t“ .- J'3'dr-Jt . ° _ o . ‘ 4mxm. 4-1- . .;-'<"v.-:7.7-':-":~. '77 :‘ 7: - . 77717.? v 1’ -'7'."_\;«. 4172;: “. . I. ‘- "'!‘.lr;"('f'~'-'o'- ."""""." L'.‘jnfi’$"“. ' 'v”‘“!‘ .3: ,% .I' U ’ '( '0 - , _ _- _ ». - J , .{!:.-. 2"".‘"-' . .I '1’. " f ‘4'.-. P: y '- J.....' "Q-“’-I..-,“ . x I. i. .t‘» I. ‘ 7 ' . to ".9 ‘ ' H s ' 7' 1 -- . .. .. ----.-7 1"30. 7M“:- ."u -. 7. .. 7’.“ n~’ ‘ .’ ".> 7*: .‘f" ‘a‘ " ”fay“? ."'“-' J' ’ " ;§' 17" 3‘32); 143;} :?'{.~'.":': '0: '£,'.." 5-‘3'Tf‘fp‘757 A ‘— . “' " 3 a - I "I : . ° .33.- “.’ I ' '- - f{.rr :’ .:’.;t' IJJj‘2} y:’ '..! i ". "€".‘\' .A::_.f:'i1’,‘;fl.' {F Q. I.“ . . j . #. c. - .. .M 5 01.)}:- ’... -' .- . )- ..:‘l "10 i.’ 1,"; A’C‘.’§, r r '0‘ ' ’ov'bl' ”-7 !°----"-7"’:-.‘ {7" - Werr’-.‘r.‘. 2" ' ’ "". In. ’t. . " .0? at "g ‘ ... ... I. ‘ c..“."~’o. a" . h .7: 1, . "#5" ?;."..'o P? ’..")? '1 I”. {mi/1:: 1.2:! . . J“) ‘1‘ ‘ ol'v.'J"’." ' ’f':::fl :4! gage! ;:’.‘0 . 9 y .Iv ...7.’ .‘ ~ -‘ ’l.;.l. I'f . FIl ‘.. '\.‘ :1 .7... ‘ A" .- ' 2.345. Afixa‘ovfflfrs ‘4’)_. .FFI}‘:?. Cap. 7.“ :53; If". I .tz':.‘b'\vfisf'. ,‘ (fi.|.)flr" I. ‘. x I, .'.i- “90"? ... -.7.”II9'V.~. '.:I. ' .'- ‘v}‘.'_ 0 .0 . -. . ' . ‘§’I'; .‘ $ .." . I.” ’9' J ‘ ‘ I v'o , ,‘g. -1 4 ’I“...r ‘ ,L.‘ 0" 51"“, I .‘ ”I «:1? I:5:I.’ {r}. ’ fr .|_ .f'U “flap-0:} . . Ff; :‘..«‘-.-..‘{ #:7‘5' -.. a ‘:?; :3." '3‘ . ‘0‘: f . . I 'f’ _ I ' ‘7'. 'Eru. ’..:f. ,1. zing. 2.1,“. '.. -’an."‘.. ! J 7’ ' w ”-51:23 ' I Q N ‘1’. .::'-.i.. 5 .. " ° ‘ - 0* . 1’: .. '“r’io‘: 'J' :‘4‘1' g. :6}?! ‘;,$ ‘-..AN‘I 'Zk’ffi‘ ‘3 ’f ‘ ’1‘ f‘e " o 1 ' 3703}: f/W 4"..."um‘3y': . .. 7 ... -.. . . ‘ v10 ~.'.'I r o“ “-..k a ""‘O.- _.. .. .__ . If! .. fig" fur—'33“ 7 q . . ‘zkh 5%:k. :5 RBI); §gi¥ ‘ O’;::;¥'J ‘0 :2‘1‘0‘ “H J..." {'56:}.1} {'2’} A ‘- ‘4‘ ‘ ’1“an “"4 0 .. ‘ ‘ .c_ . -1 - ' ... ’ § ‘ ‘ ‘ 7 7 k 9‘)»,- ." ”pig. 1' ‘ "£4." ...3.:._..:;,*€:.. .v- a I." '7" 7‘- sis? '. '7’ Ah“ ’i‘lwa'flmfigfi f’wfiugr'flhz. -,: "”".'.7.?'.'/.~31-~t~§‘;v" V "4.13“ $7 , Jami-«7.4 5'Q10'3'Q7 -'~.s'l€.~"7m~l 7.; 47 1:: Q. ‘3 Q :17} L :‘E i ? ”‘RA'Q‘F...’ ”1"". $5.5": I”:?'?M ;. J." '_ I ‘ _ . . 7 - ”Eff-P Q.J-}..":-J\'l or{‘o.::.,’..-k;:rvir‘" ,jotx’? 310,0: ; '3‘: .:C 5‘ "Jv'v.q‘ ..W I . '0 'f’ ’}’F‘1’ "‘- "J. " .' ' '..r. . . ‘ , (, ... f .0 . ‘ ‘ I? , ..7 ”:77"! 7:>n-fififypwa ,‘Z' ‘ . ' ' . 2;? {; "On'oéfléfivg. 4 '5‘ 'iL-i'fl'fi g. I 0'.\"‘,. " 3?...'r: "”.:'Aq _‘ . \ ,‘.. ‘0 . 5‘ ....K. 4:: “"2, . #77777: .77» a... a- 7'3 .7 .73-717‘7-2‘ 2:; ‘74.: 7.0/3”va ..., ‘0 . {1% "t‘ .f.[” “. ’..;I‘JI.’ .fi)‘; ""6“ '..... | .q'."‘( d '2.."‘!€.‘?..lfl. I‘VQ. ;* {_,J _u _. 9" nf“{‘f~ ..‘J’ ,‘Jr'... ‘+ "£7", “w: ‘:>'L(“.o.‘__ art; may”. :77 rv’.-.~" "if-n- f-".’(. 7 '...r; 2, ”'va :72“? 73' a Q A" - o ’ g 7 . .fi:.: ’1 IS! Vfik‘t‘ ‘ ~VJ' . "H ”'07- i" if 5 $Vl' ’1'}..'L."::‘.'0 \."'l.‘.‘( ’5 .‘ $"" .1..." ' '... ’7 .J . v“ - . :jgfikf‘b". ’ ‘1'H\. .:." 1;: ' " '. d ; o. "‘ ‘ {a .71"; 'k'JW .-zr‘lgdjry -";‘:. é:“l'{ .' "rffiJ;('fl'lffl0€a::;v' .. ,I_ " flaw» ’r ‘- O .3 . O r g .87 gs...) ~ I“ ,‘ , .:: (fl < I'v. I § :3: :7» 1‘7 37:71“ «”4 \‘I ‘$ h '3‘ "5"“ g.“ 6:532 .‘:"V . , . . :7 ' ‘00. ' ..1' 4"." ' ...' .».4- 3-3-4. 7543.77,; 229W; 7 ’71.» ’33 7m513‘7t'3}}"kf’ .1“ . 'ufir; 3n," 3W ‘34! 3W {I . 7772’; ”4:5; ’4": «2575777... «1/7"- : .... I. 'J.’ ' ‘ f" ‘;-"’;' . “3‘. o. ! '.. 'I" ‘I 5.7.7. an?! '1’! ...”! J” 5f? 3-”). ~I\l’_{:’.:v:o..£,:r ) ”,1?” 42px" - 'IIOJD. . _. . ' . . 'l ‘J “’2'; r..‘.-.’:."r0r' («'.. I r... ' 1:! 34’ {.l" g .',l :{H o Kai’s:- If '35:}; :2." "O~z‘r J r" ":1." '_'fl":£" I ";O,:;. :fi . Q‘et ‘... . f} l J ‘7- Image '.. .- u. f' '13}... ' .T g‘.‘,. :Ifh7...::.. "._;.".' 12““. .Jg...’ 'p‘l’.~r~7 F .I:"~l'.~x‘:.- : ., l.rr"\’}:’i’ " :’ .’.. ’010.J:':".;‘_‘:fl 'u .1,” 0‘ {."6'9' '1...:.‘"I.OJ(I.; {:1‘” £517.21} "“ 312:" ‘7!" '2' 7.771, I )4: fig: «’15. 0 J9," '.9 'ch' ...,” 0'01"; 0"“. .. , ‘ 2‘. I'. “So i"?! z. 1' V1": ’26!“ I"? ‘ - 0" ‘ I" 3, 0-" if " '.i‘": {'7 (r 'e.‘-‘ 3:2; ’7.I," ' ’t- J ’3‘ 7‘7"“ ' 3.47.7775). 7 ""7"-'“"}.."77 ’. 1 «'.*'-L'-"'-"'z‘ '70": :I?'{ J"? 7'7 - - g . IW’";’:|"; 1.1%}7—2'1'17; you-f 0.x. 0' I. I... '. _ . ‘1 Y!!! ‘- . ..., "r :vqu . r: '7‘. 'J.o .Gio.'..-.|| “-u r . ._ ‘. I, I, r «...? '1” . 7.7!. $717 .:;:-:r.-sr7”?5§--"§:5V $713.: 7 :7 t? {371-- .' 7 ‘7’ .557 {7‘ 7 :1! 5717-134”? .:¥.7}'1'I;’7g, 3? «2.517273%.gx-‘A. 2.71 am 976.3337172'7}; fifig 47;.”- v .".'o 7" J'.‘ v‘-.¥"".-o.:-'f." 0' ‘5" 7. "fi-"J -" ‘. 0'1” 7'; "' f N .111; .31- - ~' . :77 -. 7" '74:) . '. . 7'. J. :7- '-‘ It" - . J (I: "1.": A'f"'.‘ J'7'?‘::{v {RV/fig. L1,"? I?$p?}‘,‘- .r‘tlf'auv "1172 7r ‘ . u . 0 - I ' a O ‘ i 7'12! ' " -- 7;): ‘r.'.'77‘:‘-.’r€'T"o'o #4:??? ”£213.". tifq’fif‘ I“? ‘1'” l‘ ‘ .11.. Mn'."uwwamnmmwdm~k o ..L 7 - 3mm» 0 07am.) Inn-1H . warm l ..., .. - '0 n ' , _ "_..>:‘_"'J:.“J-P \ 1." ' ’1_.‘. .9'4. ~‘o-‘l "’.I .' ‘I‘N (.M f’ , lfi.!','{ 71""3-3.‘ ‘42:}:‘3..d~,.]' 43?: .flr‘!’ ..7 ‘..I.o,!_Sg_1'm 13%;}. ‘1’}. "‘WI.’ '03: -,4 H '. "L "if” I." I "‘5' 3127‘ '..! {11"}: “(Lg/Alta” 1 Iifiigfi'2. ° iii-iii: 37:12:." '.""‘}.>.’ Zing '1'... I. ... a. "1’. ....“ {0| ”I 4 .fl‘.‘ )vo' ' ‘ I .. 4 H i o ’ ‘ . . al‘ . r‘ .’.Y T’.' 5‘ 6’ ’..,t ‘f I .. .v1f'.:>;?"r'f6.'l1: '.;J:].v-"{:ro‘. 0{fifi241‘:!:';J{‘;'_€{I‘I‘J g" .f‘.;,:l"'w ‘ '5" .- ‘1'!” { SLf". ’ u M I. ’1: 2—1;.) 4:: ‘..". .z’nz‘i‘.’fi'f“' 1....‘."' +7 ’_ :‘ 1“. ... ’7 - ”1.... "‘;3 ?:"u' ', ‘a ."'V. I I. ' 2' N" O 'I “I; .‘ u ;‘ " J. V; .I- 3‘: . '1‘ ’. .7 ‘ .O’UI'I .3 "1...; :ff:?~r 4". '4'." ....t.'¢ ' f: .3! 'k' ';‘ I ' ‘ .'f 1"q/I¢ : ‘f r. o - .-n...’ ... -.' _. | . ‘..". .":‘. .... . IIQC ... .. I» ‘ r 0...... .. WU"? 3".- ,n’é. ~if?¢;r,o.a 1'". ’.p "’f‘rn” 1'4"”:‘3. 3‘: "'7": 717%”.27- AQQ'fAfZ' f"'§f"$!'°°:"’;?"‘gflp . .... 0. o '4'” ...-0’4 "rfiha ’3 7,: - .‘l of; .f. "'4' 0' U, Q . ' fig 1215-2737;: . 7 3.4-}; .7216“ f, r :1- ,. 21-333, - :3 ,vo- 3’3,” '2‘ 5:! ,r ,'- ,3; {1&7 - “W;.;;.;.,’p.7{:.v:; {'27}: 13:! 37.1, ' ‘9’ ‘I u 'L' ‘1“ n I'- . p01.-;..9:’7,'I7'p\-‘-‘4"f Wt” 'l'.‘ ". ‘5? 3" x'.‘ " I’ . $.53 0 ‘0. I o ‘ . alt)... 3:.0/J; (.-.... j},' . ‘.‘tf—l ’."{"'_!’x ’ "__"cf§v.' A I" ’11: I ('....3 ’ 'I. I \I 'll " (I if P“ f ... .. ' I a‘ ' 'fi I. 0".7' ’. ' " O .‘ . ‘ .; ”hf". :lY'I-gr‘ ‘pr Iii: r’f'fj‘ / ‘..;IIJ?’ ./(a:‘: 52:} :I'tz'o} fig}. ’.,'. Iy;’:Z£'fv’rf~/.,_ é. J’J. . ,pr xYE-I” :IIr'L.A.' "’té‘z'r? rfl‘ I . a a . ‘ 0 - I . " . O ‘ ’.-7'O ‘ ' . ,7 iMm-f” 7-:"t7;-7-.7"’7~ ’ =31 -.-x 7 7 I} 4.1m 4’4" y-;/ I ' Jill/3.")! a" _ru'_.- I LIBRARY Michigan State University This is to certify that the dissertation entitled MoLeculAk‘CfiAflACTeztzfiruog 0; THE lfi GENE of” ESCHEZKWA (IOU HOOD 'I'TS INTER. FERN VJl‘Tl-l THE Q9], grre 01- BMTERIDPRAQE Tq_ presentedby (game. KAO has been accepted towards fulfillment of the requirements for PHD degreein MlCl’beOLoaL‘ M a jor {tofessor Date 7/I’L/88 MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 will1111111111111lmmmmll 1293 00649 892 .. g , . .. “:RETURNING MATERIALS: ’ ”AF bV1SSIEJ3‘ 'Place in book dropfito flirty» LIBRARJES . remove this checkout from 4—!!(jl-IL your record. FINES will '9‘ '1‘ . be charged if book is returned after the date stamped below. W... Ni} 3? ’ 191.3. 1% 'A, MOLECULAR CHARACTERIZATION OE THE LII GENE OF ESCHERICHIA COLI AND ITs INTERACTION wITH THE 99% SITE OF EACTERIOPHACE T4 By Cheng Kao A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSPHY Department of Microbiology and Public Health 1988 HOLE COL l BACTE bacte props lit(C Iutal capsj defir with exPre Chara Under the t Sena I ABSTRACT MOLECULAR CHARACTERIZATION OF THE I_._I_1 GENE OF EEEEEBIQELA 99L! AND ITs INTERACTION WITH THE 99L SITE OF BACTERIOPHAGE T4 By Chang Kao Escherichia ggli lig(Con) (ggnstitutive late inhibitors of 14) mutants can completely block gene expression of bacteriophage T4 late in infection and thereby prevent the propagation of the phage. T4 mutants named ggl [grow on lit(Con)) can overcome this block because they have point mutations within a forty base pair region in the major capsid protein gene, gene 23. This forty base pair region defines a cis-acting site that, when wild-type interacts with the lit gene to cause a general inhibition of gene expression. My thesis work is to molecularly and genetically characterize the lit gene and the 113(Con) mutations, and to understand their interaction with the g9; site. The wild-type lit gene and three independently-derived li£(Con) mutant alleles have been cloned and sequenced. The lit gene encodes a 34 kilodalton polypeptide that is associated with the inner membrane of the cell. The li£(Con) mutations are up-promoter mutations that increase the transcription of the lit gene by approximately 10 times and the translation by 7 times in minicell analysis. The lit gene is molecularly mapped to a cryptic DNA element named e14 ‘ of £7 59.1. expr expre intei Eltht site of these characterization studies suggest that the lit and g2; site interact at the inner membrane and that gene expression is not affected until the lit gene is over- expressed because of the lit(Con) mutations. I have analyzed the primary effects of the lit-go; interaction. Genetic evidence suggests that the Lit protein either directly or indirectly binds to the RNA from the ggl site. This then triggers a trans effect which causes a very rapid and specific inhibition of all translation. The trans effect does not immediately affect plasmid or chromosome replication, or the kinetics of transcription. I hypothesize that the lit-go; interaction produces a factor that inhibits the function of all ribosomes in the cell and hence blocks T4 gene expression late in infection. To my beloved Laura I their Hausir 1 his it which every: I Holec fEIIOI Acknowledgements I would like to thank the members of my committee for their guidance and interest: Drs. Michele Fluck, Robert Hausinger, Leonard Robbins, and P. T. Magee. I would especially like to thank Dr. Larry Snyder for his insights, stimulating discussions, and positive attitude which made this thesis research fun, and Laura for making everything else fun. Finally, I would like to thank the Cellular and Molecular Biology program and the Barnett Rosenberg fellowship for partial financial support. Ch Su Table of Contents 2252 List of Tables vii List of Figures viii Introduction 1 Chapter 1. Literature Survey 8 Chapter 2. Cloning and characterization of the Esshsrishia 2211 lit gene. which blocks bacteriophage T4 late gene expression. Abstract 48 Introduction 49 Materials and Methods 50 Results 58 Discussion 74 Literature cited 80 Chapter 3. The lit gene product which blocks bacteriophage T4 late gene expression is a membrane protein encoded by a cryptic DNA element, e14. Abstract 82 Introduction 82 Materials and Methods 82 Results 84 Discussion 97 Literature cited 88 Appendix The interaction between the lit gene and the go} site occurs at the g9; RNA 89 Defining the structure of the ggl sequence which interacts with gplig 94 Theinteractionbetweenthe lit gene and the ggl site produces a specific inhibitor of translation. 100 ggl mutations do not totally prevent the interaction with the lit protein. 120 Summary 123 vi List of Tables Iehls Page 1 Bacterial anui phage strains and plasmids 51 2 Mapping of the lit gene by a three- factor cross and the site of inte- gration of the lit gene clone by transduction 59 3 Bacterial and phage strains and plasmids 83 4 Colors of pUCl2-PZl NTG3 transformed lit(Con) and 1150 cells on plates containing X-gal 9O 5 Effects of IPTG induction on cell survival in Lit(Con) W3110 Iq strains harboring Gol+, Gol-,or pUClZ plasmids 102 6 Lit(Con) plasmids in g. 3911 1046 can reduce the plating efficiency of ggl mutant bacteriophage 122 vii Eissrs 10 ll l2 13 Lies 2: 2152525 Southern hybridization of EggRV- pM6Pl Southern hybridization of EggRV- digested DNA probed with a cosmid clone of lit(Con) gene from MPH6 Proteins labeled in minicells prepared from plasmid-containing strains Evidence that the lit(Con) mutations residein the DNA element e14 Restriction map of e14 showing the location of the lit gene and the sequencing strategy Nucleotide sequence of the lit gene region euui surrounding sequences The lit(Con) mutations increase the resemblance of the putative lit promoter sequence The lit(Con) mutations increase the transcription of the lit gene The lit protein is enriched in the inner membrane preparations from EL 2211 The potential secondary structure of the ggl site and its homology to the ngA sequence of the lambda nut site Growth of Lit(Con) cells harboring Gol+ or 601- plasmids before and after IPTG induction Effects of the liEngl interaction on amino acids transport Effects of the lit;ggl interaction on DNA replication viii 65 69 72 84 84 85 86 86 87 95 103 106 110 '3) Eissze 14 15 16 Lies 2: Eiagrss Effects of the ‘IIEngl inter- action on transcription and translation Northern blot analysis of the effects of the li£;ggl interaction on transcripts of the rprC operon Analysis of protein synthesis before and after induction of the lit; ggl interaction ix 112 114 117 bact heri unde trip and usef curr synt diph rece POSS app: res t over 01's INTRODUCTION The T-even bacteriophages and their host, Eggherighia 3911, have played a central role in the understanding of many fundamental principles in molecular biologyu Some of the classical contributions made by studying T-even bacteriophages include the proof that nucleic acid is the heritable genetic material (Hersey and Chase, 1952), the understanding of the structure of a gene (Benzer, 1962), the triplet nature of the genetic code (Crick et al., 1961), and the demonstration that messenger ribonucleic acid is the molecule encoding proteins (Brenner et al., 1961). The usefulness of T-even bacteriophage is still very evident in current research. For example, studies of the thymidine synthase ( Hall et al., 1987) and. ribonucleoside diphosphate reductase (Gott et al., 1986) genes of T4 recently disproved the dogma that prokaryotic genes do not possess introns. III Laruqy Sruyder's labornrtory, we: beluieve ‘that bacteriophage T4 has not yielded all of its secrets in the regulation of gene expression.Tb this end we plate phage and look for mutants with interesting phenotypes. Our basic approach is to genetically screen for mutant E. 9211 that restrict phage growth, and then isolate phage that can overcome the restriction. In this way we can identify genes or sites in both E. £011 and T4 that interact. Bacteriophage T4 can encode between 150-200 genes (curtmal non-9559 to be 11 polynucl lutants normallj plaques 19711; Si CIISX apj gene exp this res‘ for the hence co This hyp Snyder, atthe t P°lynuc1 backgro, nitrogeg‘ the “0;; restrict “stable “9.5811 “"Perat all “Pp :COOIey “he. The (Guttman and Kutter3 1983). Most of the genes appear to be non-essential in the laboratory environment but are presumed to be needed in natural environments. The gene for polynucleotide kinase-3'-phosphatase (pggT) is ani example. Mutants deficient in the activity of this gene product grow normally in laboratory strains of E. 321.1, but do not form plaques in a clinical isolate CTrSX (Depew and Cozzarelli, 1974; Sirotkin et a1” 1978). This inhibition imposed by CTrSX appears to be at the level of DNA replication and late gene expression (Sirotkin et al., 1978). One hypothesis for this restriction has been that CTer may have been deficient for the cellular homologue of the pggT gene activity and hence could not complement the deficiency of pggT phage. This hypothesis has since been proven false (Jabbar and Snyder, 1984; Kaufman etia1.l987), but it was attractiye at the time and it prompted our laboratory to identify background. When a KalZ strain was mutagenized by nitrosoguanidine, and screened for the ability to restrict the propagation of pggT T4, several mutants were found to restrict these phage at 37°C. These mutations were very unstable and accumulated extragenie suppressors which, by themselves, were able to restrict wild-type T4 at temperatures under 34°C. The suppressor mutations were all mapped to a single site at 25'cn1the chromosome map (Cooley et al., 1979) and appeared to be within a single gene. The mutants were named _l__i_t_ for late inhibitor of I4 for in ackr of I do I gene also iEpc with 23)1 name 521 demo 3 for they severely depressed the gene expression of T4 late in infection. The name was later changed to EE£(Con) to acknowledge that the mutations cause constitutive production of the Lit protein, gplig. DNA replication and encapsidation do not appear to be directly affected. In addition, early genes which are normally expressed late in infection were also turned off (Champness and Snyder, 1982). Bacteriophage T4 mutations which can overcome the block imposed by EE£(Con) E. 22;; were isolated and mapped to within a 40 base pair area in the major capsid gene (gene 23) of T4 (Champness and Snyder, l982L.These mutants were named ggl since they permit growth gn EE£(Con) cells. The g2; site defines a gig-acting sequence that was later demonstrated to interact with the 115 protein. The evidence that ggl mutations are gig-acting comes from a mixed- infection experiment in which EE£(Con) and wildtype E. 3311 are each infected with two T4 strains, one of which has a gal mutation and the other the wild-type g9; sequence (Champness and Snyder, 1982L.The protein products of gene 23 were distinguishable because one phage has an amber mutation down-stream of the ggl site so that one gene 23 protein is of normal size and the other is truncated. When bOth phage infect a 1E5 mutant, the expression of both the 601+ and the 601- gp23 was reduced relative to that in wild- type EE 39;; but the Col- gp23 was not reduced as drastically. The general decrease in phage gene expression suggests that there is an inhibition in trans. The obseIV better indirec Chan IEIOVC cells 1 cells. transfo denonst this ir Thi Snyder Sequenc Calls. 1 fr°fl re base pa Has suf luSt be "”11 th E'seli Berg fron tl ‘ite 1 harbOri 4 observation that the g9; mutant's gene 23 is expressed better suggests that the ggl mutations act, directly or indirectly, in gig with the Egg protein. Champness and Snyder (1984) demonstrated that a ggl site removed from within gene 23 is sufficient to trigger the Egggg effect in _l__3:_(Con) cells. When the ggl site is cloned in a plasmid, the plasmid will not transform 115(Con) cells to antibiotic resistance, but will transform EE£+ celLs. The cloned ggl mutant sequence does allow transformation of a 1;; mutant. These results conclusively demonstrate that no other T4 gene products are required for this interaction. This transformation assay has allowed Bergsland and Snyder (submitted for publication) to define the minimal ggl sequence required to prevent transformation of EE£(Con) cells. Using a combination of Bal3l deletion, and subcloning from restriction sites, it was shown that a maximum of 75 base pairs of the golf site cloned in the _]_._a_gZ gene of pUC8 was sufficient to prevent transformation. The 75 base pairs must be translated, however, since fusions not in register with the 1532 ribosome binding site will transform EE£(Con) E- sali- Bergsland and Snyder made an intriguing observation from these fusions. They observed that, although the ggl site is cloned in frame with the E352 gene, and the cells harboring the plasmids are normally blue when the 115 protein, gplEg, is absent, a wild-type amount of the gplig 5 results in white coloniesn Since this low amount of gpllg could prevent the expression of the fusion gene without killing the cell, the cis-interaction between gplig and the ggl site is separable from the trans effect which would kill the cell by blocking gene expression. Bergsland and Snyder used monoclonal antibodies directed against the B- galactosidase protein to determine if the fusion protein is synthesized in 115(60n) cells. They found the expression of the fusion protein to be severely inhibited in EE£(Con) cells, but not in wild-type cells. One interpretation of this result is that a translational (or transcriptional) termination occurs at the ggl+ site when it interacts with the EEE protein, and the ggl mutations no longer recognize the 1;; protein, and hence do not trigger the termination event. Bergsland and Snyder (submitted for publication) further proved that the Egggg effect requires the translation of the g3; site or the peptide encoded by the ggl site. When the ggl+ site is fused in an alternate frame with respect to the IH 252 gene, so that its peptide sequence is drastically altered, it can transform a EE£(Con) E. 321;. The Egggg effect is abolished. Experiments addressing the level at which the gig effect takes place were not performed. Before I started my thesis project, the g2; site was fairly well characterized. Very little, however, was known about the 1;; gene and the mutations that precipitate the interaction with the ggl site. My thesis project was to sol hop int reg nut str the inte in late The and 6 molecularly and genetically characterize the 115 gene, in hope that this characterization will help us understand the interaction between 1;; and ggl. Several questions regarding EEE were specifically asked: How do the EE£(Con) mutations affect the Egg gene or its expression? What is the structure of the EEE gene and where does gplig function in the cell? At what level, transcription or translation, does 1;; affect the late gene expression of bacteriophage T4? What is the mechanism of the interaction between the 11$ protein and the ggl site? This thesis documents results which help answer these questions. The thesis will be presented in four parts. I will first summarize known mechanisms of gene expression in T4 and published knowledge on several systems of host-phage interaction which have general effects on gene expression in a manner similar to the effects of the 115-39; interaction. I will then present two published manuscripts. The first manuscript details the cloning of the EE£+ gene and several EE£(Con) alleles and evidence that the mutations cause the overproduction of the El; protein. The second manscript completes the physical characterization of El; and presents evidence that gplEE is a membrane protein encoded by a cryptic prophage named e14 and that the EEE(Con) mutations are up-promoter mutations. In the appendix I present results which will be published when they are more complete. The experiments presented address the mechanism of the interaction between gpllg and the ggl site. I analyzed the level at which the cis and trans effects takes place. I have evidence that the cis-effect occurs at the level of transcription and the trans inhibition of gene expression is due to the production of a substance which specifically inhibits translation. Chapter One ll pl hu H e1 be pr ma. ex] Enc Chr Pro Literature Survey The T4 Infection Process Bacteriophage T4 is a double-stranded lytic DNA virus that infects EEEhEElEhlE 321$. It has a prolate icosohedral head, a cylindrical tail, and six fiber-like projections that emanate from the base of the tail structure. The head contains the phage chromosome which is a linear molecule of approximately 160 kilobase pairs. The complex tail structure helps to punch a hole through the bacterial cell wall, and inject the DNA into the cell (Goldberg, 1983). The fibers mediate the attachment of the phage to the cell and determine the host range of the phage (Riede et al., 1985) All of these structures allow the phage to efficiently infect the cell. Once the phage DNA is within the cytoplasm, phage gene products will be expressed and release of hundreds of progeny phage occurs within 30 minutes at 30 C. What happens within the 30 minutes is a very complex and elegantly regulated developmental process that is just beginning to be understood. One minute after T4 infection, the synthesis of E; 5911 proteins and mRNAs totally ceases and the biosynthetic machinery of the cell is comandeered by T4 gene products to eexpress only phage genes. Some of the early genes of T4 encode nucleases which will specifically degrade the host chromosome in order to provide nucleotide precursors for T4 biosynthesis (Prashad and Hosoda, l972).0ther early gene products include several proteins which replicate the phage 8 9 DNA (Spicer and Konigsberg, 1983). Replication can take place about 5' post infection. After DNA replication, the T4 late genes are expressed. Most of the late genes encode structural proteins required for formation of the T4 capsid. The head complex and the tail complex are assembled separately on the inner membrane of the cell (Kellenberger, et al., 1968; Wood and Conley, 1979). Approximately 15' after infection, the heads are stuffed full of the replicated DNA (Kalinski and Black, 1986; Serwer, 1986), and preassembled tails are put on the heads to complete the phage capsid (Wood and Conley, 1979). T4 then lyses the cell to release progeny phage. Transcriptional Regulation Much of T4 gene expression is regulated at the level of transcription. The transcription pattern is classified into three general phases: early, middle and late (Christensen and Young, 1983}. The early and middle transcripts start from specific promoters and extend counter-clockwise with respect to the circularly permuted chromosome. The late transcripts are initiated from the opposite DNA strand and could be transcribed as an operon or from individual promoters (Young, Mendard, and Harada, 1981). Some RNAs Called anti-late that can hybridize to the late transcripts are made during the early and middle phases, and these RNAs may regulate the expression of the late genes (Young et a1” 1980; Elliot et a1” 1984). The early RNAs are expressed as polycistronic messages fit 19' in: Pr' {IE 30 am tl’E Ch its II Eh: be an- Co fa id In 31 10 from early promoters in two main blocks (Caruso et a1” 1979). The genes proximal to the promoters are named immediate early (IE) genes and the genes distal to the promoters are named delayed early (DE) genes. The transcription of the immediate early genes initiates within 30 seconds of infecting the cell and is insensitive to the antibiotic chloramphenicol (CM) (Brody et a1” 1983). The transcription of the DE genes are dramatically inhibited by CM, suggesting that translation.or the act of translation itself is required for the expression of these genes. The switch from expressing the IE to the DE genes is regulated, at least in part, by transcriptional termination. E; 92;; mutants that allow the transcription of the DE genes in the presence of chloramphenicol have been isolated (Stitt et al.1980; Simon et a1” 1974; Revel et a1” 1980). These mutants all have defects in the nucleoside triphosphate-dependent transcriptional termination factor, the- Transcriptional termination is best understood in the bacteriophage Lambda system (see below). Lambda has an antitermination factor named E which allows transcription of genes downstream of a termination signal (Friedman and Gottesman, 1983). T4 may also have an antitermination factor, 3230 (Caruso et al., 1979). This mutation is identified as a suppressor of a special type of E; 52;; £119 mutant named EEEC(Caruso, et al., 1979). EEEC mutants do not allow the expression of the DE genes even without CM. 3930 11 mutants can overcome the block imposed by EEEC E; 391;. The exact function of the gg_C gene product is still under investigation. Experimental observations support the hypothesis that translation (u: an antitermination factor is needed to overcome the normal termination that occurs in the junction between the IE and DE genes. Therefore, when CM is present the antitermination signal in the transcripts is exposed and the Egg factor causes termination to occur. The transcripts from the middle promoters are not well understood because their expression is temporally overlapped by expression from early promoters (Brody et a1” 1983). It is known, however, that the functions of the the T4 335A and B gene products are required. The Egg genes were originally identified as suppressor mutations in a EEE' cell (Takahashi and Yoshikawa, 1979) The gpggg polypeptides are two of the first proteins made in T4 infection. They are DNA-binding proteins that may allow the recognition of the middle promoters (Uzan et al., 1985). Cloning middle mode genes in plasmids may circumvent the need for gpggg (Schinedling et al., 1986). Transcription in these plasmid clones appears to initiate at the proper promoters. The authors hypothesize 'that the superhelicity of the plasmid may circumvent the need for gp_9__t_ in the activation of middle promoters (Shinedling et al., 1986). The gpggg protein may also play a role in antitermination (Daegelen etialu 1982).This result came from examination 12 of the expression of DNA polymerase, gene 43, of T4. When a Egg- T4 infects a EDQi E; 3211, no gene 43 RNA is found early in infection. However, if the same phage infects a 5E2; E; 221;, then gene 43 is expressed at about one-quarter of normal quantity (Daegelen, et a1” 1982). This suggest that Egg may activate the expression of gene 43 as well as antagonize the action of the Rho protein“ However, since Egg mutations seem to affect the expression of several genes in T4, gggfs effect on 3E2 may be indirect. The switches in the various pre-replicative modes may involve modifications of the host RNA polymerase (RNAP) complex (Rabussay, 1983). The IE transcription is thought to require no alteration of the RNAP because the IE transcripts can be synthesized in vitro using a purified preparation of RNAP (Brody and Geiduschek, 1970). Transcription of the late genes, however, does not occur in vitro, suggesting that the bacterial RNAP complex must be modified in order to recognize late promoters. Consistent with this observation is the finding that late promoters of T4 do not have the normal -35 consensus sequences of E; 321$ (Geiduschek et al., 1983) and shouldn't be recognized by unmodified RNA polymerase. Two polypeptides of T4, gp 33 and 55 have been Shown to bind to the RNAP at approximately 5' after infection (Kassavetis and Geiduschek, 1984). The product of gene 55 can act as a specificity factor in vitro, much like the sigma specificity factor that is normally associated with the RNAP complex (Malik et al., 1985). The 13 protein product of gene 45 also acts upon the RNAP to activate late gene transcription,tnn:it does not directly bind to the transcription complex. Throughout the entire course of infection, there are a series of modifications of the host RNAP. Most are not required for transcription of T4 genes, but may be required for the shut-off of cellular gene expression (Rabussay, 1983). DNA replication is required for the transcription of late genes ( see Geiduschek, et a1” 1983 for review). The replication process somehow makes the DNA competent for transcription. One hypothesis is that late promoters are only recognized after DNA replication when they are single- stranded or when the DNA around the promoters is somehow exposed. The late gene, 23, is also expressed when cloned in plasmids ( Jacobs and Geiduschek, 1981). Presumably the initiation of transcription of cloned gene 23 is due to initiation from plasmid promoters. Transcription of T4 late genes is usually dependent on the modification state of the T4 DNA. Normally, T4 has hydroxymethyl-cytosine (HMC) in place of cytosine in its DNA. If the phage has a mutation that inserts cytosine instead of HMC, the expression of the late genes is severely ' reduced (Snyder, et al., 1976). T4 mutants that can express the late genes in the absence of hydroxy-methyl cytosine have mutations in a gene designated 212122: because they Ellow gytosine in the DNA and affect an ggEolding of the host nucleoid. The protein product of gig/33E participates 14 in shut-off of host transcription since the mutations allow some transcription of E; coli genes (Sirotkin, 1978). gpglg which prevent glglggg function and thereby permit transcription from cytosine DNA are mapped in the £298 subunit of the RNAP complex (Snyder and Jorrisonw 1987; Ka and Snyder unpublished data). In addition, the purified gpglg binds to DNA (Snustad, 1987). Translational Eggglgglgg lg bacteriophage IS Expression of the pre-replicative genes is also regulated at the level of translation. Phage mutations in a gene named EggA will increase the expression of a specific set of early genes (Miller, E. et al., 1985) In addition, the mutant EggA protein is also overexpressed, suggesting that EggA+ inhibits its own translation (Cardillo et a1” 1979). Many EggA-controlled genes have an uridine-rich sequence in their translational initiation site. It is hypothesized that gpgggA competes with ribosomes in binding to this sequence and thereby prevents translation (Karam et al., 1981). EggA cloned in plasmids can regulate the expression of cloned target genes (Miller, E., et alul985), indicating that no other T4 proteins are required. 'for this translational inhibition. The expression of the T4 lysozyme gene is also regulated at the level of translation. The lysozyme gene is transcribed early in the infection process, but not translated. The message is translated, however, when it is transcribed late in infection. Comparison of the structure 15 of the early and late lysozyme mRNAs revealed a difference at the 5'-terminus; the late messages do not contain a leader sequence which can form an elaborate secondary structure (McPheeter et al., 1986). The sequence within the stem of the secondary structure contains the ribosome binding site and perhaps prevents the association of the ribosomal complex and the message, hence preventing translation (McPheeter et al., 1986; Knight et al., 1987). she protein translated from gene 32 has roles in several processes of bacteriophage T4, including DNA replication, recombination, and repair. It also seems to regulate its own expression at the translational level (Gold, et al., 1976). When the first seven amino acids of gene 32 are fused to the B-galactosidase gene, the fusion protein is not expressed when the intact gene 32 protein is present, suggesting that gp32 represses its own translation by binding to its own 5' sequence. In addition, the translation of several other genes that are co-transcribed in the same message as gene 32 is also inhibited” Amazingly, genes both upstream and downstream of gene 32 are inhibited. The mechanism seems, in some respects, analogous to a negative enhancer element (Young, et a1” 1980). Bacteriophage T4 run: only regulates its gene expression at the translational level, but there also is evidence that it prevents the translation of cellular transcripts. Hsu and Weiss (1969) have isolated a factor that is produced 5 minutes post infection. This factor apparently binds to the 16 ribosome and reduces the efficiency with which the ribosome recognizes cellular ribosome binding sites. 13 head assegbly Assembly of the T4 capsid occurs about 15 minutes after the infection. While the head and tail complexes are assembled in independent pathways on the inner membrane, the phage DNA is replicated into a multi-chromosome concatamer. The chromosome is then stuffed into the assembled T4 head and a nonspecific endonuclease releases the rest of the concatamer (Kalinski and Black, 1985). Finally, the head and the tail complexes couple to form the intact phage. The formation of the prohead occurs in several steps. The prohead is assembled around a membrane initiation protein, gp20. A structure composed of seven polypeptides first forms a core complex on ngO (Caldentry et al., 1987), then a "shell" structure is assembled around the core complex (Kuhn, et a1. 1987). The shell is mostly composed of gp23, which is the major polypeptide in the capsid, and gp24, which forms the vertices in the capsid. After the assembly of this complex the inner core is proteolytically digested by the protein product of gene 21. This protease recognizes 'a specific secondary conformation of the prohead polypeptides as the cleavage signal (Keller et al., 1985). When the gp23s within the capsid are cleaved, two major events are activated: 1) the shell undergoes conformational change and results in increasing the inner 17 volume by 50-1008 (Rao and Black, 1986); ii) a subpopulation of gp23 becomes activated to assume the role of the DNA packaging enzyme and non-specific endonuclease (Rao and Black, 1986). The gp23 enzyme measures out a headful of DNA, then cleaves the DNA. The assembly of the core determines the shape and size of the mature head. Mutations in the polypeptides of inner core complex can result in formation of long tubular heads. (Eiserling et al.,l984; Keller, et al., 1986; Volker et al., 1982). The ratio of gp23 and other head proteins also seems to be important and an imbalance in the ratio can affect the length of the heads (Eiserling et al., 1984; Doherty et a1, 1982). Mutations in several non-structural head genes can affect the assembly of the T4 head. Mutations in gene 20, the initiation protein, result in a significant delay of prohead assembly and formation of aberrant heads (Brown and Eiserling, 1979). Mutations in T4's gene 40 result in a similar phenotype and the gene 40 product may play a catalytic role in the assembly of the core complex around gp20 (Brown and Eiserling, 1979). In an E; 321; mutant strain that does not permit the head assembly at low temperatures, it was observed that gp23 forms an aggregated lump in the cytoplasm (Kuhn et al., 1987). Also, Laemmli et a1.(l970) have observed that mutants defective in gene 31 function also result in gp23 aggregates. Gene 31 is hypothesized to direct the assembly of gp23 at the inner 18 membrane. The signal sequence in gp23 that targets it to the site of prohead assembly hasrunzyet been identified. Bergsland and Snyder (submitted for publication) propose that an amino acid sequence that is present about one-fourth of the way into gp23 could be such a signal. Issassriesieaal 5251;22591225122 229 Esaslasiea 2t 5222 exnression is hassszieehass 122E92- The transition from immediate early to delayed early gene expression in bacteriophage T4 is regulated at the level of anti-termination. Furthermore, we have evidence that the ggl site of T4 may interact with E_._ 32;; EE£(Con) mutations at the level of transcription (see appendix, this volume) and the ggl site contains structures that bear strong similarity to the antitermination signals of bacteriophage lambda. Therefore, T4 probably uses antitermination mechanisms to guide its transcription. In this section I will review the current thinking on how antitermination is regulated. Both lambda early and late genes, transcribed in different operons, are regulated by transcriptional antitermination. The expression of the early genes is regulated by a phage-encoded protein named N which allows the RNA polymerase to proceed beyond normal transcriptional termination signals in the DNA (Friedman and Gottesman, 1983); that is, when N is expressed, the transcripts made are significantly longer than when N is not expressed. Another protein named Q regulates the expression of the 19 lambda late genes (Roberts, 1988). In this section I will restrict the discussion to antitermination in the early genes. The N protein binds to a site near the 5' end of the early genes' operon named 33;, for E ggilization site (Salstrom and Szybalski, 1978). This site is not the actual termination signal. In fact, the termination signals are located thousands of base pairs downstream of the ngg site (Friedman et al., 1973). Deletion mutations that remove the termination signals have also been identified and these mutants can express the early operon in the absence of the N protein (Salstrom and Symbalski 1978b). The structure of the BEE site has been determined from several lambdoid phages. This site of approximately 50 base pairs has several well conserved domains. There is usually a conserved stretch of approximately 9 bp called EggA followed by a less conserved stem-loop structure called E958. Finally there is another G-T-rich area named E250 (Li et al., 1984). The 2358 domain has been hypothesized to bind the N protein (Horwitz et al., 1987). The fact that E258 is not well conserved between different lambdoid phages may reflect the divergence in the li protein sequence of these phages. Mutations in either 995A or E258 would affect the ability of N to antiterminate. It is not known if EQEC is required for antitermination. Mutations in EgEA can be suppressed by cellular mutations in a gene named 235A (Friedman and Olson, 1983). The 325A gene is required for antitermination since cells without te II It at id at oi a1 (31 bi p: ac tc Ce a] ti C: 20 EEEA do not antiterminate and hence do not allow lambda growth. There are several other cellular factors required for antitermination (Friedman and Gottesman, 1983). Mutations in genes named EEEB QEED, and gggE will abolish N-mediated antitermination. The gggD mutation is in the Egg termination gene. The 235E mutation has been mapped in the ribosomal subunit, 310, although it isn't clear how a mutation in a ribosomal protein affects transcriptional antitermination. Greenblatt's laboratory has biochemically identified another factor, gggG which is required for antitermination in vitro. Friedman and colleagues have proposed that the NusA protein binds to the EQEA sequence because point mutations in RQEA seem to decrease antitermination (Friedman and Olson, 1983). It came as a surprise when other laboratories observed that deletions of 925A did not decrease N-mediated antitermination (Patterson etzial” 19870. ‘Work in Greenblatt's laboratory suggested that NusB may actually bind to 225A since NusB does not affect antitermination without EgEA. They have also demonstrated that several proteins including N and NusA bind to the RNA polymerase. In addition, two molar equivalents of the NusA protein appear to bind to the polymerase (Greenblatt, 1987) and NusA and N can bind to each other (Li and Greenblatt, 1981). Barik et a1. (1987) have examined the ability of the N protein and the transcriptional complex to associate with the RNA containing BEE site in an in vitro single round 21 transcription experiments. They showed that the N protein can become a stable subunit on the RNA polymerase as the polymerase moves through the 5335 site. Since N is normally thought to bind to EQEB- it must translocate onto the polymerase. Horwitz, et al., (1987) used reconstitution experiments to establish the following order of events that results in antitermination. They preposed that one unit of NusA binds to the RNA polymerase complex before the complex transcribes the 235 site. This binding may cause the removal of a sigma factor initially associated with the polymerase. The complex than transcribes over the 233 site allowing N, the second NusA, 810, and other Egg factors to associate with the 925A and B signals. These proteins then translocate onto the polymerase complex, further modifying it to prevent the polymerase from recognizing termination signals. After transcribing the operon, the modifying proteins presumably fall off the polymerase. This very complex regulation of transcription is not restricted to phage systems. The ribosomal RNA operon of E; 92;; is regulated by antitermination (Morgan et al., 1986). This operon even possesses the characteristic EQEA and B ‘signals but the gggA.and E mutations do not seem to affect antitermination in this system. NusB may have a slight effect in in vitro studies (Sharrock et al., 1985). The observation. that NusA. is not involved in rRNA antitermination is not unexpected. NusA also plays no role 22 in lambda late gene antitermination. Each termination system may have specific associated factors. The Egl operon involved in regulation of E-glucoside metabolism is also controlled by antitermination (Mahadevan et al., 1987). Work with the Eggillgg trytophan operon (Shimotsu et al., 1986) and even eukaryotic genes (See Roberts, 1988) have recently implicated antitermination in the regulation of gene expression. A novel twist has recently been added to the role of lambda antitermination" A lambdoid phage named HK022 can prevent transcription of infecting lambda phages. HK022 apparently produces a protein named app that binds to the Egg site of lambda to cause transcriptional termination at the 3335 site (Roberts et al., 1988). In this case, the Nus proteins stimulate Nun-mediated termination at or near the 335 site and does not stimulate antitermination. Host genes that Effect general gene expression 9E EpEggElpg bacteriophageg There are many E._ gp_ll mutations which affect gene expression of bacteriophage T4. I have already discussed mutations in the 5E2 antiterminator and mutations in the glucosylation pathway that fail to modify the T4 chromosome land result in the phage chromosome being degraded in a restrictive host (Revel, 1983). In this section I would like to focus on gene-gene interaction ofridifferent nature.I will review the literature of chromosomal elements (cryptic prophages, integrated plasmids, F-factors)‘withinIELcoli 23 that somehow prevent normal gene expression of the infecting phage. These gene-gene interactions all cause an abortive infection; the cell is killed, but because some processes of the bacteriophage are inhibited, the phage produces no progeny. These phenomena represent a type of interaction that could reveal the normal regulation of gene expression and provide perspective on the EEE-gpl interaction. There are several well understood examples of abortive infection (Duckworth et al., 1981). Three systems are especially well studied: the interaction 'between bacteriophage T7 and the pig gene on F sex factors; the interaction between T5,cn:its close relative BF233, and colE plasmids; the interaction between cryptic prophages 12222222122 2222222 2222222212222e T1 222 21.1- 22!. 2222 22 .13 222222 The gene expression of bacteriophage T7, like T4, is temporally separated into three classes (Morrison et a1” 1974). When T7 infects an E_._ 52;; strain that harbors the sex factor F, genes from the earliest class are expressed normally, but subsequent gene expression is inhibited by about 20 fold (Makela et al., 1964). The defect is caused by presence of the 2;; genes in the F factor because defects in these genes allow normal infection by bacteriophage T7 (Morrison and Malamy, 1971). The pig operon contains two structural genes named pEEA and B. Another gene upstream of pEEA and B, named pEEC, regulates the expression of the 24 operon by directly or indirectly interacting with a cis- regulatory site named pEEO.( Millerg et.al” 1985; Miller and Malamy, 1983). The subject of F—factor inhibition of T7 development remains hotly contested and the level (transcription or translation) at which the T7 gene expression is inhibited remains unclear. Malamy and his associates have physiologically characterized the effects of the pi: genes on bacteriophage T7. They examined T7 late gene expression in F+ and F- cells. Although they found the levels of late mRNAs comparable, the mRNAs are not translated in vivo. When purified, the late mRNAs are translatable in vitro and have similar stability as T7 late mRNAs produced in F- cells (Blumberg and Malamy, 1974). Based on this evidence Malamy's group suggested that T7 abortive infection in F+ cells is due to a translational block imposed by the pig proteins. They also proposed that T7 very cleverly modifies phage transcripts and that p1: proteins prevent this modification (Blumberg, Mabie, and Malamy, 1976). Young and Menard (1975) prepared translational extracts from T7-infected F+ and F- cells and used them to in vitro »translate exogenously added T7 early or late mRNAs. They found that the translational efficiencies of both cell types were similar. They also measured the amount of early and late T7 mRNAs in F+ and F- cells and found that several messages were not synthesized at a detectable level in F+ th “h P0 0b Ce: 25 cells but were present in F- cells. Their interpretation is that the pi; gene products inhibit both the transcription and translation of the T7 late genes and has no specific effect on translation. Yamada and Nakada (1975) published results consistent with those of Young and Menard, but they also observed that the translational extracts from T7- infected F+ cells lost activity faster than those from F- cells, although the loss was nonspecific and did not discriminate against T7 messages. This observation might explain the Malamy group's proposed translational defect. Two laboratories have also found that the permeability of the T7-infected F+ cells increased after infection and that ATP may have leaked out of these cells (Condit and Steitz, 1975; Britton and Haselkorn, 1975). Since gene expression is highly sensitive to energy levels in the cell, it is not surprising that transcription and translation were affected. Condit and Steitz, (1975) and Britton and Haselkorn (1975) also observed that the inhibition of gene expression occurred at about the same time as the increased permeability of the cells. These laboratories believe that the interaction between 2;; proteins and some function in the T7 early genes affect the integrity of the cell membrane ,which then cause a non-specific leakage of the nucleoside pools and hence inhibited gene expression. In support of tflua Malamy group's results is the observation that nucleoside leakage occurred even when cells are pEE- and T7 plated normally (Blumberg, Mabie, and 26 Malamy, 1976; Schmitt and Molineux, 1987). Furthermore, Remes and Elseviers (1980) have isolated suppressor mutations which allowed the plating of T7 phage on Pif+ cells without preventing the leakage of' cellular nucleosides. These mutations map either in the gglU locus which encodes a function that alters the T7 receptor on the bacterial cell surface or in the EREL locus which is involved in translational elongation. The exact role of these mutations in the inhibition of T7 development in F+ cells is unclear. It does appear, however, that the ATP leakage may not be a satisfactory answer to why the inhibition occurs. 12222222222 2222222 2222222222222 22 222 2222 21.222222 When bacteriophage T5, or its close relative BF 23, infects an E; gpll harboring a colE plasmid, early phage gene expression is normal, but the late genes are not turned on, and all RNA and protein synthesis abruptly cease at about 10' post infection (Cheung and Duckworth, 1977). Physiological studies revealed that these infected cells have sustained membrane damages; the cells no longer take up labeled amino acids, and they release ATP and B- galactosidase into the culture medium (Cheung and Duckworth, -1977; Glenn and Duckworth, 1979). The effect gplEl has on T5 is not due to a stable alteration of the RNA polymerase because the RNA polymerase isolated from infected cells can transcribe templates in vitro (Cheung and Duckworth, 1977). The membrane damage induced in the cells containing a 27 gp_lE plasmid is due to a plasmid-encoded function (Altier et a1” 1986). The identification of the gene responsible was, however, not straight-forward.lfl:is known that the 321E plasmid can synthesize the antibiotic colicin, encoded by the Egg gene. The colicin will kill plasmidless cells. The original cell is protected from the action of the colicin because.of the function of another plasmid gene named Egg. BF23 was believed to induce the production of colicin (McCorquodale et al., 1979). Furthermore, BF23 that are defective in the function of an early gene named A3 could productively infect plasmid-containing cells (McCorquodale et al. 1979). This lead to the hypothesis that the function of A3 inactivates the product of the Egg gene, and hence allows the produced colicin to cause membrane damage and kill the cell. Although attractive, this hypothesis is at least partially incorrect. Colicin may run: be the molecule responsible for inhibition of gene expression in bacteriophage T5. Mitomycin exposure or UV-irradiation can also induce cell killing in a manner very similar to the phenomenon induced by T5. Cells containing ggEEl are killed by much lower doses of mitomycin than cells without ggEE plasmids (Zhang et al., 1985). The killing functions are activated by SOS-induced cleavage of 135A proteins which normally repress the killing functions (Ebina et al, 1983). It came as a surprise when it was discovered that plasmids that can not produce colicin are still sensitive to mitomycin (Hull and Moody, 1976). Another 28 gene located downstream from the Egg open reading frame appears to be actually responsible for the killing, hence it is named Ell. The El; gene is also responsible for the cessation of macromolecular synthesis, loss of capacity for active transport, and disruption of membrane potential (Suit et al, 1983; Altier et a1, 1983L It is apparently induced when T5 infects the cell. The El; gene and the Egg gene define an operon with the origin of transcription beginning with the Egg promoter (Sabik et al., 1983). The open reading frame encoding Ell overlaps with the open reading frame for the Egg gene, but the 1-! gene is transcribed in the opposite direction to that of the 22931911 operon. This motif is found in several members of the gplE plasmid family (Jakes and Zinder, 1984) Conclusive proof that the induction of the EEE gene function causes cell death came from a series of experiments from Luria's laboratory. Luria and his associates have used transposon mutagenesis to distinguish the functions of Egg and El; (Sabik et al.,l983; Suit et a1” 1983). They have mutants which disrupt either the function of Egg or Ell but not both. The 932-, kil+ plasmids showed all the phenotypes associated with cell killing. They have also (cloned the El; gene and under control of an inducible lactose promoter. This clone was maintained in an unexpressed state in a lactose repressor overproducing strain, W3110 Eq, and Ell is synthesized only after induction of the lactose promoter with the gratutitous 29 inducer, IPTG. The induced cells exhibited all the phenotypes observed with induction of the intact gp__lE plasmid, thus conclusively demonstrating that the Ell gene may be the only gene required in the barrage of effects observed when bacteriophage T5 induces the gplE plasmid. The killing effect of the Ell gene is somewhat reversible. If the inducer is removed from the cell after only a 10' exposure, about 10% of the cells still retain the ability to form colonies. Also, Mg++ at certain concentrations will delay cell killing. The effect of Mg++ is observed in other systems (see below) and is interpreted to provide an enhanced stability of the bacterial membranes. Even in this system where most of the effort has been focused on membrane damage, it has not been proven that membrane damage is the most direct effect of the 5;; gene. 2212222 22222222 222222 222 22 c011 chr0208022 that cause novel DNA elements that contribute interesting functions to the cell. The first DNA restriction-modification system discovered was a gene within a cryptic DNA element that mapped at 25' (Brody and Hill, 1988). There are also several prophages that are similar in genetic organization to bacteriophage lambda scattered around the chromosome (Redfield and Campbell, 1987L.The first cryptic lambdoid prophage was discovered by its ability to induce recombination in a strain that was supposedly recombination- deficient (Gottesman et al., 1974). This phage was named 30 Egg and it maps at 29 minutes on the chromosome. All of the cryptic lambdoid phages seem to be defective phages and have suffered large losses of genetic material. Phages, like plasmids, may have evolved mechanisms to protect the host cell from further invasion by other phages of similar types. For example, a cell containing a lambda prophage can not be infected by another phage of a similar immunity (Toothman and Herskowitz, 1980). Several functions assigned to prophages have also been demonstrated to prevent infection by related phages. Examples within this category that I would like to review are : l) the restriction of the p55 locus of bacteriophage T4 defective in the function of polynucleotide kinase, 3'phosphatase. 2) the restriction of EII- bacteriophage T4 by lambda prophages. First, I would like to summarize the biology of the cryptic DNA element e14, which contains the 1;; gene. The cryptic DNA element e14 has been mapped to 25' on the E; 32;; chromosome. It is so named because it is approximately 14 kilobase pairs (Greener and Hill, 1980). e14 is called a DNA element since it is not clearly established whether it is a prophage or an integrated plasmid. It is known that e14 can excise from the chromosome -as a circular molecule after UV induction and reintegrate into the same position (Brody and Hill, 1985). The DNA region around the integration site has been sequenced (Brody and Hill, 1988) and it is highly homologous to the host chromosome over a 216 bp region. The actual integration 30 Egg and it maps at 29 minutes on the chromosome. All of the cryptic lambdoid phages seem to be defective phages and have suffered large losses of genetic material. Phages, like plasmids, may have evolved mechanisms to protect the host cell from further invasion by other phages of similar types. For example, a cell containing a lambda prophage can not be infected by another phage of a similar immunity (Toothman and Herskowitz, 1980). Several functions assigned to prophages have also been demonstrated to prevent infection by related phages. Examples within this category that I would like to review are : 1) the restriction of the p5; locus of bacteriophage T4 defective in the function of polynucleotide kinase, 3'phosphatase. 2) the restriction of Ell-'bacteriophage T4 by lambda prophages.F1rst, I would like to summarize the biology of the cryptic DNA element e14, which contains the 1E5 gene. The cryptic DNA element e14 has been mapped to 25' on the E; ggll chromosome. It is so named because it is approximately 14 kilobase pairs (Greener and Hill, 1980). e14 is called a DNA element since it is not clearly established whether it is a prophage or an integrated plasmid. It is known that e14 can excise from the chromosome ass a circular molecule after UV induction and reintegrate into the same position (Brody and Hill, 1985). The DNA region around the integration site has been sequenced (Brody and Hill, 1988) and it is highly homologous to the host chromosome over a 216 bp region. The actual integration 31 site is probably an 11 bp region at one end of the conserved sequence. (Brody and Hill, 1988). There are several genes identified in e14 that are of interest. Many of these genes seem to duplicate other functions of the cell. The pig gene can actually complement the invertase function of the Mu phage gene, gig and is functionally similar to the flagellin invertase gene (van de Putte et a1” 1984). Another e14 gene named, EEEC, functions in blocking cell wall formation and is functionally analogous to the gElA gene (Jaffe et al., 1986; Maguin et.alq 1986). The e14 element also contains the Egg, methylcytosine restriction-modification system (Brody and Hill, 1988). It was discovered because some E; 231; strains can prevent plaque formation by bacteriophage T4 that did not modify its DNA with glucosyl moieties (Revel, personal communication). The restriction enzyme has been renamed __5A and is specific for 5-methyl cytosine containing DNA (Raleigh and Wilson, 1986). A clinical isolate of E; 92;; named CTrSX can restrict the infection of bacteriophage that are defective in any of several functions, including polynucleotide kinase (also a '3' phosphatase) (Depew and Cozzarelli, 1974; Sirotkin et al., 1978), RNA ligase (Jabber and Snyder, 1984), and the product of the 33; gene (see Jabber and Snyder, 1984). The restriction is partially temperature-dependent, and is more severe at 30 C than<42- ‘ “F? 1““ r} . A .._.. - 1‘. . . . vs -- I“ r . . .- - - ‘v-‘r - ABCD EFGH 11 K [M FIG. 9 The lit protein is enriched in the inner membrane prep- arations from E. coli. The plasmid-encoded proteins were labeled in maxicells. and the cells were fractionated into periplasm. cyto- plasm. and membrane components as described in Materials and Methods. A portion of each membrane pellet was then banded in an isopycnic sucrose gradient. Lanes: A. total proteins from the pABglII+4 plasmid which has a frameshift mutation in the lit gene: B. total proteins from the pM6E1 plasmid containing the (it gene: C and D. total membrane pellet proteins from pABglll+4 and pM6E1 plasmids. respectively: E. to G. outer membrane fractions from the isopycnic gradient of the pM6E1 cells: H to J. mixed membrane fractions from the gradient: K to M. inner membrane fractions from the gradient. The gplir is the diffused band marked by the arrow at the top of the gel. 1"..qu qt 7‘. _ a. A.. A We labeled the plasmid-encoded proteins by the maxicell technique and then fractionated the cell into cytoplasmic. periplasmic. and membrane components. A portion of the membrane fraction was further separated into outer and inner membrane fractions with isopycnic sucrose gradients. There were usually five distinct bands due to the plasmid- encoded proteins: the two forms of B-lactamase. chloram- phenicol acetyltransferase. and the two insert-encoded poly- peptides of 34 and 17 kDa. The 34-kDa polypeptide is the (it gene product identified previously by three criteria: (i) it is the correct size. (ii) it gives the same difl‘used band as the 34-kDa protein overproduced in the minicell experiments (12). and (iii) it is missing in cells containing the pABglIl+4 plasmid that has a frameshift mutation at the BgIII site in ORFI (Table 3). When the maxicell-labeled proteins (Fig.9) were fractionated into the cytosol. periplasm. and inner and outer membranes. the antibiotic resistance proteins segre- gated as expected. The chloramphenicol acetyltransferase was enriched in the cytosolic fraction. and the mature and precursor forms of B-lactamase polypeptides were enriched in the periplasm and the cytoplasm. respectively (data not shown). The only detectable membrane protein encoded by the plasmid vector is the tetracycline resistance protein. which is 40 kDa. However. in these EcoRV clones the re! gene is disrupted and the protein should not be produced. The amp proteins segregated with the outer membrane. as apparent from the stained~gel pattern (data not shown). In contrast to the other proteins. gplit was enriched in the inner membrane fraction. The Lit polypeptide was enriched in the total membrane fraction in the pM6E1-containing cells (Fig.9. lane D). and when a portion of the membrane pellet was banded on an isopycnic sucrose gradient. a difl‘used band of the same molecular mass as gplir further segregated into the inner membrane fractions (Fig.9. lanes K to M). Some of this band also appeared in the mixed-membrane fraction (lanes H to J). but is largely absent on the outer CHARACTERIZATION OF THE Lit PROTEIN membrane fractions (lanes E to G). The lower band present in lanes L and M probably represents some of the chloram- phenicol acetyltransferase which was trapped in the total membrane pellet. DISCUSSION We have shown that the Lit(Con) phenotype for blocking gene expression late in infection by bacteriophage T4 is due to the overproduction of a 34-kDa protein. gplit. encoded by the cryptic DNA element e14. The lit gene straddles the HindIII site on the left side of e14 closest to purB (4) and is transcribed from left to right with respect to the integrated e14 map. We have also sequenced the lit gene and deter- mined the base pair changes due to three lit(Con) mutations. all three of which were CG-to-TA transitions at the same base pair. 60 nucleotides upstream of the presumed start codon of gplr'r. These mutations result in a promoter which more closely approximates the consensus promoter in E. coli and cause an increase in the amount of the lit gene RNA by Northern analysis. We also present evidence that gplr'r is an inner membrane protein. since it has the characteristic hydrophobic domainsand segregates with the inner mem- brane during cell fractionation. All of our evidence thus far supports the idea that the (it gene is unregulated and normally expressed at low levels from a weak promoter just upstream of the gene. It is of interest why e14 expresses this gene in the dormant state. but this could be difficult to determine. since all that is known about e14 is that it can excise and reintegrate and that it does not contain genes essential to E. coli (4. 11. 15). It seems likely that e14 is a defective prophage or integrated plasmid. and many of its functions could be vestigial. The Lit protein. even when grossly overproduced. has no apparent effect on the E. coli host. It is only after infection by bacteriophage T4 that its presence becomes known. The mechanism through which gplr‘r blocks gene expres- sion late in T4 infection still remains a mystery. Its efl‘ect becomes apparent when the late T4 gene 23 is first tran- scribed and translated. The interaction between gpli't and a short sequence within T4 gene 23. the go! site. blocks the expression of gene 23 in cis (6: Bergsland and Snyder. submitted) and other genes in trans (7: C. Kao and L. R. Snyder. to be published). Considering our evidence that gplir is an inner membrane protein. this interaction presumably occurs on the inner surface of the inner membrane. The product of T4 gene 23 is the major capsid protein gp23. which is thought to bind to the T4-encoded inner membrane protein gp20 during the normal assembly of the procapsid (3). How gp23 is targeted to the membrane is unknown. From other work (Bergsland and Snyder. submit- ted). we know that the Lit protein seems to cause the expression of gene 23 to terminate in the go! region. Perhaps the go! site is the sequence that targets gp23 to an assembly complex which includes T4 gp20 in the membrane. Accord- ing to this hypothesis. translation (and/or transcription) normally pauses at the go! site until the site is properly docked at the membrane assembly complex. Then. and only then. can translation (or transcription) continue. When gplir is overproduced. it somehow interferes with the docking event. and the translation or transcription of gp23 is inhib- ited. accounting for the cis efl‘ect of gph't on the expression of gene 23. Also. according to this model. T4 go! mutations change the sequence of the go! site so that it no longer interacts as strongly with gp/i'r. thus averting the more permanent block of gp23 synthesis. This interpretation is KAO AND SNYDER consistent with all our evidence so far on the gol-gph‘r interaction (6. 7. 12: Bergsland and Snyder. submitted). Much more difficult to explain. however. is how this inter- action causes the observed trarrs inhibition of the expression of other genes. Perhaps the arrested gene 23 transcriptional- translational complexes somehow induce a trans inhibitor of all gene expression. The normal function of this rrarrs inhibition may be to reduce the expression of other T4 late genes and therefore delay the progress of the infection. if T4 heads are not being assembled properly. There are other known situations in which gene products encoded by two different phages or by a phage and a plasmid interact to cause a general efi‘ect on gene expression and/or a general shutdown‘of cellular functions (2. 19. 28). In at least one such instance. the rent and rexB genes of phage lambda may encode membrane proteins (5). The rex genes of A are also expressed in the lysogenic state. and although they normally inhibit only rII mutants of T4. they will even inhibit wild-type T4 if overproduced (24). Because of their similar- ities. it seems possible that gplir of e14 and other proteins such as the rexA and/or rexB gene products of A have a common function. ACKNOWLEDGMENTS We thank Kristin Bergsland for helpful discussions. Charles Hill for generously providing the e14 subclones. Jerry Dodgson and members of his laboratory for materials and advice in Maxam- Gilbert sequencing. and Robert Brubaker and members of his laboratory for advice on separating membrane fractions with isopyc- nic sucrose gradients. C.K. acknowledges support from a University Cellular and Mo- lecular Biology assistantship and the Rosenberg fellowship. This work was supported by Public Health Service grant GM-28001 from the National Institutes of Health and by National Science Founda- tion grant DM8617142. LITERATURE CITED 1. Alba. 1].. S. Adhya. and B. deCrornbrugge. 1981. Evidence for two functional gal promoters in intact Escherichia coli. J. Biol. Chem. 256:11905-11910. ' . Benzer. S. 1955. Fine structure of a genetic region in bacterio- phage. Proc. Natl. Acad. Sci. USA 41:344-354. 3. Black. L.. and M. Shove. 1983. Morphogenesis of the T4 head. p. 219-245. In C. K. Mathews. E. M. Kutter. G. Mosig. and P. B. Berget (ed.). Bacteriophage T4. American Society for Microbiology. Washington. D.C. 4. Brody. It. A. Greener. and C. Hlll. 1985. Excision and reinte- gration of the Escherichia coli chromosomal element e14. J. Bacteriol. 161:1112-1117. . Campbell. J. H.. and B. G. Rolfe. 1975. Evidence for a dual control of the initiation of host-cell lysis caused by phage lambda. Mol. Gen. Genet. [55:1-8. 6. Char-onus. W.. and L. Snyder. 1982. The gal site: a cis-acting bacteriophage T4 regulatory region that can affect expression of all the T4 late genes. J. Mol. Biol. [55:395—407. 7. Champness. W.. and L. Snyder. 1984. BacteriOphage T4 go! site: sequence analysis and effects of the site on plasmid transforma- tion. J. Virol. 50:555-562. 8. Cohen. S. N.. A. C. Y. Chang. and L. Hsu. 1972. Nonchromo— somal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. USA 69:2110—2114. OJ 'A 88 9. 10. 11. 12. 13. 14. 15. 16. 17. J. BACTERIOL. Cooley. W. C.. K. Sirotkin. R. Green. and L. Snyder. 1979. A new gene of Escherichia coli K-12 whose product participates in T4 bacteriophage late gene expression: interaction with the T4-induced polynucleotide 5'-kinase 3’-phosphatase. J. Bacte- riol. 140:83-91. Finlay. B. L.. L. Frost. and W. Parachych. 1986. Origin of transfer of IncF plasmids and nucleotide sequence of the type II oriT. rraM. and (MY alleles from colB4-K98 and type IV rraY alleles from R100-1. J. Bacteriol. 168:132—139. Greener. A., and C. Hill. 1980. Identification of a novel genetic element in Escherichia coli K-12. J. Bacteriol. [44:312-321. Kao. C.. E. Gumbo. and L. Snyder. 1987. Cloning and charac- terization of the Escherichia coli lit gene. which blocks bacte- ri0phage T4 late gene expression. J. Bacteriol. 169: 1232—1238. Laemmli. U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685. Lin. L. S.. Y. J. Kim. and R. J. Meyers. 1987. The 20 bp directly repeated DNA sequence of broad host range plasmid R1162 exerts incompatibility in vivo and inhibits R1162 DNA replica- tion in vitro. Mol. Gen. Genet. 208:390—397. Maguin. E.. H. Brody. C. fill. and R. D’Ari. 1986. SOS- associated division inhibition gene sfiC is part of an excisable element e14 in Escherichia coli. J. Bacteriol. [68:464—466. Maloy. S. R., and W. D. Nunn. 1981. Selection for loss of tetracycline resistance in Escherichia coli. J. Bacteriol. 145: 1110-1112. Maniatls. T.. E. F. Fritsch. and J. Sambroolt. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory. Cold Spring Harbor. N.Y. . Mann. A.. and W. Gilbert. 1977. A new method for sequencing DNA. Proc. Natl. Acad. Sci. USA 74:560-564. . Miller. 1.. and M. Malamy. 1984. Regulation of the F-factor prf operon: pifO. a site required in cis for autoregulation. titrates the wit product in trans. J. Bacteriol. [60:192-198. . Miler. J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory. Cold Spring Harbor. N.Y. . Miller. V. L.. R. K. Taylor. and J. Mekalanoe. 1987. Cholera toxin transcriptional activator tarR is a transmembrane DNA binding protein. Cell 48:271-279. . Sancar. A., R. Wharton. S. Seltzer. G. Kacinskl. N. Clarke. and w. D. Rupp. 1981. Identification of the uvrA gene product. J. Mol. Biol. 148:45-62. . Sanger. F.. S. Nicklen. and A. R. Coulson. 1977. DNA sequenc- ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. . Shinedllng. S.. D. Farms. and L. Gold. 1987. Wildtype bacte- riophage T4 is restricted by the lambda rex genes. J. Virol. 61:3790-3794. . Southern. E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503—517. . Stuart. P.. M. Ito. C. Stewart. and S. Conrad. 1985. Induction of cellular thymidine kinase occurs at the mRN A level. Mol. Cell. Biol. 5:1490—1497. . Straley. S. C.. and R. R. Brubaker. 1981. Cytoplasmic and membrane proteins of yersiniae cultivated under conditions simulating intracellular environment. Proc. Natl. Acad. Sci. USA 78:1224—1228. . Toothman. P.. and l. Herskowitz. 1979. Rexcdependent exclu- sion ot‘lambdoid phages. I. Prophage requirement for exclusion. Virology 102:133-146. . Yanisch-Perron. C.. J. Vieira. and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide se- quences ot'the M13mpl8 and pUC19 vectors. Gene 33:103-119. Appendix Interaction betassa the lit gene and the 321 £122 12221225 the as; EEA The work of Champness and Snyder (1982) and Bergsland and Snyder (Submitted for publication) suggests that the lit gene of g; 9911 and the 321+ site of bacteriophage T4 interact at two levels. There is a cis-effect which inhibits the expression of the gene containing the 521+ sequence, and a trans-effect which inhibits all gene expression in the cell. Bergsland and Snyder have evidence that the trans-effect requires the translation of the gol+ site. This result was obtained from a very elegant experiment with two clones that containtthe ggl+ sequence fused in the 1592 gene within the pUC12 cloning vector. One of the clones was translated in the correct reading frame for both the gol-encoded sequence and the 1392 sequence. The second clone, pUC12P21, was cloned with the ggl sequence in the -1 reading frame so that the peptide through the ggl region was drastically changed.The reading frame through the 1222 sequence was not altered in this clone. Bergsland and Snyder found that the clone translated in the -1 frame transformed Lit(Con) g; 39;; while the clone in the right frame did not. The interpretation from this experiment was that in-frame reading of the ggl sequence is required for the inhibition of transformation into a Lit(Con) cell. The inability to transform a Lit(Con) cell is due to the trans inhibition of gene expression. To determine if the correct amino acid sequence is 89 90 Table 4. Colors of pUClZPZl-transformed Lit(Con) and LitO cells on plates containing IPTG and X-Gal Color of cells transformed by: Genotype pUClZ 'pUCIZP21(30C) pUC12PZl(37C) JMlOl, 1150 Blue Blue Blue JMlOl, 1150, pACYC184 Blue Blue Blue JMlOl, ligo, pAElPAC Blue White Faint blue JMlOl, lit6, Blue White Faint blue 91 required for the cis-effect, I transformed pUC12PZl or control plasmids into lit-6 and lito JMlOl strains. An active beta-galactosidase enzyme complex will only be formed if transcription and translation continue through the ggl sequence into the downstream lggZ gene when grown on plates with IPTG and X-gal. At 30 ° c, the pUClZPZl-transformed JMlOl ligé strain was white, while the pUClZPZl-transformed IH 150 strain was blue (Table 4). The total number of transformants was comparable. The same experiment was performed in a JMlOl lito strain that either contained the plasmid pACYCl84 or the li£+ gene cloned in pACYClB4. The subclone containing the li£+ gene is phenotypically Lit(Con) at 30 °C. The color of the transformants at different temperatures was also informative. The pUClZPZl-transformed Lit(Con) cells were noticeably bluer at 37 06 than at 30 °C. This observation is in agreement with published reports that the Lit(Con) phenotype is more apparent at temperatures under 34 0C (Cooley et alul979; Chapter 2, this volume). Based on these results, we propose that the liEngl interaction involves the RNA of the 3331 site. The peptide sequence through the g9; site is completely altered in pUC12PZl, but the RNA is essentially the same. Since the lit protein can still prevent the expression of the downstream lggz gene in pUC12PZl. the cis effect must not have been abolished in this construct. The trans effect has beeml abolished since the pUC12PZl clone transforms Lit(Con) 92 cells as well as LitO cells. The cis interaction between gplig and the ggl RNA could prevent the translation of the fusion gene by two different mechanisms. The interaction could cause a transcriptional termination of the EQlLlEEZ transcript, or could prevent the translation of the message by preventing the ribosomes from reading past the hypothesized gpliE-ggl RNA complex. Iam currently trying to determine if transcriptional termination occurs at or near the ggl site by both biochemical and genetic assays. Biochemicallyq I have examined whether truncated Gol-LacZ messages are visible in Northern blot analysis of Lit(Con) cells harboring the pUC12PZl plasmid. Preliminary Northern analysis could detect neither a decrease in the transcription of the fusion message nor the presence of truncated messages, suggesting that translation is blocked. I am also assaying the activity of a promoter-less reporter gene, chloramphenicol acetyltransferase (CAT), fused downstream of the ggl site in pUClZPZl. iLf the lit-g3; interaction. causes a. transcriptional termination, the CAT gene should not be expressed. This construct in a Lit(Con) genetic background does produce at least enough CAT activity to confer chloramphenicol resistance to host cells, suggesting that there is no transcriptional termination at the g2; site when gp_lig interacts with the 321 site. Since the Northern data agrees with the CAT-fusion experiment, I believe that the cis-effect does not cause transcriptional termination. 93 However, further experiments, such as quantifying the amount of CAT-activity in Lit(Con) and LitO cells, and eliminating the possibility that there are cryptic promoters servicing the CAT gene, need to be performed before other interpretations can be ruled out. Literature cited: Champness, N., and L. Snyder. 1982. The g3; site: a cis- acting bacteriophage T4 regulatory region that can affect expression of all the T4 late genes. J. Mol. Biol. 155:395- 407. Cooley, W., K. Sirotkin, R. Green, and L. Snyder. 1979. A new gene of Escherichia coli K12 whose product participates in T4 bacteriophage late gene expression: interaction with the T4-induced polynucleotide 5'-kinase, 3'-phosphatase. J. Bacterio. 140:83-91. # 94 Refining the attestsre 2f the Gel sesnente shish interacts sigh sslit- The maximum size of the gel site required to prevent transformation has been defined by Bergsland and Snyder (Submitted for publication) as a 75 base pair sequence. The RNA sequence within this 75 bases contains a potential hairpin structure with six perfect base pairings in the stem of the hairpin. The structure is even more intriguing because several ggl mutations map in the stem of the hairpin (see Fig. 10). Bergsland and Snyder propose that gplit may bind to the hairpin» This binding may then cause the cis- inhibition of gene expression. I noticed that the potential RNA secondary structure of the ggl site bears considerable resemblence to the n It: In antitermination signal of bacteriophage lambda. The 3 Ir: In site contains at least two conserved areas (see literature survey). There is a well conserved nine base sequence named ngA followed by a hairpin structure which is less conserved in sequence among different lambdoid phages. Three bases upstream of the hairpin structure of the ggl site is a sequence that has considerable homology to the ngA sequence of the 935 site. The putative ngA sequence in the ggl site has perfect identity in six of nine bases to the lambda ngA sequence (fig 10). One of the nonidentity bases in the Egg site, the C at nucleotide 9, has been changed to an A without affecting antitermination (Robledo et a1” 1987). In total, at least seven of the bases within the gp__l 95 Figure 10. The potential secondary structure of the ggl site and its homology with the ngA sequence of the lambda 235 site. The RNA sequence of the lambda BEE site (Morgan, 1986) is presented in the upper line, and the ggl RNA sequence on the lower line. Nucleotides in the 321 sequence that have identity with ngA nucleotides are marked by filled-in circles. The open circle represents a conservative nucleotide change from the analogous base in 25A. Nucleotide 9 of ngA is present in some, but not all, IU’ lambdoid ngA sites. The open reading frame in the g2; site is marked by an arrow under nucleotide 3 to 5. Positions of known ggl mutations are marked by stars. 96 : 59.1 F p. P» mbda P 0 rgan, RNA (3 C P- lence G ( I Fm «NSGCUCUUUAC 511' C. O 0.08 sit, "UICGAUAUUUJaUGGU —p * M.A---- (29:90,.“ ‘ a “C34 *4 203 U“ C 34 C Fi9.‘|0 97 site's ngA sequence can theoretically function as a ngA sequence in lambda antitermination. Another interesting observation is that all of the ggl mutations sequenced thus far, except one, map either in the putative ngA-like sequence or in the the hairpin structure. In order to determine whether the hairpin structure is required for the cis interaction between lit and ggl. I made oligonucleotide site-directed mutations in the hairpin. The first mutagenesis scheme was designed to disrupt the stem of the hairpin without affecting the amino acid code by changing the wobble base of the'valine codon from anIJto an A at nucleotide 17. If this change allows transformation into Lit(Con) cells and prevents the cis-effect, then we would have a ggl mutation that does not affect the protein structure, lending credencetu>the hypothesis that the RNA is required for the liEngl interaction. The mutagenesis.of the‘ggl+ sequence cloned in M13 was performed by the protocol of Kunkel (in Ausubel et a1” 1987) using a synthetic oligonucleotide purchased from Operon Technologies Inc. Twenty-five randomly chosen potential mutants were sequenced and two were found to contain the proper desired mutation. I then cloned the mutated ngI-fliggIII insert fragment out of the replicative, double-stranded, form of the M13 phage, into the 1532 gene in pUC84, which has the B59111 site in th polylinker filled in. Plasmids containing the mutated insert were identifiable since the insert restores the reading frame with respect to 98 1322 and the transformants are blue. The mutant plasmid was then transformed into JMlOl ligO which also contained either the plasmid pACYC184 or the li£+ insert cloned in pACYClB4. The two site-directed mutants which theoretically disrupted the stem of the hairpin transformed like Col+ plasmids. They transformed 1150 cells, but did not transform Lit(Con) cells. Thus, the stem of the hairpin does not appear to be required for the li£;ggl interaction. Presumably it is the sequence of this structure and not the ability to form a hairpin which is essential for interaction with split- To determine further whether the sequence or the hairpin structure is required for 601+ activity, I used site-directed mutagenesis to reform a stem structure in a M13 clone that already contains a T to C ggl mutation at nucleotide 25 (see Fig. 10). I mutated nucleotide 19 from an A to a G. The double ggl mutant was found by probing potential mutant M13 lysates'boundIu)nitrocellulose with 32P-end-1abeled synthetic oligonucleotide, then identifying the more promising clone by successive washes at higher temperatures. One clone out of six sequenced contained the desired mutation. When the g2; insert was recloned in frame with 1532 in pUC84, and the plasmid used to transform l_£(Con) and ligo g; 2211, the double mutant did not transform as a 601+ plasmid even though the stem of the hairpin was restored. This result further suggest that the hairpin structure is not sufficient for interaction with 8? CE II 99 gplig. There is, instead, an absolute requirement for certain sequences in this area of the ggl site. The observation that the hairpin structure is not required for the lit-g2; interaction raises doubts as to whether the ggl site is functionally similar to the 925 site of lambda. However, it should be noted that careful mutagenesis of the hairpin structure in the gut site has not been done, and it is possible that lambda does not require its hairpin. The ggl-lig interaction does not appear to require the protein factors necessary for lambda N-mediated antitermination. Some factors required for lambda antitermination are gugA, EggB, and EggE. We obtained these mutants from David Friedman, and tested their effect on the lit-321 interaction. I transformed these strain with a lit subclone, which makes the cell Lit(Con) in phenotype, and plated T4 or gng4 phage on these strains. No effect on plaque number or plaque morphology was observed. Also, the presence or absence of gplig in these cells had no effect on the plating of lambda phages. It may not be too surprising that the lit-521 interaction does not require these specific Egg factors, since every antitermination system may use specific Egg factors (see literature review). Literature cited: Ausubel F. H.. et. a1. 1987- £22222t. 222222212 i2 221222122 biglggy. Wiley and Sons Publishers, New York. Robledo, R., M. E. Gottesman, and R. A. Weissberg. 1987. Lambda 333R mutations affecting lambda pN and HK022 pNUN activity- In 92.1.2 finrins 222222 2121222122 22222222 25. Bacteria and their Phaggg. Cold Spring Harbor, New York. 100 222 2222322 22222222222 222222 2 22222222 2222222222 22 translation. Since bacteriophage T4 has many gene products that affect g; ggli gene expression, it was always difficult to distinguish the direct consequence of the interaction between the lit gene and the ggl: site. The recent cloning of the lit gene and the 601+ and Col mutant sequences in the 1932 gene in pUC plasmids (Bergsland and Snyder, submitted for publication) made analysis of this interaction possible. In addition, the cloning of the 321 site under the lag promoter has other advantages, including the ability to tightly repress gene expression from the lag promoter with an overproduction of Lac repressors (due to lq mutations) in the cell. We can then examine the immediate consequence of the lit-g2; interaction after inducing the expression of the ggl site with the inducer, IPTG. To examine this interaction, I first transformed W3110 with an lq mutation (gift from Dr. Panayotatos) to chloramphenicol resistance with a li£+ subclone in pACYC184, pAElPAC. This plasmid produces enough of gplig to prevent T4 plaque formation at 37 ° C. I then transformed these cells to ampicillin resistance with the 601+ subclone, pUC84PZl, or the 601- subclone, pUC84PGl, (see Bergsland and Snyder, submitted for publication), selecting for transformants resistant to both ampicillin (amp) and chloramphenicol (cm). When the Lit(Con) cells containing Gol+ and Colo plasmids were streaked on plates containing IPTG, they did 101 not form visible colonies (see Table 5}. This experiment demonstrated that the lag repression due to the Iq mutation in W3110 can sufficiently repress expression of the 521+ sequence to allow a 601+ plasmid to exist in a Lit(Con) cell. I then examined.the effects of the li£;ggl interaction on the kinetics of cell growth (fig 11). In these experiments, the various strains were grown in tryptone broth in the presence of Cm (20 ug/ml) and Amp (25ug/ml) and the optical density was monitored by a dual-beam spectrophotometer over time. At 50' after the first measurement, IPTG was added to 1.0 mM concentration to all cultures. There was an obvious, almost immediate, cessation of growth in cells containing both Lit(Con) and 601+ plasmids. Furthermore, when the culture induced for two hours was plated onto plates containing Cm, Amp, and IPTG, the number of colony forming units was decreased at least 6 logs from the number of colonies present at the beginning of the experiment (Table 5L.No effect was observed in the strains containing the pUC12 plasmid. In the cells harboring the Col mutant plasmid, the growth rate had noticeably slowed after IPTG induction, but the culture continued to increase in density over time. The decrease in the rate of growth in the 001-, Lit(Con) cells is not surprising, considering that ggl mutations do not totally alleviate the interaction with the lit protein (see below). In summary, this growth experiment clearly demonstrated that 102 .Uocaauouec uoz .m .ouscwa 0 us ucomoun Duo mo ewauseouem ecu ou poumanvmoum mucosa: any .onH usonuwa waves :0 coueaa can zuoun ocoummuu Sous poses: amnu .eawu we assess mousse on» you Azaa ouv onH an“: pecans“ owes mHHeo on» uceauuonxo munu CH .N onH usonuua use no“: canoe co mafia: wcaauom hcoaoo .H m N a m Hos x n «ca 2 me.H A-HoucHNMWmmww ass «ma mos x «2.2 mos x s~.H A+Hoovsomsmo=a az maz was 2 mo.H mos x «o.H «Home ----E-..cm:::umvm ........ 2.. ........ 3--------M_.:.:..-m?? “Naoauoaecfi oamH uuumm ease ammo mascauucoo ofiams .mUHEmmHn «Hung uo .-Hou .+Hoo wcuuonumn mcwmuum Afioovuug cm OHHmB cw ~m>w>u5m same so GOwuospSH onH mo muoommm .n MAQ