completely covering all possible positive . / $% $ / 4=5& which will score low in binding prediction and * $$ / $ $ & " $ / " !$ * (/$ ;\ (Wq)>(): $% $ 7):2:; * 2 : important point is that to date the majority of binding prediction algorithms allow only predictions of nonameric and decameric peptides 7): $ 8 $$& fact that a substantial number of class I ligands 2 ; ;R' 7):2:; $ 3PN>A8(78 / 4 5 $$ $ () / 4'2;Q5 3$"& $$ $ * $$ $ / ( incorporation of only proteasomal C-terminal excision in the procedure results in falsely negative prediction of the unknown fraction of CTL epitopes that are C-terminally liberated by a proteasome-independent mechanism 4;&;;5 8 / $$ ! mental digestion results, for instance, that the likelihood of a proteasomal cleavage after a lysine is very low, although a high number $" 7):2:< * 2 $ $" 45 ( $ $" " (& # $ 2 terminal generation predicts the intracellular generation of most class I ligands, but will miss those that are C-terminally liberated by cytosolic endopeptidases like nardilysin and (^> =& /$* 8-.

(17) tion of nardilysin-dependent processing in 2 " $ # $ $ /$* : additional category of class I ligands being missed by reverse immunology are peptides $ $$" # 4;<&;E5 or are produced intracellularly by uncommon $ % $ 4;&;=5 In summary, when applying experimental veri# * $ prediction phase, reverse immunology is ex$" *$$ $$" 7): class I presented ligands of which the immu " # / / # ( $$ ^ the other hand reverse immunology will miss an unknown percentage of ligands/epitopes of which the restriction may be biased to certain $$$ 7):2:< N $$"& $$" " $ 7): $ 8 $ecule, the length of the source antigen is obvi$" $ N & ;& $" 7):2:;2 * >?:@A * $ QR 4Q5 :$ / $" full picture, this would very roughly mean one QQ $ $ 8 $$$ * ! $" QS ! $" ;QQ /$ .

(18) Although the reverse immunology strategy / $/$ # 7): $ 8 & " / /" 6 ( genomics and proteomics in the recent past has enabled the introduction of large-scale high-throughput screening methods, both for tumor antigen discovery and T cell epitope. # 8 ! $" increasing power of mass spectrometric tech 0 * $$ $ 2 # 2 # 7): $ 82/ |$ Y (& # 7): $ 8 $ by reverse immunology-based predictions may $$" / /" /" # of cell surface presented peptides with mass " & }'E :$" ( $$ # 7): $ 8 $ $" test the immunogenicity of the epitope, and $ $ $$ ! : 0 /" mass spectrometry of peptides eluted from the cell surface is that it will probably identify, next to numerous novel ‘conventional’ ligands, $ 7): $ 8 $ * 2 $ binding motifs, extraordinary length or post$ $ # @ ligands may have been produced by nonconventional enzymatic mechanisms, possibly 2 These categories of class I presented peptides * $$ / # /" $" . Post proteasomal and proteasomeindependent class I antigen processing : }EE=& $ important issues are still unresolved in class 8 : ] 0 involvement of cytosolic endopeptidases in $ 8 An accumulating body of published evi # / proteolysis independent of the proteasome 4;&;;&; 2<Q5 7*& $ without exception use proteasome-inhibitors as primary tool, and because it is known

(19) .

(20) / $$ $$* # residual proteasome activity, the real contri/ & / 0$ $" 0 $"& of non-proteasomal proteolysis remains to / (& = primary goal was to identify a CTL epitope that is indisputably made at its C-terminus /" 2 Using a reverse immunology approach and * 6 " 7):2:< / >?:@A RQ2R' * excised at its C-terminus by the proteasome, $" $ (^> were demonstrated to jointly liberate the C >?:@ARQ2R' $"sin has the preference to cleave before or in the $ / ~<Q $ 4<5 ( "$ is implicated in the N-terminal excision of the 7):2D; 2 * / 2 $ =& N ED&&

(21) ?%/$"& 7):2D; * ously found to present a high proportion of 2 / 4<Q5 7*& * " * =& N ED& st, 2nd, 4th and 7th peptide), only the N-terminus was lib /" $" N & the epitope is located at the C-terminus of the th peptide), explaining proteasomeindependence, but the other three peptides apparently need either another endopeptidase $ / 2 $ (>>88 * !$ 4<;5 $ 2 " /$ 7 # $" (^> $ 8 & 0 *& and if so, how often, completely proteasomeindependent CTL epitopes really exist is not " * 8 . ! /* 2 independent generation of the epitope’s =

(22) . 2 * 0 2 %* tion of epitopes) as a result of post-proteasomal processing, on the one hand, and completely proteasome-independent epitope-generation N / & existence of the latter category of epitopes will remain an open issue, because complete inhibition or silencing of the proteasome is impos /$& $ $ / #$$" and completely inhibiting one catalytic activity $/$ =& $ N << ( 0tion is also strongly related to one of the other unknowns of class I antigen processing, being the precise source and nature of the substrates *" ?$ $"

(23) ? > " / "$" * " involvement of the proteasome, because such substrates would not need to be unfolded and are possibly within the length restrictions of class I presented peptides that are C-terminally liberated in a proteasome-independent * $ 8 # /" " of all peptides from one source protein that are presented on the cell surface of single-class I $$$ ! $$ /0$" lyzing their C-terminal liberation by in vitro proteasome-mediated digestions will provide 0 ( 0 *& $" /$ perspective because it possibly challenges the / 0 most C-terminal epitope excisions, but also 2 # $ : & * $" $ * >?:RQ2R' epitope = (>>882 789 4;;5 $" ! !$.

(24) / $ " (>>88 * 4<E5 $ / 7):2:< * 2 $ $" $ % 789 4;;5 ( *$ % / $ $" 6 " $ / $" 45 7*& /$ . 4<;&<2 <'5 () 0 $$" $ $" " (>>88 / Reverse immunology will also help to shed $ 8 $ opes predicted to be C-terminally released by $" =& N E: # / $$" lysin-dependent fashion, this would indicate $" $ 6 W / opes will be found to be C-terminally liberated /" $ : /" $ 4;&; &<Q5& " / relation between the chemical nature of the 2 $ 0 2 A $$" 7):2:< 4;&; 5 7):2D; 4;&<Q5& * / $$" harbour peptides with a basic C-terminus, were / 2 / ( " /$" 2$ @7 $ 8 2/ 7*& $ (^>& * $" $ $ 0 # " 4<R&EQ5& * 2 $ % #$ 2 $ /" $ < =& 2 independence may not be skewed too much * # 2 $ / Therefore, it will be especially interesting to in * $ 8 $$ $ % 7):2 A2, which has not been found to be especially proteasome-inhibitor insensitive, presents. peptides that are C-terminally excised by non$ "$" $ * $ (^> $ 8 processing was considered to be destructive, however this notion is heavily based on results from over-expression experiments that skew " 4E2E<5 ( " (^>Y ! A)N3P)8Aq2 = $ $ $ (^> $ 8 \ limits the presentation – by partial destruction – of epitopes whose correct C-terminus has al" / /" $ % 388NAq) 4E<&QR5& & & produces class I binding peptides by trimming 12–14 meric C-terminally extended epitope precursors that lack a C-terminal class I bind ( $ with a recent biochemical study demonstrating (^> / " $ $ 8 $ 4Q5 : 0 "$ * ! (>>88 4;;&EE2E=5& $" (^> $$ cytosol-resident endopeptidases are involved $ 8 $ ( swer here is that there is no reason why any cytosolic endopeptidase would not be involved in protein degradation, and, thus, possibly $ 2 A!$ " $" 4E 5 $ " 4E'5& $ " been suggested to be associated with the pro 4ER5 $ $ $ 4Q&5. Practical and theoretical implications of processing The knowledge that nardilysin is capable to excise epitopes can be used for practical pur $" 2

(25) .

(26) be incorporated as selection screen in the prediction procedure of the reverse immunol" () # A $$" 2 $ ! & because of redundancy in N-terminal process :$ $" 2 $ are much more predictable than proteasomal cleavages, the actual cleavage can still occur either before or in the middle of a dibasic (& $ % $ $& nardilysin-dependent excision of a predicted epitope should be tested experimentally by in vitro digestion of long epitope-encompassing ( 0 / whether nardilysin-dependent processing is ! #$" / # () ( * $" * & >?:@A QR >< <R< & / / $" >< was found to contain 14 class I binding peptides epitopes with a di- or tribasic motif at 2 & >?:@A $" 2 $ 8 $ * : !& " 6 " / $ 8 $$ / / 2 ) % >?:RQ2R' () =& nardilysin-dependent CTL epitopes may also be over-expressed in cells that are treated with / >2 / insensitive targeting of CTL epitopes could be used as adjuvant immunotherapeutic strategy in relapsed/refractory multiple myeloma, which is currently treated with the proteasome / D / 4;5 Another promising practical relevance of nardilysin-dependent processing is envisaged to be the insertion of dibasic motifs between epitopes in multi-epitope ‘string-of-bead’ vac 0 4<&E5 / ( *$ 6 $ '

(27) . of the vaccine-epitopes, thereby promoting " 6" 8ingly, proof of concept of this approach can be $ 42 5 8 ies, the insertion of a dibasic motif was shown to strongly enhance presentation of both CD4+

(28) '+ ( $$ 8 basic motif reversed a cryptic CD4+ T cell epitope from mouse lysozyme M into an immu ! 4=5 In this study, the motif was inserted N-termi$$" $" ] 7& " antigenic determinant may also be attributable to dibasic-motif-induced pre-processing – by nardilysin possibly, but involvement of socalled proprotein convertases in the secretory *" 4'5 / !$ 3 2 processing may facilitate the accessibility to 2! | $Y $" $ A similarly enhanced presentation of a CD4+ T cell epitope that was induced after C-terminal insertion of a dibasic motif has been observed !$" $ $"" 45 Importantly, insertion of the dibasic motif directly C-terminally of a model CTL epitope 7892 [ $" production and presentation in living cells 4 5 > $" qqPq 6 $" 2 2! =& N E:& st : (^>2 2 $ ! $ < 2 4R&=Q5& / $" " 02 # " 4<R&EQ5& # (^>2$ /$ () 2 " /" | Y2 & $/$ ( $ " $ (^>2 N $ & $ $" (^> 2-.

(29) duction adds again another level of diversity $! " $ 8 3 " & $" $ % 0$$" / 2 preferences of the highly polymorphic class I $$& # $ " " 8 789& * /] " high mutation rate, antigen processing escape 0$" * D& importantly, this does not result in accumu$ 789 4=5 The total number of predicted CTL epitopes 7892 $ D $ $" $ <Q " ( /" $$ $ # /" 4=;5& /& $"& $ 4=5 /" * ==S escape are released from immune pressure due to the highly diverse and polymorphic process . $ $( and T cell-based immunotherapy > $& / $& 2 $ 7): $ I presented peptides that can be used as target $$" () 8sight in antigen processing is needed to identify and choose appropriate target epitopes $ !$ /& magnitude, immunodominance and immunohierarchy) of the T cell response and escape /" The procedure to identify CTL epitopes >?:@A D?2:D)& / ; & / $$" similar fashion for the discovery of epitopes $$$ ^/ $"& ( cell-based immunotherapy for cancer relies on. the same basic rules as T cell therapy for intra$$$ & * . ! *$$. Taking advantage of the full potential of the dormant anti-tumor T cell immunity of the patient will in principle greatly enhance the $ $ 6" " studies indicate that the natural anti-tumor T cell repertoire is directed towards multiple () . 4=<2=R5 ( !$ plete potential of anti-tumor T cells can be achieved by adoptive transfer of ex vivo ex #$ $"" (8) * $ $$ $"& 73>& ?: 7*& / erate high numbers of autologous TILs, and often the failure to obtain tumor samples, severely hampers the application of these forms 2# $ " (& *$* . $" (::2# 2# form of immunotherapy – either by adoptive >D) ! # (? /" * () ( $$* munotherapy to be standardized, less labori /$ ( # 0 $ & by several criteria, of the TAA and the T cell (:: : !$ " # ( cell-based therapy for chronic myeloid leuke @) ( > $$ $ 2 induced BCR-ABL fusion protein has a direct $ $% 8 * believed to be an advantageous target antigen because antigen escape variants that may

(30) R.

(31) arise under immune pressure would not be 7*& $ & $" * 2 /% 2 encompassing peptides in the variant BCRABL fusion regions are both binding to a class I molecule and C-terminally generated by the ( $ been conducted with non-processed class I-binding BCR-ABL fusion-peptides seem to / * " 4 Q2 ;5 [ our results, it is no surprise that natural T cell immunity against CML, which interestingly / $ 2 4 <& E5 $ /$*& * 4 <5 $" 4 E5 to be directed towards BCR-ABL breakpoint : /$ in CML, and other leukemias, will likely be (:: >?:@A :$ >?:@A $" ! @) 4 5& leukemias, this protein contains a multitude of () ;& /$ $ N& $ " # / 4 =&. 5& $" $ >?:@A !sion may decrease the rate of distant metas $ 4 '2';5 @ $$"& >?:@A * / instrumental in cancer progression through its binding to retinoic-acid receptor thereby in / 2 2 . & * & 4 R5 N 6cious immunotherapy, apart from being presented on the cell surface, CTL epitopes should / 6 $" (& * tested ex vivo T cell responses towards the four # 7):2:;2 >?:@A ; /$ $" < A $$" >?:@AQQ2Q' was demonstrated to be & *

(32) '+ T cells reactive to the three other epitopes were only found $* 0 ( $ 160 | Chapter 7 – Discussion. complete accordance with similar studies con /" 4'<2'=5 2>?:@A / %* * >?:@AQQ2Q' >?:@A<QQ2<QR 4'5 >?:@A<QQ2<QR >?:@AE;2 4'=5 ( $ 1 # 1 ( $$ " #$ " . : / tude of the CTL response, relevant factors to ( $$ " . / " / q @) !$& $" >?:@A * $$ / $ $ * /$ $ $ # ( $$ " ( $ $ (::& @) >?:@A& W( < 4 <5& ( *$ !$ the full potential of the multi-epitopic tumor #

(33) '+ T cell responses that occur natu$$" @) 4 <& E5 Only a few cancer types express TAA that are both directly linked to carcinogenesis and suf# $" $$* $ immunotherapeutic-targeting of only a single (:: 9 $ $ % 7>9 A= A " !$ gens, as is supported by clinical regressions of lesions from vulvar intraepithelial neoplasia 98 /" 4' 5. Therapeutic vaccines for infectious cancer vaccines Major health problems worldwide are the pandemic infectious diseases originating from # " 789& D 7D9& 79& "/ /$ M. tuberculosis), $ $ N & 7891:8

(34) 3 /$ (D.

(35) annually cause approximately three million and * $$ & $"& ;QQ< two billion individuals were infected with M. tuberculosis 4''5 *" ;QQ $$ $$" * 7D9 4<5 7 . $" * development of not only prophylactic but also therapeutic vaccines for these life-threatening 4;&<5 9 " / ./$ " D" & * / $ ing the pathogen, thereby selecting for resis& ; ! < ./$ patients – could become containable and hope$$" /$& ( $$ "& $$"

(36) '+ T cell responses, have been shown to be essential in natural immunity partly controlling and lim 7891:8

(37) 3 4'R&RQ5& 7D9 4R5 79 4R;5 4R5& (D 4R<5& $ 4RE5 Thus, unlike the protective vaccines that most$" $" $$" 6 $" humoral immunity, therapeutic vaccination & 0 . () % $$ $$ N$ . /* cines and vaccines against pathogens lie in the characteristics of the antigens that need to / ( ! a relatively limited number of shared tumor (:: $ $" /$& / $*" $$" 2 # and they may be lost upon immune pressure & & 0 are expressed in the tumor of only one patient & } ( of pathogens are the source of a wealth of potential antigens, all harbouring numer D2 ( $$ N & EQQQ proteins were predicted to be encoded by the. M. tuberculosis 4R5 8$"& vaccines for viral, bacterial and parasitic infections need to deal with the sometimes high / $ "

(38) . $ & * # $ $& and antigenic drift both contribute to vari/ $ " 4R=5 789& & /" lacking proof-editing of the genetic code, can $$" . " 4R 5 ( $$*& a strategy that is called ‘reverse vaccinology’, the search for conserved antigens with limited antigenic variability in the full genomes of dif & $$* $ 4R'5 Obviously, these conserved antigens, and spe #$$" / () do not lead to escape under pressure of natural immunity, are the favourable targets to include 789 . Conclusion and prospects for T cell immmunotherapy ( / # () epitopes has the advantage that it allows precise assessment of pre-existent immunity and accurate monitoring of therapy-induced im " N& * $$ $ 6 $" guide the CTL responses towards multiple antigenic determinants, including immuno / ( will result in a broader exploitation of the full T cell potential, thereby hampering immune loss of antigens, mutations or class I down $ T cell inducing vaccines for both cancer and infectious pathogens have to meet the same tremendous challenges of overcoming immune evasion strategies and mounting T cell responses that are superior to the – apparently 6 $ " 4RR5 ( $ $& $ % Chapter 7 – Discussion | 161.

(39) avoidance of tolerance induction through non $ & ; 6 $ /"

(40) E+ Th cells) and $ & < * $tion of negative regulatory T cell responses, & $/ & E propriate multipotent target choice, will help 6" ! & adoptive T cell transfer is an important im / some viral infections) that will also strongly /# () # Important developments in cancer treatment can be expected and are needed in the application of combined conventional therapy and immunotherapy, for instance by exploiting the /# $ . 2 " radiotherapy and some chemotherapies that 1 " ? evidence indicates that the host immune system also contributes to therapeutic outcome of conventional chemo- and radiotherapy / N & 2 cyclines, which have been used to treat a broad range of cancers, can boost the host’s immune " 6" " 4QQ5 D responses can be induced by dying tumor cells 4Q&Q;5 ( 2 then help to eliminate residual cancer cells and can potentially maintain micrometastases in a " 4Q<5 $"& $lar and cellular bases of the immunogenicity of cell death that is induced by cytotoxic agents / $" $$ W / learn under which circumstances cellular de ( nogenicity of dying tumor cells is a function of $$ $ "\ & or necrotic cells may be either immunogenic, immunologically silent or tolerogenic for the " 4QE5 8 "& 162 | Chapter 7 – Discussion. for instance, has been demonstrated to force$$" $ 882 " 4Q5 2 4Q;5 7& of autophagy in cancer cells may be exploited for the induction of anti-cancer CD4+ Th cells 4Q=5 ? $" "2 @7 $ 88 sentation mediates resistance to pathogens and is targeted for immune evasion by viruses and / 4Q &Q'5 8 "& " tumor cells has also been shown to be indispensable for the cross presentation of tumor /"

(41) 4Q;5 Thus, an optimal treatment of cancer has to $ $ . .

(42) References ; . < . E . . = . . ' . R . Q . . ; . [$ D&

(43) N )\ ( $$ D $ ; \;R2<R& ;QQR 3$ @& 7 $$ @?\ ( \ $ " * > $ : 3 3 : Q 3$ ;\ER& ;QQE : D& $ [& / D&

(44) / >\ ( 3 <Q\;Q2;Q'& ;QQE P*$$ XW& D % X?\ 8 major histocompatibility complex class I-restricted ( $"" : ? 8$ \2 ''& RRR ( 3& > D& D$ % 3& 3 ^& )$ & 3 @@& q$$ >@& ? 7[& 3 $ 7& 7$ 7[\ @$ @7 $ 8 pathway by combining predictions of proteasomal $& (:> @7 $ 8 / $$ @$ ) 3 =;\Q;2Q< & ;QQ

(45) >& q$/ ^\ 8 $ ] @7 $ 8 *" > 3 E\;<;2;EQ& ;QQ ) @9& ) & )/ q& D 3& D% 3& ) ^& $ @\ : () \ / $ @7 $ 8 / & (:> 6 "& $ $ A X 8$ <\;;R2;<Q<& ;QQ & W >& q P& 72: >& D X& D >A& D 77& D 3& N% $ 3& [/ X& ) ^& ) & $ @& >% X& 3 :& & > D\ 8 / $" 8A

(46) D2:? $ : ? <=\W<2W'& ;QQ'

(47) " 8:& [ >& N$*

(48) ?\ A X\ $ ( $$ D@ D \<2E& ;QQ= q$ X7& D% X& D29$ 3:& 9 ]% >& 9$ >:& q$2X :@& 9

(49) & D [X& q @[& 3 ] :&

(50) ] XW& ^ N& ^. ?& @$ X@\ A6 # $ 7):2:Q;Q2 presented cytotoxic T lymphocyte epitopes in * $" ! >?:@A /" 2 $" X A! @ R<\ <2''& ;QQ P*$$ XW& ? A& X\ @% \ 0 @7 $ 8 ? 8$ <\R;2R=& ;QQ< q$ X7& D29$ 3:& 9$ >:& ? :& 7 ]/ 8X& @& @$ :& ^. ?&

(51) ] XW& )% ^& ^ N& @$ X\ D?2:D) $ $ $% 2 #

(52) ' (2$$ )% ;Q\ <'2 Q& ;QQ=. < 7 " P& [ 2> N& q ?& ) 9& [ N& ) N:& [& A >@\ 8 # $$" 7):2:;2 -restricted proinsulin epitopes by reverse immu$"

(53) / E\;Q<2;QR& ;QQ E 3$ X)& @ $% 3& 3 @X&

(54) . [& X q&

(55) $" X&

(56) $ X& [ // X[& $ )@\

(57) EQ2 D $$\ $ $" 6 $$ $ 2 # ( $$ " X $ 8 QQ\; 2 ; =& RR :""/ @& 3 3& 3 & [ $$ >& 3 & ? $

(58) & 9$

(59) & ? >& X& ? 7[& > @& 3

(60) & )" N\ >2 # 332;2 /" 2 () #$ $ X 8$ ='\ 2 ;;& ;QQ; = D/% :& ) N^& 3 (& A $2 @ @& N $ @& [ :& 3" W& W >& W$ >\ 8 # $ melanoma-associated T cell epitopes for cancer " X 8$ E\= =2= ;E& ;QQ ? 7& D X& A >& D ^:& 3 3\ 3PN>A8(78\ / @7 $ 8 Q\;<2;R& RRR ' >/2q @& 3/ 9& q $ ?& [$ D& )" & D N& " X>& W" D& D (& 9 A" DX\ : $ open reading frame of the human macrophage colony-stimulating factor gene is independently translated and codes for an antigenic peptide of 14 /" 2 #$

(61) ' ( $"" X A! @ R<\'R2R'& ;QQ R (" NA& D* 3?& D%$ :@& $ 3& D :& @ $ XX& D (& W % X& W $ @& 3 $ 3)& D* X@& q]2 $ )& q% )& >$$ :W& @$%" X& ?] X\ ( $$ receptor recognition of a ‘super-bulged’ major / $ " $! $ 82/ 8$ =\E2;;& ;QQ ;Q 3 P& )

(62) & [ $ 3& 3 )& A >@&

(63) $ 9$ @\ : $ 2 $$" ! / @7 $ 8 $ $ 789 $ X D $ ;'\=<'2=<=& ;QQ= ; )%" X& @ X:& > @& 7$$ A& W N@& ) $ X

(64) & 3/* X& 7

(65) N& A$ 97\

(66) . ! $ 8 @7 $$$ / X 8$ = \;;2;;& ;QQ.

(67) =<.

(68) ;; 3 & @ & 3$ :&

(69) XN& W$% )& X% q& 7%$ >&

(70) 3& : @& $ 37& W % (& 3 $ 7& )

(71) & [$" :& 7 [& q$$ >@& ? P& 7$ :\ : $ $ "$ @7 $ 8 8$ E\< 2< R& ;QQ< ;< 3% X& 7 % ?& [$ >7& D 9& 9 >$ :& P& 3/* X& W$$ (& 3$ $. )& X& D (& 7

(72) N& A$ 97\ : 7):2:;2 " $ $$ $ $ $ #tion and suggests a novel pathway for processing of / X A! @ '<\; 2<E& RR= ;E A$ 97& :$ 29$ @& ^% @& $ :)\ >2$ $ # $$" @72/ ^ 8$ '\R;2R & ;QQ= ; 9 & 3/ 9& X& ^ :&

(73) [& @$ 3& D>& D (& 9 A" DX\ : /" $ 3 <QE\' 2 RQ& ;QQE ;= 7 q& P*$$ XW& P X\ 8 nition of a human renal cancer antigen through 2$ $ $ E; \;;2;=& ;QQE ; D :@& [ @& X\ :$$$ ferences in the relationship between proteasome " @7 $ 8 $ X 8$ =\'<2'R& RR' ;' )%" X& q [@& @ X:& 9%* 3& @ DN& 9)& $$$ (:& 3/* X& 7

(74) N& A$ 97\ > " @7 $ 8 \ for nonproteasomal epitope generation in the "$ X 8$ =\;2;& RR' ;R 3* q& [ $ ?& 3 % [& q% 3&

(75) D@& q qD& * @& q ?& [ @\ ( $ / $" and epoxomicin can be used to either up- or down$ ! X 8$ =E\=E 2= & ;QQQ <Q @ $$ @& $ XX& ) X:\ >2 7):2D; $ $" $$ / @$ $$ > =\R;<& ;QQ < * q@& A& X$ @:& 3 >"% X& X$ )& 7 )D\ 3 / # " $" 2 / X D $ ; \RE2R& ;QQQ <; @ $$ @& 9 $$ $ A@& X:\ ( dyl peptidase II is dispensable for the generation of both proteasome-dependent and proteasome-. 164 | Chapter 7 – Discussion. << . <E . < . <= . < . <' . <R . EQ . E . E; . E< . $ 7):2D; $ 8 $$ A X 8$ <'\=<2=<R& ;QQ' 3* >N& 3$ A& 7 ?^& den Nieuwendijk AM, Leeuwenburgh MA, van der @$ [:& q$ D@& ^%$ 73& q $ :N\ : $$2/$ / "2/ probe for the caspase-like activity of the protea D @ ) \<EQ;2<EQ& ;QQ [ A& > [& W $ @& ) 27 @& D W& A q& [\ : ant protease with potential to substitute for some 3 ;'<\R '2R'& RRR [ $ 3& ? 2 @& : $ N& 9 $$ $ A@& : )&

(76) $ 9$ @\ ( "$2 88 @] 7 / $ " $! $ 8 9 $ : > * >

(77) $ X D $ ;'\<RR;2 <RR<E& ;QQ= W" AX& [$ (& @ 3A& 3 /" [& @A$ @X& 3%"

(78) & A $ )\ Re-evaluating the generation of a “proteasome @7 $ 82

(79) ' ( $$ X 8$ =\;;ER2;;=& ;QQ= P% 8:& D & 3& [$/ :)& ?% q)\ ( "$ > 88 8 @] > ( ) : >& / 8 ?0 @ @7 $ 8 : > X 8$ . \E<E2EE<& ;QQ= D$ @& [ @\ $ $ "$ 88 )@92 ( $$ A X 8$ < \'R=2RQE& ;QQ ^$ 9& @& @$ ?)& N A3& :& X$ @:& X$ )\ 3/ # " characterization of recombinant metallo oligo $ $" D " EQ\EE 2EE;& ;QQ 3 X:& >* (7& (/$ :9& X

(80) & [$% @X& W$ :X\ N$! / $ " / recognition by thimet oligopeptidase as revealed by D X <''\;2;=& ;QQ 3 (& D X& [ 8& :% (& ?% q)& [$/ :)\ @] / $ " $! class I-presented antigenic peptides are degraded in cytosolic extracts primarily by thimet oligopep X D $ ; =\<=E E2<=E'& ;QQ q 38& >/ :& 3* (:& [$% @X\ Regulation of cell-surface major histocompatibility complex class I expression by the endopeptidase A<E;E $ D X < \2;Q& ;QQ< P% 8:& @ :& ) q& W& 3 P& :/ ?& 3 (& [$/ :)& ?% q)\ ( cytosolic endopeptidase, thimet oligopeptidase,.

(81) destroys antigenic peptides and limits the extent @7 $ 8 8 " '\E;R2EEQ& ;QQ< EE ? A& ] X& 7/ & D% ] W& X )&

(82) ] XW& X\ : ] $ (>>88 $ @7 $ 8 8 " ;Q\ER2Q=& ;QQE E (% D& ) :\ ( "$2 88\ $ 2 8 X D $$ D $ < \R<<2R< & ;QQ E= A >\ ?$ "$ 88 @7 $ 8 ` A X 8$ <'\=QR2=<& ;QQ' E D* q& @ q& ) W& D% @?& ($/ W

(83) & ?

(84) W\ 3 $" $ $ $ / > $ : 3 3 : R'\<; 2<<;& ;QQ E' 3 P& X % :& ? @?& ( WX\ 3tures of human insulin-degrading enzyme reveal * / EE<\' Q2' E& ;QQ= ER D ?[& 7$ N[&

(85) %* W\ 8 # $ "$ $" complex containing insulin degrading enzyme and $ $" D D " ? ;Q;\QE 2Q<& RRE Q

(86) %* W& D ?[& 7$ N[\ : / " . $ "$ teolytic complex containing insulin-degrading " $ $" X D $ ;=R\;E 2;E'Q& RRE 7$ N[& D ?[&

(87) %* W\ ?$tion of multicatalytic enzyme activity by insulin $ 2 " A $" <R\EQ=2EQ==& RR' ; ^$*% ?& q

(88) X\ > / "\ $ # $ ? E\=ER2= & ;QQ' < ( ?A& 7/ ?& 9 A8& ? @A& A/ :X& @$ X& ^. ?\ >tive anti-tumor immunity induced by vaccination with recombinant adenoviruses encoding multiple tumor-associated cytotoxic T lymphocyte epitopes 22/ > $ : 3 3 : RE\E==Q2E==& RR E ) D

(89) & * @& & @q "

(90) & ?& 3 :\ ^ cessing enhances immunogenicity of multiepitope

(91) : 9 R\E=;2E==Q& ;QQ 3 3& ^ X& N % )& [$ D& [ X& : :& 3 AA&

(92) 7\ \ $ % . egg lysozyme results in enhanced T cell determi $" X 8$ =\;Q2;<& ;QQQ = 7& ) q& " X& 8 (& @ $ q

(93) \ 8 / % of a cryptic self-determinant leads to activation of 2 # ( $$ X 8$ \;;;2 ;;=Q& ;QQ ) [$$ 3& 3 >& W$% D

(94) \ >/$ % 0 $ () X $ 8 \<=<2< & ;QQ ' 3

(95) N\ ( ^ D $ ;\<2<R& RR' R q [&

(96) >@& D :X\ ( $ # "\ $ cleavage near the C-terminus and product inhibi % $" "$" D X <Q' > \E2Q& RR =Q ^$ 9& [ ?& ? $ 9& N A3& 3 :& :& X$ @:& X$ )\ ( $ . combinant metallooligopeptidases neurolysin and $ A X D ;=R\E<;=2 E<<E& ;QQ; = 3 D9& q & D ?X\ ( # " $" @7 $ 8 $/$ 7892 (:> >)3 ^A <\<;& ;QQ' =; 3 3@\ 789 () \ * ` ? $" \'& ;QQE =< P % [9& @$$

(97) W& & ^ (& ( )& >$" X& [$ 7& :0 $ W&

(98)

(99) & ? W& > XW& A$ 97& N 3& 3$ $. )& X\ 30 $ cape and shifting of T cell responses in a long-term $ X 8$ E\='=<2=' & ;QQ =E & X

(100) P& ( :@& 7 & 7 )& @ X& @ W& ?// >N& >$$

(101) @& X. A@\

(102)

(103) ' (2$$ $ cell carcinoma antigens in patients treated with an autologous granulocyte-macrophage colonystimulating factor gene-transduced renal tumor $$ ? =\Q R2Q''& ;QQ = q 7(& W X& ?/ 3:\ 8 # $ $ /" 2 #$ $"" $ \ /" $ $ @7 ! X 8 ; \'E2RQ& ;QQE == ) 9& N @& [ $ & N" ?:& ) :& N$

(104) & W$$ & 7/ & W$$ (\ ( of autologous T cells to a human melanoma is /" > $ : 3 3 : Q;\=Q<2=Q'& ;QQ = > [&

(105) N $ :& $$ )& $$ \ 0 \ / $".

(106) =.

(107) $ $ $ X 8$ '\R 2R R& ;QQ =' 3$ $. )& X& $$$ (:& ( )& [

(108)

(109) & 3% X& $*$$ X& D %. )& q $

(110) X&

(111)

(112) 7& ^ & 7 )& 7% A)& 7

(113) N&

(114) * ()& A$ 97\ @$mas with concordant loss of multiple melanocytic . \ " / /" 0 # 8$ 8 E'\==2= ;& ;QQQ =R q 7(& ?/ 3:\ >2! " " 2$ (?>2;& * (?>2; & P2A3^2 $ a patient with a dramatic response to immuno" X 8$ ='\R2R=& ;QQ;. Q > $$28/ X& q& q (& 3 3& D @& X& ) )& X X& q$ X& 3 /

(115) :\ 9 * chronic myelogenous leukemia with bcr-abl onco /% # D$ R\ '2 ' & ;QQQ. q& > $$28/ X& q (& 3* X& %$ 9& >$ AD& 3 /

(116) :\ A multivalent bcr-abl fusion peptide vaccination $ * "$ $% D$ Q<\Q< & ;QQ<. ; D @& [ $ 3& :/ A& N$$ :& 8$ N& (/ $ :& :/ $ @& N N& [ :& ?

(117) & : 3& ) N\ A. of a p210 multipeptide vaccine associated with imatinib or interferon in patients with chronic "$ $% $ \ $ / $ $ ) <=\= 2 ==;& ;QQ. < [ @& :/$ @& @ $$ :3&

(118) $$ 3& )/ @& [ 3& q ?& [& 3 )& [ $ N& )& 9 D& D" :& $$2 3\ A! $ 2 #

(119) ' ( $$ * "$ $% \ $ vaccine development and adoptive cellular immu" X 8$ E\';Q2';'& ;QQ. E [/ N& @ %] 9& @ %] 9& @$$ @?& D (& D >\ D?2:D) immunodominant antigen in chronic myelog $% ? ==\'R;2RQQ& ;QQ=. >" 3& ( q& P 3& 3"$ [\ >?:@A ?: $$ * $% \ $ $ * $ )% ? <\<=2<=R& ;QQ . = (] & [$ X)& ) @& 9 3 @& (/$ D& [ $$" >\ (2: : > $$" A! : @$ >?:@A 8 28 $$. 166 | Chapter 7 – Discussion.

(120) 8 ? ( " 8 ? =\ <E'2 <& ;QQ. 3 /

(121) & >. & W 3& [ D\ >?:@A ! * *2 regulation of retinoic acid signaling in primary "$ $% [ " . \2E& ;QQ . ' | 9 )X&

(122) 7& 9 ] @X& 7 P

(123) & 7 ::& @ @& > 7)& qq& @ @X& W :(& 3 / [X& q% ?@& ?/ & ) $" >3& D ?& N 37\ [ ! #$ $ $ / E\<Q2<=& ;QQ;. R A @(& W )& A$ @X& $ )& 7 @& D ?\ ( >?:@A is a dominant repressor of retinoic acid receptor $ $$ ;;\'<2'E & ;QQ 'Q A @(& D ?\ : $ $ man tumor antigen preferentially expressed anti $ ? ==\Q=<R2 Q=E;& ;QQ= ' A @(& 7 ::& [$ :@& q ] ^& D ?\ >?:@A ! $ $ / D X RR\<R'2EQ<& ;QQ' '; >" 3\ 8 " %* $$ / >?:@A` )% ? <;\<=2< & ;QQ' '< [ X& 3 @& ) )& [ $ q& D q& 3 :&

(124) q& 3$% ?N& >$$% X?&

(125) 7& D$$ )\ A! 2 "$ $% \ $ # D$ Q'\EQR2E & ;QQ= 'E ) )& [ $ q& ? >& (/% * X& 3 :& [ X& ?$ % X& 7 8&

(126) "% :& W @& 3 @\ 8" for patients with acute myeloid leukemia using autologous dendritic cells generated from leuke /$ 8 X ^$ ;'\'2'=& ;QQ= ' $$ &

(127) [&

(128) :$ D& 7" 9& @ @& ) )& 7$ 7A& ?" @& > N& 3$ D\ "! ( $"" the preferentially expressed antigen of melanoma >?:@A "$ $% D$ ;\' =2''& ;QQ' '= ? q& P :3& (*/ :& X D& A ?& @ $% 3& 3 D& q" q& ) P& q$ ?& D :X\ A!2 $"$$ "

(129) ' (2$$ >?:@A2 # * lymphoblastic leukemia and acute and chronic "$ $% D$ <\;;E2;;& ;QQ' ' q [[& W$ @X>& 9$ ] :?& )* % @X& D2 @

(130) @& 9$ :>& A N& N )@& ^. ?&

(131) ] XW& W$ :?& ^ X& N$ [X& D 37&.

(132) @$ X@\ 9 7 > $$ = 9$ 8 $ $ $ A$ X @& 8 >\ ;QQR '' q 37& @ $ :X\ :$$ $ \ :8

(133) 3 /$ @ \3<<23EE& ;QQ 'R q ?:& 3 X(& P& :* :& @) [& D%*%" W& N & 7

(134)

(135) \ ($ association of cellular immune responses with the initial control of viremia in primary human # " " " X 9 $ ='\E=Q2E=& RRE RQ D X@& ) @& 3 9\ ( $

(136) ' (2$$ 789 :8

(137) 3 ? =\ R2 ''& ;QQE R W/ [X& ? 3& D*

(138) & ^ [3& X )& 3 3)& W $$ ?&

(139) % [& D$ :\ ) $ $"

(140) ' ( $$ # for structural and nonstructural hepatitis B virus * D\ $ " X 9 $ '\ Q 2 R& ;QQE R; q$ >& 3 \ $$$ \ / [ \RE2R=& ;QQ= R< )$ :& D% ?& W $% ?X& @$ :3& > ::& : >&

(141) %$$ 7& >$ [& 7 $$ :9\ 7 "$" 2

(142) ' ( $"" # @"/ /$ > $ : 3 3 : R\; Q2; & RR' RE @ :& $ N\ A. "

(143) ' ( $$ " $ 8$ ? ;Q\;R2<Q<& ;QQE R $ 3(& D ?& >% $$ X& [ (& & 7

(144) & [ 39& A $ q& [ 3& D" A& 888& (% N& D% q& D

(145) & D*

(146) & $$ * (& ?&

(147) ?&

(148) $ q& N$*$$ (& [$ 3& 7$ & 7$" 3& 7/" (& X$ q& q :& @) X& @$ 3& @" )& ^$ q& ^/ X& $ @:& ?] @:& ? X& ? 3& 3 q& 3%$ X& 30 ?& 30 3& 3$ XA& ("$ q& W 3& D$$ D[\

(149) / $" @"/terium tuberculosis from the complete genome 0 <R<\< 2EE& RR' R= ?$ ?\ D %*$ D $ ;\<=2<==& ;QQ R @ $ :X\ 789 : ? 8$ ;E\;; 2;& ;QQ= R' 3

(150) & ?$ ?\ >2 $ NAD3 ) 'Q\;R'2;RR;& ;QQ= RR

(151) ?\ > :8

No results found