. "710 '0 I O- n.“ 3-17:- v . . ‘vr ‘Tvvq’ ‘1 .TY‘v‘r'("=_ - $V'.'."'3 'VTVV‘V‘W I ._ .aA-Agg.‘ :‘g q-T‘. 1 L.... V .9 o" .A;. -...-.-.- w... ',’ I ." own».- - \ 91009-- 90' “4194.90Iolow-p.99,_m” 6"?”“0300096 - v "9 9mavneozzm-«uu ' 94.1%”! I” . . . . . . . . . . .. . . ; . . . o ... . . . . o . . . . . _ . _ . I . . . . 0. . . a .. ... .. 0 0.. .7. . . _. .I ..II .. . A . . I . . II. . .. . t . . 0 A . . . I Ilrdtlrcbdi-olfll It. .’,0. . . . I. . a ... II . . , . . . s. I .y . .+ ... .I o :1 r. . c . . . . . . _ 4 to. .IIIICIIrfl a . a s . . . I. _ a . - . I. . . , _ I . I . .. . . . . . . , _ . . . . .o. I‘m-ta “ . . o .u . .. .. o c 0 . t . . O:OS§;’I . I A O I. ‘ :0 0 . . s I. . . I. . . .. . . . . a g . . . . . . I I . . 0 u 0 I . I . V. . - o .. a a _ . v. . . T. . O \ . . c o I . g 0 . I o. . . _ ‘ . I b. . o . I a A. . . c .00 . . . I . I . . n 0 . a I . I . 1 o I 0. .. o . . o . . . . a 0. . . .0 I. o 0 v . v . . I I . . 0 I . l .v . . . . .. o .. _o . . O o . I v a . . . a u . . . . I l o. . . . . o . I . . c. .o . 6 . . . . . . a . I . o. n . 0 5 .. . _ . C . 0 . . . . _ . . 0 . u m. o . . . . . : I . a c I . . .. . a. .IIO'Dmolni. 30.0.3.0 . . . ... o . . . . .... . . . . ... . 21...... :. $.53?! 5...}. . v. . . . . I. . . .. II; .29.... 3'0. I . .. I . I O . . 0.0 . _ . . a . . . . . 2. o .I r a . . . . . . 0 - . u . I. 1 . mi . 0 . 0 u 0 ' I ‘ O I ’ 0 . . c ._ U . . . .4. I v . t I 0 I .l . . O O I - I n - o . . I . IV 0 . .o a . . . . v - . . . . . .. x ‘ . :0 . . I 9 I u . A . 00 . u . 1. . 0 c. . .0 - _ . . . . . . . . . . . - . ._ . . v.0: . . c A , . . . 6 - . . . . a I . . . o In . . . . . . . I v . . . . . .. I. o 0 n .‘ I0¢0I.i..I I... . . q .A . ‘ . . I O .6 . I . . K.~O.v.. 2...}... Ta... . It". .0. . . . . . .o I T . . . o n I . m 0 .. v n . . u . Y 0 O I . I "O”; u y! _ t | . u . o . _ _ . . ...I. o. .4 . .. . 0.. IIPI4:91. .. ‘v . . . . . . .ui 9 . u . I . . . I. .II . .. a... Guava: .nr 0 II 9.0a . n I . . . I . u r . . I . 0 . ..0 (.9 o. I 0.... 00.-.w‘.v . , . . o‘I..0308.00 4» .- . . . _ . . a . I I . . o . . 0 . 053.0. . $2.35.“. 5 . . v . I . o I I o .‘I O 3‘...- .. ‘a . . . .... . I . 0 0o . 00 U. U . . I o yin-yr. 4 .I I .. . I .0 . p . c . 40 .I 0. o 0 .n O . 6-. .o . I O. . . .. . . . .. . o .. .. I 9 I . . . . ... . . . .. u. a.. .. I 6 .. oma$lgln . . . u . In . . . . s ...009 . . .00I.III.0I.'0 I..Ir . . _ . . .. I ... . 2... . . .. .I a. I .‘v. ..II _ .Ié-npnflnbvi. “ . u . . I . o A.0 . I ..Q..I. - .. n0 3.... .207..I.... ...I I... :0: . I.ab.IOI’-l.) .(I o I . _ . I I. . a . . u .I . . I o . . . .. I .\I o. It. P... I . Pf: ..P~S0.¢.o8 3.30.0. I 0):.“ . I . . . . ._ . . . . V. I . . I I I I .. of... .. ..I':I . . of, » .II Izbzouazo...‘ . . . _ . I i . 0 . . I I . I I . 0o. . . . Pf‘IIOpo 0.I~0’ 0‘00iou'os‘i. 0. ~17 . . . . a 4 . 4 . . 6.. . . I . . . _ It; I. 2.1-1 0...... I.oI_I I .II: '30-} 1.. 0:05... 9.! .l . _ . I . _ a .A I ‘ . o 0 . .a .0... hot. I. o .01.. II. V . ia-..bs:\3 . ..l... . . . o a .. . . . v2.00 :1... c .7 I 302.! I II :2Ir .I 0 . a . . . I I . . 0. 0... 3. I II. (I .‘.III. ..-.v-.A..'. .)0.OI . u... r. A . . . . . . . a I II . . . 1 0.. , o. f. .o’gt0.0¢é.vr0.,i. .P...’ 3... 0.. I“ u . _. .. . I . I . . . . . n i... . .c .l. 0.0.... . 0' .v . “91'..._.;b.3 0 . .o . . . . . .. I . A . . . . - I . . .Ifiizpofzoc i}. .0, , .o‘ . I o. I I o . . .I _. . 0. . IAII"‘.00,OOC. 1‘ . . . . . . . o n .9 . o . .. . . . . .II .J . .If’a31! 3. In! . o . g I I . o . . . . . . ~ I. . . 0 . . u 0 .. 0.9 I: 13". I o . 0.0.. I I. «:0g50. . .0 . . . . . . . . . o 273-037... ... 2o . . . o u . .. . . . . w . .qI I ict‘bi It I u . n . . a; 4 cl.- .I~.-a. . . . . . . . . . . pg .99... 0‘ . . . . I I 9. I: QH . .I . . . . A a . X . . . . .. . I 1 ooh-0.3.. .* . ~ . . o o I c . . I30: IIIOIQI. . . _ - o _ I... I . . . . . . . I.. o»;€n.nl’..9..i . .o I . 9 0... n o .. 05.! €er 3' 3- y . .. . a . o 0 . a I ..I :0... 3.03:"? .I.‘ . . . _ . . I . . o I . .II .‘ I . . . . A .u I. _.Io. .I .I‘OIIX: .00 . I .I . I... 0.3... I h.. :I' . .0 . . I I . I . . . ‘3 I . I L... Inc.-. r...oob . . | . 0 so . . . :0 . 0.0. . u o . ._ . . .. . . . o , . _ . . . . I . I 0 . . I I . . I. .. .. . t . I .. o. . O . . . . I .. .~ I. . v o 0 .V 0 . . . . . . O 0 I I c O . 0. . I . I I . I .. 0 I0 _ . . . . . . . . . r .. . 0 . . A . n O . O C .u I 0 .d I I o . c . . . . O I . o O I u . . C . I I. o. . I4 I . . I I.. I I I I .0. I I . . . 0 ~ 0 D . 0 .0 .0 0. . 0 o . . I 0 .0 . . . . o . . ¢ o I... I .0 . . . 0 .0 . Q Q . 0. . ‘ if... ‘. .. . . . .yo. I r n I I .. .. . .c. . . 000...:0..I . It: . 0 I I . a . n c 3. . . . I . . . . no I. . I o . I . . I; I 6...? .o....p_o.o . ...-3.I.I.I.I.I. I. . . I. o . V I u I o. I... I I . o 1:55 5.3.0.!!9; I I‘ . I 0 I v I . O . . I I1 0". D i 03.. :- OI .' . .I 0 I. . .00 I:l¢!v..0..¢..~'. :OIIOI" I . ‘ . . . o ....OOI0P‘.P0.¢\:I$.. ‘0 0I . . I. I .30. c web... 3. . v0 0 O . I I a . .. .. . v. n ' 0.0I. : 9.0., O. I D .t . .0 . a. .. . . 0 o I .I ‘ n . 3! 0 . o I .3 O u . O . . L. . . . of . . .. .‘0I0‘ 000-0. . I II.II.- t 21-0 c.‘ 3.0:... :0. .0. .01.: [.0 . u. . I ~ I 0 .0. .I .. .02.. 90.. . o I III.- .II: .o‘c..I'Ia-o.;-IOI lo‘.V 0. I .00 I II. ._ . I To... 0 ,f‘.. I .34‘4 .o.n‘vOI.n I A a. | . 3.1. 3940:}... .II .000. I . 9 0. v a ‘ .‘to u . l I. c O. I I C l _ .. _ .c I 0:.- I\ I n...‘ I . o .. v I. 9 .90 s I... c I 0 4 .III I o . I. I .._ 96., n I . . .. I . . . .0 0-. . I u . I 0.. I. . 00. a I I. t . I'..'¢o!oo .......c... Ia! . . I. a. .r J.1r .‘ . 0.33.... .4331 I. . \«J. y I. . 9. . . 3. I .(O. . . .. _ .. . . . _ ,. 0?)... .y 0. . I. . : :10. 3.I .00 .0 ou': I. 4 . I . 0. . . In: 0 . o I... ..v). O u 0 ... .‘ a . . Z... I. I. I. A" o. n . I... . I . O I I. a I. .‘II. . . . II . v .00....00. .0o.o':£ ...I ~ .. . .OV . . . 3:15. ’3‘ o . 0 I o. M I . o... . . . . o I I . _ v I.. . _.. I“ u 3. .546...‘ I. 0- I. . l .4 I 0 _. . .u; 2 . o 0.. (out. __I 0 .I o.— . . o. o. I: . n l. n , .009. O o I .V o I. . I .0 o 0.1... .2 .0 A O I. I . L a. I 0 I n . o . I '0. i .00. .. m . .fI. . I I29: 7. . V . .. . o . . I I 0 I: .I II II. . . 0 out. ..n Ii... «2...... < 0 A I .0 It I . I I. . o. In a I 0. I .I 0‘ no . .IC 0 ...I ... .. .a . p 0 5.0 VI. . I 0 . .. t. . . '0... . I . .I goo-I... o. . . . .uA .- I... . If. I . o n .I an... . u I. . I00 . . A. 4 v... . . _I'0 Idovl. . a. .. c I .I. . . 0 0 I I I. \’00 I I. o. 0’. .0 o. .ooIIOIOOI... III!..0.AI.. I’I-.09.0-7. .i I . a. s.... 00.55.9919 .‘0.. . 8.3.0. . at. I... I... . 11‘s... o...:¢0.oa a. .06 . o 0 .I.‘-_4 5...... o. .3. III {co-55.33.10? . 0.. I...) I I a . I. A 2.... $13!: «I: . :10: f... . 1o.- . a. o I. o I . 0-. . .II I! I... Eueo: J0..0DJ.'0.I- ..I I; In I I . I ‘ 4 0 I. a? 0.90.03... . 0 .~ '. . .0 .0 o n. ,- . . . I. I 00-. .4 no. .PO.OI}V I. 0.4 0.). I). I o‘ I I.DI IV . . . I n . ‘ o . It. . 4 ... .o .3 5:5 .I‘.....; .to:...o.-II.« .9... .00.?r..- on... o . . I . .0... .o .9 2......... o. 3v.-. ’o..'........ 2.... v1 - I! i NUNHHIIHIHIUlHHHINJWIHIHHI”NHllHlHllUll W 3'72 1293 00781 5743 LIBRARY M'Ch'Qan State University This is to certify that the dissertation entitled Isolation and Characterization of the Escherichia coli K12 hemA Gene presented by Debra Elizabeth Verkamp has been accepted towards fulfillment of the requirements for Ph.D. degreein Microbiology and Public Health Major professor ( s/lo/‘c‘? MS U is an Affirmative Anion/Equal Opportunity Institution 0-12771 ISOLATION AND CHARACTERIZATION OF THE ESCHERICHIA COLI K12 HEMA GENE By Debra Elizabeth Verkamp A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1989 bo4quY ABSTRACT ISOLATION AND CHARACTERIZATION OF THE ESCHERICHIA COLI K12 HEMA GENE By Debra Elizabeth Verkamp The initial step in heme formation is the biosynthesis of 5-aninolevu1inic acid (ALA). Mutations which block ALA synthesis in Escherichia 9911 have been mapped to two loci, 22;; and 2229. The original characterization of the beg; mutation did not allow an unanbiguous assignment for the role of its gene product in ALA formation, nor was there biochenical evidence that defined the ALA biosynthetic pathway that operates in E. coli. Biosynthesis of ALA occurs in biological systens by either a C4 route (via the ALA synthase reaction, with succinyl CoA and glycine as starting substrates) or a C5 route (a multi-enzyme reaction sequence that transforms glutamate into ALA). It was generally assuned that the ALA synthase pathway operates in bacteria, including 5. ggli. In this dissertation, I describe the isolation and characterization of the beg; gene from E. ggli. Genetic analysis, gene replacement mutagenesis, and naxicell experiments identified the cloned gene as h2g5. The nucleotide and predicted amino acid sequences of hag; gene bore no resemblance to ALA synthase sequences cloned from other organisms and no ALA synthase activity was detected in g. 52;; extracts, including extracts prepared from strains carrying the cloned hem; gene on high copy number plasmids. These results suggested that the C4 route does not normally function in g. 92;; for the synthesis of ALA and that a re-evaluation of the role of the hggg_gene product in ALA formation was required. Included in this dissertation is a discussion of my results in relation to biochemical studies on ALA biosynthesis in 5. 92;; performed by other investigators. The combined studies support the conclusion that ALA synthesis normally occurs in g. 99;; by the C5 route and that the hem; gene encodes one of the components of this reaction sequence. Results of SI nuclease protection experiments showed that the hem; mRNA appeared to have two different 5’ ends and that a nonoverlapping divergent transcript was present upstream of the putative distal hemA transcriptional start site. to my parents iv ACKNOWLEDGMENTS I would like to acknowledge Barry Chelm for providing support, instruction, and encouragement. I feel very fortunate to have worked with Barry and to have been a member of the Chelm lab. Special thanks go to Larry Snyder for guidance and advisement during the latter stages of my research. I would also like to recognize the other members of my guidance committee, Mike Thomashow, Shelagh Ferguson-Miller, and Wendy Champness for advice and assistance. Thanks are also extended to Hans Kende and the PRL for providing financial support and to Tom Elliott for sharing results prior to publication. The various members of the lab that I worked with over several years are duly recognized for making the years of my graduate career always stimulating and occasionally enjoyable. John Scott-Craig, Mary Lou Guerinot, Todd Cotter, Bill Holben and Prudy Hall deserve special thanks for sharing ideas and expertise that helped me to complete various aspects of my research project and writing endeavors. TABLE OF CONTENTS List of Tables List of Figures CHAPTER 1 Introduction CHAPTER 2 Isolation and Characterization of the hemA Gene from Escherichia coli K12 Introduction Materials and Methods Bacterial Strains Culture Media and Growth Conditions DNA Methods Gene Replacement Mutagenesis Cosmid Libraries E2131 Deletions and DNA Sequencing Maxicell Experiments Enzyme Assays Results Isolation of the hemA Gene DNA Sequence Analysis Maxicell Experiments Construction of hemA Mutant by Gene Replacement Pl Transductions vi PAGE viii ix 15 16 16 16 18 18 19 20 21 22 22 22 26 31 36 42 ALA Synthase Assays Discussion CHAPTER 3- Fine Structure Analysis of the Escherichia coli hemA Transcriptional Unit Introduction Materials and Methods E. 331; Strains and Growth Conditions RNA Isolation Sl Nuclease Mapping Results Discussion CHAPTER 4 Summary and Conclusions REFERENCES vii PAGE 42 43 49 51 51 51 51 53 63 69 77 LIST OF TABLES TABLE PAGE 1 E. coli strains 17 viii CHAPTER FIGURE 1 1 LIST OF FIGURES PAGE The biosynthetic pathway 2 from ALA to protoporphyrin Ix Postulated biosynthetic pathway 11 of ALA formation from glutamate and ALA synthesis via the ALA synthase reaction Restriction endonuclease map 25 of plasmid subclones of the isolated hemA region Southern hybridization analysis 28 of cosmids pMRCS, pMRClO, and genomic DNA Partial nucleotide and 29 predicted amino acid sequence of ORFC Nucleotide and predicted 30 amino acid sequence of ORFD Nucleotide sequence of the 33 hemA gene and the predicted amino acid sequence of the hemA polypeptide Autoradiogram of plasmid 35 polypeptides labeled in the maxicell procedure and schematic depiction of pMR81 and pMRELABgl Inserts of plasmids used for 39 site-directed mutagenesis Southern hybridization analysis 41 of EVBI genomic DNA ix CHAPTER FIGURE PAGE 3 1 Nucleotide sequence of the 5’ 55 end and the flanking upstream region of the E. coli hemA gene 3 2 31 nuclease protection analysis 57 3 3 SI nuclease protection mapping 60 of the ORFD transcriptional start site 3 4 I High resolution 81 nuclease 62 mapping of the hemA transcript 3 5 Promoter sequence comparisons 64 3 6 Nucleotide sequence of the 67 hemA-prfA intergenic region CHAPTER 1 Introduction Tetrapyrrole derivatives are distributed throughout nature and play a central role in the biology of both prokaryotic and eukaryotic organisms. With rare exceptions, members of this diverse group of compounds are present in all organisms and perform a variety of vital physiological functions. A schematic depiction of the tetrapyrrole biosynthetic pathway by which 5-aminolevulinic acid (ALA) is converted to protoporphyrin IX is presented in Figure 1. Excluding the formation of ALA, which is the first committed intermediate unique to the tetrapyrrole biosynthetic pathway, the enzymatic reactions that comprise this pathway appear to be conserved in biological systems. (For a review of tetrapyrrole synthesis and regulation, see Granick and Beale, 1978.) Most biologically—active tetrapyrroles exist as metal complexes and can be classified according to the metal which is associated with the porphyrin nucleus. Iron is inserted into the protoporphyrin IX ring to form heme and the insertion of magnesium produces the biochemical precursor of chlorophyll and bacteriochlorophyll in another branch of the pathway. Corrins (including vitamin B12) constitute a class of tetrapyrroles that is complexed to .x. 5.392.305 0. <4< E0: 3352. 03055303 0:... . w 052“. :. «u- .31 L in. . fiesuoetaeeoasea :00w :000 28¢ :00 «:w «:0 «:w «x. «:0 «:0 «:0 a: 9:0 z000~z0~z 0:0 «1w .3. :000 60005300898 30. 25.33.05.3(6 « a «:2 :0000: x :2 u _ u a z z :0... : «zw 0 w — " All All “ d\ All 0 u 0 a w I a N «:0000: a . :0 a w 2w z . x 2000 :000 w :000 :00 3 cobalt and diverges from the common pathway at the point of uroporphyrinogen III synthesis. The regulatory aspects of this multi-branched pathway pose interesting biological questions. Requirements for tetrapyrroles vary among organisms and can be distinctly different between tissues of a single organism, depending on the physiological roles that they fulfill in a particular cell. Information about this regulation is rudimentary at present, even in the well-studied bacterium Escherichia 921;. Results of investigations into porphyrin synthesis and its regulation in 5. £21; are discussed below. Tetrapyrroles function in E. 3911 in the cellular processes of respiration and oxygen defense. As the prosthetic groups of cytochromes, hemes are essential for the operation of the electron transport chains required for cells to derive energy from non-fermentable carbohydrates (Poole and Ingledew, 1987). Name also forms the prosthetic group of catalase, an enzyme that converts the toxic aerobic metabolic by-product hydrogen peroxide into oxygen and water (Gottschalk, 1986). By this mechanism, catalase protects cells against the deleterious effects of reduced oxygen intermediates. Siroheme is an iron tetrapyrrole that is derived from uroporhyrinogen III (Jacobs, 1974) and forms the prosthetic groups of both sulfite reductase, an enzyme required for assimilatory sulfate reduction (Kredich, 1987) and nitrite reductase, which reduces nitrite to ammonia during anaerobic growth (Lin and Euritzkes, 1987). Research into the genetics and biochemistry of heme biosynthesis in 5. £21; dates back to the 1960’s. Heme-deficient strains of E. ggli_were isolated initially by investigators studying small-colony-forming mutants selected on the basis of their resistance to low levels of aminoglycoside antibiotics (Sasarman and Horodniceanu, 1967; Sasarman gt al. 1968a). These cells are resistant to aminoglycosides such as streptomycin presumably because they are unable to synthesize a functional respiratory chain and are unable to generate the transmembrane electron potential required for the uptake and accumulation of this class of antibiotics (Chopra and Ball, 1982). Upon further characterization, these small colony, respiratory-deficient mutants were determined to lack catalase and cytochromes because of their inability to synthesize heme. Advances in this area of investigation were somewhat hindered because of the impermeability of E. 92;; cells to all compounds in the heme biosynthetic pathway beyond ALA (Sasarman st 51., 1968a). Further progress was facilitated after McConville and Charles (1979c) isolated heme-permeable mutants that could be manipulated more easily for genetic analysis. Additional mutants deficient in heme biosynthesis were identified because the cells accumulated large quantities of the heme precursors synthesized prior to the block in the pathway (Cox and Charles, 1973; Powell gt 31., 1973; 5 1., 1975; McConville and Charles, 1979a; Sasarman gt McConville and Charles, 1979b). The mutations which blocked heme synthesis were characterized based on the pattern of precursor accumulation, the absence of enzyme activities, and the ability of intermediate compounds to rescue the mutant phenotype. Data compiled from these studies provided the genetic scheme for heme synthesis in E. ggtt summarized below. At least eight enzymatic steps are required for heme biosynthesis in E. ggtt. The associated genetic loci have been designated Eggt through EggE and correspond to the following enzyme activities: ALA synthase (EC 2.3.1.37) (Evidence provided by this dissertation and other reports argues strongly against this assignment), ALA dehydratase (EC 4.2.1.24), porphobilinogen deaminase (EC 4.3.1.8) (formerly uroporphyrinogen I synthase), uroporphyrinogen III cosynthase (EC 4.2.1.75), uroporphyrinogen decarboxylase (EC4.1.1.37), coproporphyrinogen III oxidase (EC 1.3.3.3), protoporphyrinogen oxidase, and ferrochelatase (EC 4.99.1.1) (see references in Bachmann, 1983). Genes involved in heme biosynthesis map to several scattered regions of the chromosome, and, with the exception of one apparent heme biosynthetic operon (see below), are not physically linked. Two other loci, 2229 and pggt, also appear to be involved in porphyrin synthesis, but have been less well characterized. 6 A systematic investigation into the regulation of heme biosynthesis has not yet been conducted, but it appears from several reports that this metabolic pathway may be controlled at multiple levels. It has been shown that wild-type strains of E. ggtt accumulate protoporphyrin and coproporphyrin when grown with excess ALA (Sasarman gt gt., 1975), suggesting that the formation of ALA may be a rate-limiting step in the pathway. However, since these cells did not accumulate heme, there may be additional control points later in the pathway, possibly at the ferrochelatase step at which iron is incorporated into the tetrapyrrole ring. In a study by Ishida and Hino (1972), the quantity of cytochromes and protoheme was shown to be several—fold higher in aerobically grown cells of E. ggtt as compared to anaerobically-grown cells, suggesting that some aspect of general oxygen control may contribute to the level of heme and hemoproteins in E. ggtt. In another study, Poulson gt gt. (1976) reported that protoporphyrinogen oxidase levels are subject to catabolite repression. In summary, while genes involved in heme biosynthesis have been mapped in E. ggtt and some investigations have been conducted into the regulation of heme formation, knowledge about this pathway has lagged behind that of other biosynthetic pathways. This is somewhat surprising in light of the importance of heme to cellular metabolism. In recent years, however, renewed interest in the genetics, 7 biochemistry, and regulation of heme biosynthesis in E. ggtt has resulted in substantial progress toward an understanding of this fundamental process. The second step in the heme pathway is catalyzed by ALA dehydratase (ALAD), coded for by the EggE gene. The reaction involves the condensation of two molecules of ALA to form porphobilinogen (PBG), with the removal of two molecules of water. The hemB gene has been isolated and sequenced (Echelard gt gt., 1988; Li gt 1., 1988; Li gt gt., 1989a) and the predicted amino acid sequence of the polypeptide displays extensive homologies with that of ALAD cloned from yeast (36*), human liver (40%), and rat liver (40*) (Li gt gt., 1989a). These studies also revealed a relationship between the activity of ALAD and the activity of the enzyme which catalyzes the third step in the pathway, porphobilinogen deaminase (PBG D), the product of the Eggg gene. Umanoff gt gt. (1988) discovered that EggE mutants (lacking ALAD activity) also have extremely low PBG D activity, and, furthermore, that hemA mutants are also deficient in PBG D activity. The activity of PEG D is restored to the hemB mutant by introducing the cloned hemB gene into the cells on a single-copy plasmid or by growing the mutant on media supplemented with PEG. Likewise, PEG D activity is restored to a Eggt_mutant if it is grown with exogenous supplements of ALA or PBG. Further experiments demonstrated that the availability of PEG probably controls the appearance of PEG D activity at some 8 posttranscriptional level, possibly by binding to the protein and acting as a prosthetic group to activate the enzyme or by protecting the enzyme from degradation. This is the first demonstration that the product of one gene in the pathway, ALAD, is required to synthesize a biochemical intermediate (PBG) that is necessary for the activity of a second enzyme in the pathway, PBG D. The results imply that a complex and intricate array of regulation may exist to control the activities of enzymes in this pathway. The Eggg and EggE genes, encoding the enzymes PBG D and uroporphyrinogen III synthase, respectively, have been isolated (Thomas and Jordan, 1986; Jordan gt gt., 1987; Sasarman gt gt., 1987; Jordan gt gt., 1988) and are constituents of the Uro operon. This operon is located at approximately 85 min on the E. ggtt linkage map, adjacent to and transcribed divergently with respect to the gxgt gene, which encodes adenylate cyclase (Sasarman gt gt,, 1987). Nucleotide sequence analysis revealed that the Eggg gene is located promoter-proximally and that the start codon of the EggE gene overlaps the terminal nucleotide of the Eggg coding sequence. Expression of EggE is dependent upon the Eggg promoter, but studies on transcriptional regulation of the Eggg EggE operon have not yet been published. At the third step of the heme pathway, PBG D catalyzes the head-to-tail polymerization of four molecules of the monopyrrole, PBG, into the linear tetrapyrrole intermediate, preuroporphyrinogen. Uroporphyrinogen III synthase rapidly converts this highly unstable intermediate to uroporphyrinogen III (Jordan gt gt., 1986), the first cyclic tetrapyrrole formed in the pathway to heme and all other biologically-active tetrapyrrole compounds. The result of the sequential and coordinate actions of these two enzymes is the formation of uroporphyrinogen III. In the absence of uroporphyrinogen III synthase or in the presence of excess PBG D, the nonphysiologic isomer uroporphyrinogen I is formed in a non-enzymatic reaction following the polymerization reaction catalyzed by PBG D (Jordan gt gt., 1988). The concerted expression of the hemC and hemD gene products, suggested by the physical arrangement of their coding sequences, may exist in order to ensure that their catalytic functions are coordinated and collaborative, analogous to a multi-enzyme complex. Two additional open reading frames, denoted Egg; and Egg! have been detected downstream of the EggE coding region (Alefounder gt gt., 1988; Sasarman gt gt., 1988). These open reading frames direct the synthesis of polypeptides (Aldea gt gt., 1988) that may be involved in heme biosynthesis. The EggE locus, which encodes protoporphyrinogen oxidase, has been mapped to this region of the E. ggtt chromosome (Bachmann, 1983) and is a possible candidate for one of these genes (Jordan gt gt., 1988). At this time, there have been no further studies on the hemE, hem? and hemH genes. 10 The primary focus of this dissertation is the molecular characterization of the Eggg gene in E. ggtt, which is required for the synthesis of ALA, the universal precursor of the common tetrapyrrole biosynthetic pathway. The synthesis of ALA occurs in biological systems via at least two different pathways. In one pathway, the condensation of succinyl-coenzyme A and glycine to form ALA is catalyzed by ALA synthase (Figure 2). This enzyme has been studied extensively in Rhodobacter sphaeroides and other photosynthetic bacteria (e.g., Burnham and Lascelles, 1963; Lascelles, 1964; Burnham, 197D) and in avian and mammalian tissues (e.g., Marver gt gt, 1966; Bottomly and Smithee, 1968). This pathway has also been demonstrated in other eukaryotic and prokaryotic organisms and the genes which encode ALA synthase have been isolated from the following sources: Rhizobium meliloti (Leong gt gt., 1982); Etgdxrhigobigg japonicum (Guerinot and Chelm, 1986); yeast (Keng gt gt., 1986); Rhodobacter sphaeroides (Tai gt gt., 1988); Rhodobacter capsulatus (Biel gt gt., 1988); chicken (Borthwick gt 1., 1985; Yamamoto gt gt., 1985; Maguire gt gt., 1986); mouse (Schoenhaut and Curtis, 1986); and human (Bawden gt gt., 1987). The enzyme displays significant evolutionary conservation across a broad spectrum of organisms as illustrated by the 48.8% amino acid identity observed between the predicted amino acid sequence of ALA synthase from the symbiotic nitrogen-fixing soil bacterium ll £03000. 000553 (.3 05 0... 0.00556 <._< 0:0 30805.0 Eo: 00:05.0. 44¢ .0 6.5500 02055303 00.0.3000 . N 050.“. 00.300.02.000 0_0< <4< . 70.060020 <25..>E0S.0 02:03.0 «xzuzw 0 u “.0... .. .0...sz zoom 0 1': w NIZT—w o H w Omfluflzac>m NT—ZIM—V N N N N 10 000.0.00030554‘ :0 00000005300 IZIU (25.38320 10 a . m a . A u . 402.9202... n. I... I... a. z. I... :000 :000 «rm :000 1000 000503 <4< Ioow <00.On_0 N:zuz0 «zw + a 05020 1% 1000 <00-_>:_00:m 12 Bradyrhizobium japonicum and from chicken embryonic liver (McClung gt gt., 1987). In a second, multi-enzymatic pathway, the intact 5-carbon skeleton of glutamate is converted into ALA. This route has been demonstrated in plant chloroplasts (Beale and Castelfranco, 1974; Beale gt gt., 1975; Beale, 1976), Chlamydomonas (Wang gt gt., 1984; Huang and Wang, 1986), 0 ('0' 1., 1988), algae (Neinstein and cyanobacteria (O’Neill Beale, 1985; Weinstein gt _t., 1987), an anaerobic archaebacterium, Methanobacterium thermoautotrophicum, (Gilles gt gt., 1983) and Euglena (Neinstein and Beale, 1983), which also has ALA synthase activity (Beale gt gt., 1981). The postulated biosynthetic sequence of catalytic reactions that transforms glutamate into ALA has been t a1., 1981; elucidated by several laboratories (Wang Kannangara gt gt., 1984; Wang gt gt., 1984; Huang and Wang, 1., 1988; Schneegurt and Beale, 1988) 1986; Kannangara gt and is depicted schematically in Figure 2. The glutamic acid is activated by the ligation of a tRNAlI“ in a reaction that is thought to be identical or analogous to the aminoacylation of tRNA catalyzed by glutamyl tRNA synthetase during protein synthesis. Following this activation at the a-carboxyl, a NADPH-dependent dehydrogenase catalyzes the reduction of this activated glutamate to glutamate- l-semialdehyde (GSA). In the final step of this pathway, glutamate-l-semialdehyde aminotransferase transfers the amino group to the terminal 13 carbon to yield ALA. This pathway has been studied extensively in barley chloroplasts, where the component enzymes of this ALA-synthesizing system are soluble and located in the stroma of plastids (Gough and Kannangara, 1977). The glutamic acid tRNA ligase and the glutamate-l-semialdehyde aminotransferase have been purified (Bruyant and Kannangara, 1987; Kannangara gt _t. 1988), but the dehydrogenase has not yet been purified to homogeneity. The tRNA!lu which participates in ALA synthesis has been sequenced from barley and is encoded by chloroplast DNA (Schon gt gt., 1986). Isolation of the genes which encode the component enzymes of this pathway has not yet been reported. At the time the series of experiments described in this dissertation was initiated, it was generally accepted that E. ggtt synthesized ALA via the ALA synthase reaction. However, reports of ALA synthase activity in E. ggtt have been inconsistent. While some investigators have failed completely to detect ALA synthase activity in tg gttgg enzyme assays (McConville and Charles, 1979b), others report values which range from 0.025 nmol/mg protein/hr (Schoenhaut and Curtis, 1986) to 65 nmol/mg protein/hr (Tai gt gt., 1988). Because of this disparity, ALA synthase activity in E. ggtt has remained controversial and a clarification of the role of the hemA gene product in ALA synthesis was warranted. l4 Mutations in E. ggtt that cause defects in ALA synthesis map to two different loci on the genome; the Eggt (Sasarman gt gt. 1968b) and 2229 (Powell, gt gt. 1973) loci at approximately 27 min and 4 min, respectively. Eggg mutants display a heme-deficient phenotype that can be rescued by exogenous ALA. However, the exact nature of the defect was not characterized and there was no definitive evidence that the Eggt gene encoded ALA synthase. The role of the 2229 gene product in ALA synthesis has likewise not yet been defined. The primary focus of this thesis project has been the isolation and characterization of the Eggg gene in E. ggtt, with the aim of defining the role of its gene product in ALA synthesis. Experiments described herein were also designed to characterize the transcriptional unit of the Eggg gene and to identify elements which might participate in the control of its expression. I also present the nucleotide sequences of additional open reading frames which have been identified in the region of the genome surrounding the hemA gene. CHAPTER 2 Isolation and Characterization of the hemA gene from Eschgrichia coli K12 Introduction Heme serves two major functions in the cellular metabolism of E. ggtt: respiration and protection from toxic oxygen metabolites. The initial step in heme biosynthesis is the formation of ALA, the first precursor unique to this pathway. Although mutations which result in ALA auxotrophy have been isolated (Eggg, gggE), the roles played by the gene products of the associated loci in the genetics and biochemistry of ALA production lack definition. The objective of experiments recorded in this chapter was the isolation and characterization of the Eggg gene from E. ggtt with the aim of clarifying its function in ALA biosynthesis. The cloning strategies and molecular genetic techniques used to isolate the Eggg gene from E. ggtt cosmid libraries are described herein along with the results of experiments that identified the cloned gene as Eggg. The nucleotide and amino acid sequences of the cloned E. coli hemA gene display no significant similarity to ALA synthase genes which have been isolated from other organisms. 15 16 Materials and Methods Bacterial Strains. E. coli strains used in this study are listed in Table l. Egtture Medig and Growth Conditiggg. E. ggtt strains were grown routinely on LB medium or M9 minimal medium (Davis, gt gt., 1980). Carbon sources were added as indicated at the following concentrations: glucose .010 E, acetate .020 E, succinate .016 E. Minimal medium was supplemented as required with L-amino acids at 40 ug/ml and thiamine at l ug/ml. Strains SASX4lB and EV61 (Table l) were grown routinely on medium supplemented with 50 ug/ml ALA. Antibiotics were incorporated into LB medium when indicated at the following concentrations: ampicillin 80 ug/ml, kanamycin 50 ug/ml and tetracycline 5 ug/ml. Anaerobic conditions were achieved by sparging the cultures with N2 in closed, stoppered bottles for 30 min. These cultures were then incubated at 37° in anaerobic GasPak jars (BBL Microbiology Systems). Aerobic cultures were grown in 2 liter flasks at 37°C with vigorous shaking. For ALA synthase assays, the appropriate E. ggtt strains were inoculated into 400 ml of M9 glucose, succinate or acetate medium with 4 ml of a mid-log culture grown in the same medium and then incubated aerobically or 17 Table 1. E. coli strains Strain Genotype Source of Reference CR63 sugD60 lamB63 Bachmann (1987) R3101 hstZO (rn‘,ms') recAl3 Boyer and leu-6 thi-l era—14 Roullane- proA2 lach galKZ Cussoix (1969) rpsL20 (Sm?) xyl-5 mtl-l supE44 DH5 endAl recAl Hanahan (1985) hstl7 (rx'mx’) supE44 tEi-l gyrA96 reltt? JC7623 thr-l are-14 leuBS Bachmann (1987) aigpt-proA)EE lach tax-33 supE44 galK2 rac‘ hisG4 rfbDl m 1-51 rpsL31 kdgK51 xyl-5 gtl-l argE3 thi-l recEEt recC22 gchlS gthZOl JM103 hst4 A(lac-pro)/F’traD36 Messing (1983) proAB lacIQZ M15 thi strA supE endA sch 0RN125 zcg::TntE thr-l leuB Spears gt _t. thi-1 AKargF-lac)U169 (1986) malAl xyl-7 are—13 mt1-2 gal-6 rpsL fhuA2 supE44 SASX4lB HfrPOZA hemA4l metBl relAl (B.Bschmann) EV61 Same as JC7623, except hemA This study MC1024 araD139 AKara-leu)7697 Casadaban AKlacZ)M15 galU galK and Cohen strA recA56 srlzzTntE (1980) EVIO Same as SASX41B, except This study recA56 srl::TntQ JK268 trpE trpA dade purB Radar gt gt. (same as J3266) (1976) 18 anaerobically. Cells were harvested at the mid-log phase of growth. Pl transductions were performed according to methods described by Miller (1972). EE_ Methods. Small scale plasmid isolation was performed by the method of Holmes and Guigley (1981) and large scale isolation was by CsCl ethidium bromide equilibrium centrifugation (Clewell and Helinski, 1972). Genomic DNA was extracted and purified using the method of Marmur and Doty (1962). Restriction enzyme digests were performed according to the suppliers’ specifications. Ligations and gel electrophoresis were done according to standard techniques (Maniatis, 1982). Restriction enzyme-cleaved DNA was size-fractionated by electrophoresis on 18 agarose gels, transferred to cellulose nitrate filters by the method of Southern (1975), and hybridized to 32P-labeled probes as described (Adams gt gt., 1984). Eggg Replacement Mutagenesis. A method described by Ninans gt _t. (1985) was used to construct a defined mutation in the Eggt gene. The entire Eggg gene and flanking sequences were present on a 6.0-kilobase pair (kbp) EtngII fragment that was cloned into pUC19 to create pMR57 (Figure l). Plasmid pMR6l is a derivative of pMR57 in which a 278-base pair (bp) EgtII fragment within the coding region of the hemA gene has been deleted and 19 replaced by a 1.3-kbp EggHI kanamycin resistance cassette derived from plasmid pRL161 (designated kanamycin cassette C.Kl in Elhai and Nolk, 1988). The ggttt gene is oriented in the same direction as that of the putative Eggt open reading frame (ORFA in Figure 1). Approximately 2 ug of this plasmid was digested with EtngII and transformed into E. ggtt strain JC7623 (Table 1). This strain can be transformed readily with linear fragments of DNA because of the loss of RecBCD nuclease activity, but is proficient for recombination due to a compensatory mutation in exonuclease I (gEgE). A marker for drug resistance flanked by E. ggtt DNA sequences transformed into JC7623 on a linear DNA fragment is maintained only after a double recombination event between the E. ggtt sequences introduced on the fragment and the homologous chromosomal sequences. Transformants were selected on LB containing kanamycin and ALA, tested for the absence of the ampicillin resistance marker present on vector sequences, and assessed for ALA auxotrophy. Southern hybridization analysis of total genomic DNA digests was performed in order to verify that the deleted copy of Eggt with the gpttt cassette insert was integrated at only one location in the genome, replacing the chromosomal copy of the Eggt gene. The construction of other plasmids used in gene replacement experiments (see Figure 7) made use of the afragment, a 2.0-kbp antibiotic resistance gene (streptomycin resistance / spectinomycin resistance) that 20 is flanked by transcription and translation termination signals (Prentki and Krisch, 1984). The cassette was isolated as a EggI DNA fragment and ligated to pMR19 DNA which had been digested with various restriction endonucleases; EggRV for pMRl9zi, EggI (partial) for pMR54 and pMR55, and EggI (filled in with the Klenow fragment of DNA polymerase I) for pMRlQQ. Cosmid Libraries. Methods used to prepare cosmid libraries were modified from those published (Adams gt gt., 1984). The first of two cosmid gene banks was prepared from genomic DNA purified from CR63 and partially digested with EggBAl. Size-fractionated DNA fragments greater than approximately 20 kbp were ligated to a cosmid vector, pV35 (C. P. Nolk, unpublished), that had been digested with EgtII and phosphatase—treated. This vector contains the origin of replication and the ampicillin resistance gene of pBR322. Products of ligation were packaged tg gtttg, and plated on E. ggtt strain SASX41B on LB agar containing ampicillin and ALA. Resulting colonies were screened for complementation of the Eggg mutation by the resident cosmid. Colonies were transferred to media with and without added ALA, and cosmids were isolated from those. which grew well without added ALA. A second cosmid library was prepared by the method described above from genomic DNA extracted from E. ggtt strain ORN125. This strain possesses a transposon TntE insertion at 26.5 min that is approximately 82x 21 cotransducible with the Eggg locus (Spears, gt gt., 1986). Purified DNA was partially digested with Egg3A and cloned into the unique EggHI site in pNH4, a cosmid vector that contains a lambda replicative origin (Herrero gt gt., 1984). Recombinant transformants of HB101 were selected on LB plus kanamycin and screened for tetracycline resistance and complementation of the hemA mutation. EgtEt Deletions ggg EE_ Squgpcing. Egt3l exonuclease was used to generate a series of overlapping deletions spanning the DNA fragment to be sequenced (Poncz gt gt., 1982). The appropriate DNA fragments were then cloned into Ml3mpl8 or Ml3mp19 (Yanisch-Perron gt gt., 1985) with the Egt3l-deleted and inserted nearest the primer binding site in the tggE coding sequence. Single-stranded DNA template was prepared in JM103 (Messing, 1983) as described (Messing gt gt., 1981) and sequenced by the dideoxy chain termination method (Sanger _t _t., 1977). The nucleotide sequence of both strands was determined at least once. Materials and enzymes used for sequencing were purchased from either Bethesda Research Laboratories or United States Biochemical Corporation and used according to the supplier’s recommendations. End-labeled probes used in 81 protocols were sequenced by the method of Maxam and Gilbert (1980). 22 Maxicell Experiments. Plasmid-encoded polypeptides were identified using the maxicell technique of Sancar gt gt. (1979) except that E. ggtt strains HBlOl and DH5 were used as the plasmid hosts. Plasmid pMRBl consists of a 2.5-kbp EgtI-Eth insert that includes ORFA plus 895 base pairs of the downstream gttg coding region and 336 bp of upstream sequence (See Figure 6) cloned as a blunt-ended fragment into the unique EtchI site of pUClQ.‘ This plasmid was digested with EgtII and re-ligated to yield pMR8LABgl, a plasmid with a deletion of the 278-bp EgtII fragment within the ORFA coding sequence. Enzyme Assays. ALA synthase activity was assayed in a 1 ml reaction containing 0.5-2.0 mg of protein for 2 hours at 37°C. Composition of the reaction mixture was that described by Guerinot and Chelm (1986). Cells were pelleted by centrifugation at 4,000 x g for 10 minutes at 4°C, washed one time with Hepes buffer (200 mE Hepes pH 8.0, 33 mE MgClz, l mE 2-mercaptoethanol, 10 mE pyridoxal 5-phosphate) and resuspended in 3—5 ml of Hepes buffer. Cells were disrupted at 4°C by two passages through a French pressure cell at 12,000 lb/inz. Cell debris was removed by centrifugation at 17,000 x g for 20 min at 4°C. The supernatants were used for determination of protein concentration by the method of Lowry (1951) and for the measurement of enzyme activity. ALA synthase assays were 23 also performed by the method of Burnham (1970) as modified by Tai gt gt. (1988). Results Isolation gt the hemA ene. Two schemes were devised to isolate the Eggg gene on a cosmid vector. The first approach was the complementation of an E. ggtt Eggg mutation with a cosmid gene bank prepared from E. ggtt strain CR63. The second strategy was to clone a TntQ tetracycline resistance marker that was inserted adjacent to Eggg on the E. ggtt chromosome in strain ORN125. reasoning that some of the transformants which harbored TntE on a recombinant cosmid would also contain Eggg. The desired cosmids were isolated as described in Materials and Methods; one cosmid obtained by the first screen (pMRCZ) and two cosmids isolated from the second (pMRC5 and pMRClO) were characterized in detail. The apparent complementation of SASX41B by the cosmid pMRCZ required that transformants were selected in the presence of exogenously-added ALA. Transformants could then grow in the absence of ALA at rates comparable to wild-type, suggesting that recombination was required for complementation. This was confirmed by showing that the apparent complementation did not occur in strain EVlO, a derivative of SASX41B made RecA' by P1 transduction from MC1024 (Table 1). In contrast, pMRC5 and pMRClO complemented SASX41B with no 24 Figure 1. Restriction endonuclease map of plasmid subclones of the isolated E. coli hemA region. The plasmid pMRlS is a subclone of cosmid pMRCZ; the genomic DNA insert is carried on the cosmid vector pV35. The EgtII site delineates one boundary of recombinant DNA present on the parent cosmid. The entire hemA and prfA genes and flanking sequences are carried on a 6.0-kbp EtngII DNA fragment that was cloned from pMRC5 into the unique EtngII site of pUC19 to yield plasmid pMR57. The location and direction of transcription of each ORF is indicated by a solid arrow; the DNA sequence was determined for the region delineated by these arrows. A 278-bp EgtII fragment within the hemA coding region was deleted and replaced by a nptII cassette to create plasmid pMR61 that was used for gene replacement mutagenesis. Restriction enzymes: B=EgtII, E=EggRI, Hp=EggI, H=EtngII, S=Sa11. 5 2 3min 5.0.120 pecan S... <52. 0000 0000 5.00 9.00 A." A .IIIY - a n m L: 0 .u 1: udl ‘ u u - g ml w 0r 0 00 m a: m z n u . _ . . . . 1| - d 1.. “1- d a q d z m a: u “m m. a: u x c. "X Aw =.ac _ _maocm . 26 requirement for prior selection in the presence of ALA or recombination function. Plasmid subclones of pMRCZ (pMR15) and pMRClO (pMR57) were obtained and these exhibited the complementation characteristics of the parent cosmids described above. A partial restriction enzyme map of each of these plasmids is presented in Figure 1. The EgtII site at the far left end of pMR15 represents the endpoint of the recombinant DNA contained on its parent cosmid, pMRC2. The presence of common restriction enzyme sites between the plasmid subclones suggested and Southern hybridization experiments verified that the DNA sequences isolated by the two different schemes were homologous. These data showed that pMR15, a plasmid subcloned from pMRCZ, hybridized to the 6.0-kbp EtngII fragment present on cosmids pMRC5 and pMRClO that was subcloned to create pMR57 (data not shown), as well as to a 4.2-kbp EggRI fragment and a 6.5-kbp EggHI fragment common to pMRC5, pMRClO and genomic DNA (Figure 2). The plasmid pMR57 therefore contained the recombinant DNA insert of pMR15, along with additional flanking DNA that appeared to be necessary for hemA activity. EEE Sequence Analysis. The DNA sequence of the region delineated in Figure 1 was determined. An examination of the sequence data revealed two large ORFs, herein designated A and B. These two ORFs, along with their coding capacity and direction of transcription are shown schematically in Figure 1. If ORFA corresponds to the hemA 27 Figure 2. Southern hybridization analysis of codmids pMRC5, pMRClO, and genomic DNA. EggRI (lanes 1-3) and EggHI (lanes 4-6) digests of pMRC5 (lanes 1 and 4), pMRClO (lanes 2 and 5), and genomic (lanes 3 and 6) DNA were transferred to cellulose nitrate and hybridized to 32P-labeled pMR15 DNA. 2.7“ 29 prfA End GCAGGAA[TA AITG GAA TAT CAA CAC TGG TTA CGT t‘ M E Y 0 H N L R ORFC so . CAA CTT CAG GCG AGC GAA AGC CCG CGG CGT GAT 0 L 0 A S E S P R R D 100 t CTG GAG CAT GTT ACC GGC AGA GGG CGT ACT TTT L E H V T G R G R T F 150 I GGT GAA ACG CAG CTG ACT GAC GAA CAA TGT CAG G E T 0 L T D E G C 0 zoo t CTA CTG ACA CGT CGT CGC GAT GGT GAA CCC ATT L L T R R R D G E P I ECORI CGG GGT GCG AGA ATT C R G A R I Figure 3. Partial nucleotide and predicted sequence of ORFC. GAA GCA E A GCT GAA A E ATT CTC I L CAA CTT 0 L GCT CAT A H ATA ATC GCC GAT TTA L amino acid AGC CTG TTT GCG AGC S 30 ACGGTAACGCTAGCATTAAGGGTTATAACTGCAACGTATCTCAAGGACTTGTCATCACT +1 ORFD ATG M GCT A GGC G GTG V ATT I ACC T AGC S GTC V ATG M CGC R CTG L GGC G CAA 0 CCC R CAC CCC P CTT L AAA E CGC R TCT S GGC G ACG T GAC D ATT I CAG 0 GAC D AAA E CCT P ATC I ACT 45 CTG CCC GAT TTT CGT CTT ATC CGC CTG CTA CCG CTG GCT L GTG V AGC S AAT N GAC D CAG 0 GAA E AAT N GGC G TGG N GAC D AAC N 000 A AAG K GGC Figure 4. ORFD. P CTC L CCG P CTT L CAA 0 CAT D CTG L AAA K AAA K ATT I CAG 0 TGG N ATG M TTA L CTC ACT T GAT D AAT N CAA 0 CCC R GAG E GGT G TTG L TTA L TAC V AAG K CCA P AAA E TCC F GCC A TCG S CAG 0 AAA K TAC Y CTG L CAG 0 ACC T GGT G CGC R GTT V GCC A ATG M GGC TGT C CCA P TAT Y GTG V CGT R AAT N CGT R GGA G TTA L CTG L GTT V AAT N GAT D AAA L TCC S CAA Q CAG Q TAC Y CTG L GCT A TAT Y ATG M 006 P AGC S TAT Y ATG M AAC N ACT I GTT V TGG W ACT T GCC A CTG L CAA 0 ACC T CCA P GGT G GAA E GGT G GAA E TGG W TAA Nucleotide and predicted ACC T CGT R 000 R CGC R CTC L CCG P GCC A ATT I GAT D ATT I GGT G CTC L ATA I TCT L ACG T CAG Q GGC G TTT F ACT T GGT G GAT D CCG P GCA A ACC T TAT Y ACC T GTG V GTT L CCC P CAT H 000 A TTC F AAC N AAC N GAC D CTC L ACC T TAC Y GAC D GAC D AAA K TTT P AAA K CAG 0 TTC F TGG W CCA P GTC V GCC A AAC N GAC D AGC S ACC T GGT G TAA End ATA L GGT G CAA Q GCT A CAG Q TTG L GAG GAA E AGC S TAC Y CAG 0 AAA K GGT G TGC CAT amino acid sequence A so CCT P 135 GAC D 130 TAT V 225 CAA 0 270 GGC G 315 TTA L 3so GAG E 405 TTG L 050 AAA E 405 AAT N 540 A00 T sss CAA 0 GGA TGC 31 gene, it would explain why pMR15 does not display Eggt activity, since it lacks a portion of this open reading frame. The experiments described below supported this interpretation and indicated that ORFA corresponded to the Eggg gene. Two other ORFs of undetermined lengths (ORFC and ORFD in Figure l) are also located in this region and their nucleotide sequences are presented in Figures 3 and 4. These sequences did not display significant similarity to any sequences contained in the GenBank data base and they have not been characterized further. A computer search of GenBank sequences disclosed that ORFB corresponded to 251g, the gene that encodes peptide chain release factor 1 (RFl), which had been isolated and sequenced previously (Craigen gt gt., 1985). Since the locus for the RFI gene has been mapped precisely to 26.7 min on the genetic linkage map (Ryden gt gt., 1986; Lee gt gt., 1988), and hemA maps to about 27 min (Bachmann, 1983), the cloned DNA sequences came from the Eggg region of the chromosome. The coding sequence of the putative Eggt gene, ORFA, along with upstream flanking sequences are presented in Figure 5. Assuming that the coding region begins at the first ATG, ORFA could theoretically code for a polypeptide consisting of 418 amino acids with a molecular weight of 46,312. The 4l-bp intergenic region separating the ORFA stop codon and the first codon of 21;; are also presented in the figure. 32 Figure 5. Nucleotide sequence of the hemA gene and the predicted amino acid sequence of the hemA polypeptide. The sequence of each DNA strand is shown for the region that separates hemA from the putative translational start site of the divergent ORFD. As described in Chapter 3, the proposed transcriptional start sites of the two major hemA RNA species are indicated by asterisks and sequences which comprise possible promoter elements are underlined. The likely promoter sequences of the divergent ORFD transcript are underscored with dashed lines and arrows indicate the nucleotides which correspond to the 5’ end of the ORFD transcript. -220 -180 -120 -60 +1 61 121 181 241 301 361 421 481 541 601 661 721 m 841 901 961 1021 1081 1141 1201 1261 33 GGGGCATAGTGATGACAAGTCCTTGAGATACGTTGCAGTT CCCCQTATCACTACTGTTCAGGAACTCTATGCAACGTCAA ORFD +1 hem_A2 araacccrrAArocraoccrraccorccocrircorcrarcrrCAAorrorcrraarroc TATTGGGAATTACQATCGCAATGGCAGGCGATAGCAEATACAAGTTCAACAGAATTAACG AAA CAGAATCTAACGGCTTTCGGCAATTACTCCAAAAGGGGGCGCTCTCTTTTATTGATCTTA GTCTTAGATTGCCGAAAGCCGTTAATGAGGTTTTCCCCCGCGAGAGAAAATAACTAGAAT least I CGCATCCTGIAxQAIGCAAGCAGACTAACCCTATCAACGTTGGTATTATTTCCCGCAGAC GCGTAGGACATACTACGTTCGTCTGATTGGGATAGTTGCAACCATAATAAAGGGCGTCTG hemA ATGACCCTTTTAGCACTCGGTATCAACCATAAAACGGCACCTGTATCGCNECGAGAACGT M T L L A L G I N H K T A P V S L R E R GTATCGEhTTCGCCGGATAAGCTCGATCAGGCGCTTGACAGCCTGCTTGCGCAGCCGATG V S F S P D K L D O A L D S L L A 0 P M GTGCAGGGCGGCGTGGTGCTGTCGACGTGCAACCGCACGGAACTTTATCTTAGCGTTGAA V O G G V V L S T C N R T E L Y L S V E GAGCAGGACAACCTGCAAGAGGCGTTAATCCGCTGGCTTTGCGATTATCACAATCTTAAT E 0 D N L 0 E A L 1 R N L C D Y H N L N GAAGAAGATCTGCGTAAAAGCCTCTACTGGCATCAGGATAACGACGCGGTTAGCCATTTA E E 0 L R E S L Y W H 0 0 N D A V S H L ATGCGTGTTGCCAGCGGCCTGGATTCACTGGTTCTGGGGGAGCCGCAGATCCTCGGTCAG M R V A S G L D S L V L G E P 0 I L G 0 GTTAAAAAAGCGTTTGCCGATTCGCAAAAAGGTCATATGAAGGCCAGCGAACTGGAACGC V K I A F A D S 0 K G H M K A S E L E R ATGTTCCAGAAATCTTTCTCTGTCGCGAAACGCGTTCGCACTGAAACAGATATCGGTGCC M F 0 K S F S V A K R V R T E T D I G A AGCGCTGTGTCTGTCGCTTTTGCGGCTTGTACGCTGGCGCGGCAGATCTTTGAATCGCTC S A V S V A F A A C T L A R 0 1 F E S L TCTACGGTCACAGTGTTGCTGGTAGGCGCGGGCGAAACTATCGAGCTGGTGGCGCGTCAT S T V T V L L V G A G E T I E L V A R H CTGCGCGAACACAAAGTACAGAAGATGATTATCGCCAACCGCACTCGCGAACGTGCCCAA L R E H K V 0 K M I I A N R T R E R A 0 ATTCTGGCAGATGAAGTCGGCGCGGAAGTGATTGCCCTGAGTGATATCGACGAACGTCTG I L A D E V G A E V I A L S D I D E R L CGCGAAGCCGATATCATCATCAGTTCCACCGCCAGCCCGTTACCGATTATCGGGAAAGGC R E A D I I I S S T A S P L P I I G K G ATGGTGGAGCGCGCATTAAAAAGCCGTCGCAACCAACCAATGCTGTTGGTGGATATTGCC M V E R A L K S R R N 0 P M L L V D I A GTTCCGCGCGATGTTGAGCCGGAAGTTGGCAAACTGGCGAATGCTTATCTTTATAGCGTT V P R D V E P E V G K L A N A Y L Y S V GATGATCTGCAAAGCATCATTTCGCACAACCTGGCGCAGCGTAAAGCCGCAGCGGTTGAG D D L Q S I I S H N L A 0 R K A A A V E GCGGAAACTATTGTCGCTCAGGAAACCAGCGAATTTATGCCGTGGCTGCGAGCACAAAGC A E T I V A 0 E T S E F M A N L R A Q S GCCAGCGAAACCATTCGCGAGTATCGCAGCCAGGCAGAGCAAGTTCGCGATGAGTTAACC A S E T I R E Y R S 0 A E 0 V R D E L T GCCAAAGCGTTAGCGGCCCTTGAGCAGGGCGGCGACGCGCAAGCCATTATGCAGGATCTG A K A L A A L E 0 G G D A 0 A I M 0 D L GCATGGAAACTGACTAACCGCTTGATCCATGCGCCAACGAAATCACTTCAACAGGCCGCC A H K L T W R L I H A P T K S L 0 O A A CGTGACGGGGATAACGAACGCCTGAATATTCTGCGCGACAGCCTCCGGCTGGAGTAGCAG R D G D N E R L N I L R D S L G L E ‘ TACATCATTTTCTTTTTTTACAGGGTGCATTTACGCCTATQAAGCCTTCTATCGTTG rl prfA 34 Figure 6. Autoradiogram of plasmid polypeptides labeled in the maxicell procedure and schematic depiction of pMR81 and pMR81ABg1. The [35S]-methionine-labeled extracts were analyzed on a 10* sodium dodecyl sulfate polyacrylamide gel. The plasmids harbored by the host E. coli strain DH5 are identified above the appropriate lanes and the chromosomal DNA inserts contained on the plasmids are illustrated schematically below. The dashed lines in the hemA coding sequence represent the region of the 278-bp deletion. The electrophoretic mobility of the protein molecular mass markers egg albumin (45,000) and carbonic anhydrase (29,000) are indicated. The arrow identifies the position of the 46-kilodalton polypeptide (middle lane) that is coded for by the hemA gene. A polypeptide of approximately 40 kilodaltons that probably corresponds to the truncated RFl protein is synthesized by both plasmids. The intact RFl protein migrates as a 48-kiloda1ton polypeptide in this gel system (my unpublished result). In addition, a band that migrates below the 28,000 dalton 0-1actamase protein is unique to pMR81ABg1 and probably represents a product of the deleted hemA gene. Restriction endonuclease sites are EgtI (B) and EygI (P). 35 0: P U 2 a 'pMRBi pMRB1ABgl kDa <— 45- ' 29h- ? P I l hemA H prfA j pMR81 9 I? 1 r I AhemA U milk 7 pMRBtABgI g—A 5000p 36 Maxicell Experiments. Maxicell labeling experiments were performed in order to determine whether a polypeptide of the anticipated size was produced by ORFA. A polypeptide of approximately 46 kilodaltons was synthesized, in maxicell experiments in which ORFA, the presumptive Eggg gene, was cloned into pUC19 on plasmid pMR81 (Figure 6). There are no other ORFs present on this plasmid which have a coding capacity sufficiently large to encode this polypeptide. Furthermore, an identical plasmid that was deleted for the 278-bp EgtII fragment within ORFA (pMRBIA Bgl) failed to synthesize this polypeptide (Figure 6). These experiments showed that ORFA can direct the synthesis of a polypeptide the size of which is in agreement with that predicted by the DNA coding sequence. Construction gt Eggg Mutant Ey Eggg Replacement. The following experiments demonstrated that the 46-kiloda1ton polypeptide encoded by ORFA is required for ALA synthesis. If ORFA codes for a polypeptide necessary for ALA synthesis, then a deletion in the coding region of the chromosomal copy of this gene should result in ALA auxotrophy. Plasmid pMR61, (see Materials and Methods and Figure 1), unlike its parent plasmid, pMR57, could no longer complement the Eggg mutation and did not produce the 46-kiloda1ton polypeptide in maxicells (data not shown), indicating that the gene had been interrupted. The gene replacement technique described in Materials and Methods 37 was used to introduce the inactivated copy of ORFA on plasmid pMR61 into the genome, replacing the chromosomal copy of ORFA. Fourteen of the 14 kanamycin resistant, ampicillin sensitive transformants selected by this procedure displayed the characteristics of ALA auxotrophs. These strains were respiratory-deficient, being unable to grow on a non-fermentable carbon substrate such as acetate in the absence of exogenous ALA. Aerobic growth in LB broth without ALA was undetectable even after prolonged incubation. When strains were grown on LB plates without ALA, they formed minute colonies that displayed no detectable catalase activity. These three phenotypes could be completely rescued by the exogenous addition of ALA to the growth media and thus were due to ALA auxotrophy resulting from the inactivation of ORFA. One of these strains, designated EV61, was characterized further. In parallel experiments, it was not possible to introduce into the genome a Eggg disruption in which the gpttt cassette was inserted in the opposite orientation. Unsuccessful attempts were also made to introduce the deletion and insertions depicted in Figure 7 into the genome. Possible reasons why these gene replacements were not recovered are presented in the Discussion. Total genomic DNA was extracted from EV61 and JC7623 and used in Southern hybridization studies in order to verify that the expected gene replacement event had occurred. The DNA samples were digested with EcoRI, 38 Figure 7. Inserts of plasmids used for site-directed mutagenesis. A restriction endonuclease map for the plasmid subclone (pMR19) used to construct plasmids for site-directed mutagenesis is shown along with a schematic depiction of genes contained on the plasmid. The dotted line denotes the segment of the hemA Open reading frame not present on pMRl9. The dashed lines represent either a region of E. coli DNA deleted and replaced by the eragment or a site into which the eragment was inserted. Restriction sites are EgtII (B), EggRV (V), Eng (Hp), gygI (A). EQRI (E). 39 . 30.2.. m c n 30.2.. n . e .. 0252.. u . 0 i ...................................... 3.5.2.. .380 . m a: > < a: >> 0 > 0 AH ......... 5:. (a... 40 Figure 8. Southern hybridization analysis of EV61 genomic DNA. EggRI digests of chromosomal DNA from strain JC7623 (lanes 1 and 3) and strain EV61 (lanes 2 and 4) were transferred to cellulose nitrate and hybridized to two radiolabeled probes. Lanes 1 and 2 were hybridized to the 1.4-kbp nptII gene and lanes 3 and 4 were hybridized to the 1.8-kbp Epgl fragment containing the E. coli hemA gene. kbp 5.2— 4.2— 1 41 34 42 size-fractionated on 18 agarose gels, transferred to cellulose nitrate and hybridized to two different DNA probes: the 1.3-kbp gpttt gene and a 1.9-kbp Egg I fragment which contains the major portion of the ORFA gene along with 700 bp of upstream DNA sequences (Figure l). The gpttt probe hybridized to a 5.2-kph Egg RI fragment of DNA derived from the mutant but did not hybridize to DNA isolated from the parent strain (Figure 8, lanes 1 and 2). The Eng probe hybridized to a 4.2-kbp EggRI fragment in DNA derived from strain JC7623, but, as expected, this fragment was shifted to 5.2 kbp in DNA derived from strain EV61 (Figure 8, lanes 3 and 4). This analysis showed clearly that the wild type chromosomal copy of the ORFA gene was absent in EV61 and had been replaced by the inactive copy. Et Transductions. If ORFA corresponds to the Eggg gene, then the kanamycin resistance marker present in EV61 should map at approximately 26.7 min and should be co-transduced with ttp at approximately 27.5 min (Bachmann, 1983). To test this, a P1 lysate was prepared from JK268 (tr A, tr E, Eggt’) and used to transduce strain EV61, selecting for HemAt (ability to form normal colonies on LB with no ALA supplement) and then testing for the Trp' phenotype. The Eggg_and tgp markers were co-transduced at a frequency of 15.43 (10/65). All of the HemAt transductants of EV61 were sensitive to kanamycin, indicating that a wild-type ORFA 43 allele had replaced the inactive ORFA gene containing the gpttt insertion. These genetic data corroborated the DNA sequence analysis that located ORFA to the 26.7 min region of the E. ggtt linkage map. For unknown reasons, we were unable to transduce the kanamycin resistance marker from EV61 into various recipient strains (see Discussion). gt_ Synthase Assays. Extracts from various strains of E. ggtt, including strains which harbored the putative Eggt gene on high copy number plasmids, were tested for ALA synthase activity by the protocols described in Materials and Methods. Although strains which harbored the plasmid-bourne Eggt gene appeared to accumulate tetrapyrroles in liquid culture and on solid media, I was unable to detect activity in any of these strains. However, ALA synthase activity could be readily demonstrated by these methods in an E. ggtt strain which contained the Bradyrhizobium japonicum Eggg gene cloned into p009 (data not shown), indicating that the protocols were adequate for ALA synthase detection. Reports of ALA synthase activity in E. ggtt_are inconsistent (Schoenhaur and Curtis, 1986; Tai gt gt., 1988) and recent experiments indicate that ALA is synthesized by the 5-carbon glutamate route in E. ggtt (see below). Additional experiments are required to resolve this question. 44 Discussion This chapter describes the isolation and nucleotide sequence of the hemA gene from E. coli. The cloned gene was identified as hemA based on the following results. (i) The gene complemented the hemA mutation of SASX4lB. The apparent complementation of the mutant Eggg allele by truncated copies of the cloned gene required RecA function, indicating that the cloned DNA corresponded to Eggg and eliminating the possibility that the cloned gene suppressed the mutant phenotype of SASX41B by a mechanism involving a different locus. (ii) The gene was localized to approximately 26.7 min on the E. ggtt linkage map, the site mapped previously for the Eggt locus. Sequence analysis of the cloned DNA disclosed that the Eggg gene lies directly upstream of and is transcribed in the same direction as 251g, which encodes RFl and maps to 26.7 min. Transduction mapping using P1 also localized the cloned gene to the 27 min region of the chromosome, and, in addition, the restriction map of the DNA sequence surrounding the isolated Eggg gene corresponds to the chromosomal restriction map at 26.7 min (Kohara gt gt,, 1987). (iii) Inactivation of the chromosomal homolog of the cloned gene created an ALA auxotroph. Strain EV61 contained a defined insertion in the hemA gene that removed 278 nucleotides of the coding sequence. This strain 45 exhibited the characteristic phenotype of a Eggg strain; it was respiratory-deficient and lacked catalase activity, but formed small colonies on agar plates if provided with a fermentable carbon source. The mutant phenotype was completely reversible when the media was supplemented with ALA. Thus, the Eggg gene probably does not encode a polypeptide vital to the survival of E. ggtt, although the possibility that strain EV61 harbors a compensatory suppressor mutation or that some ALA is synthesized via an alternate pathway has not been ruled out by these experiments. The inability to construct a gene replacement mutant in which the gpttt cassette is transcribed in a direction opposite to that of Eggg implies that this insertion may be polar on the downstream pttg gene. (In an extensive genetic analysis, Elliott and Roth [1989] failed to recover Eggg insertion and deletion mutations in Salmonella typhimurium, although point mutations at this locus were readily isolated.) This polarity effect may be less severe or absent when the gpttt gene is transcribed in the same direction as pgtg. In this orientation, the gpttt gene may provide a promoter that allows pgtt transcription, or, alternatively, the insertion of opposite polarity may introduce a transcriptional terminator which prevents pttt from being expressed from an upstream promoter (hemA or nptII). 46 Attempts to recover gene replacement mutations in which pgtg was disrupted by insertions containing transcriptional and translational terminators were unsuccessful, as were efforts to create chromosomal deletions that extended into the pttt coding region, suggesting that RFl is essential for the viability of g. ggtt. This i. not surprising, in light of its role in polypeptide chain termination. Termination of protein synthesis and release of the completed polypeptide chain require the activity of release factors which recognize specific termination codons, bind to the ribosome and facilitate peptidyl-tRNA hydrolysis. There are three release factors in E. ggtt which participate in this process: RFl catalyzes termination at codons UAG and UAA; RF2 specifically recognizes termination codons UGA and UAA; and RF3 enhances the activities of RFI and RF2 (both binding and dissociation) but does not recognize nonsense codons (Hershey, 1987). In the absence of RFl, amber (UAG) termination codons cannot be recognized, and its function would therefore be essential to survival. This hypothesis is supported by the fact that pttE, which encodes RF2 is indispensable for the growth of E. ggtt (Kawakami gt gt., 1988). The above observations and considerations, in conjunction with the inability to transduce the Eggg insertion from EV61 into different genetic backgrounds, suggest that some insertions in Eggg may be polar on prfA and that strain EV61 may possess a 47 secondary mutation that suppresses the polarity effect on pgtt expression. The computer analysis program of Pustell (1984) was used to compare the DNA and amino acid sequences of the E. {ggtt Eggg gene and ALA synthase sequences which had been reported from several organisms. The E. ggtt_EggE sequence exhibited no significant similarities with the ALA synthase sequences of E. ’a onicum, chicken, mouse, or human (data not shown). This observation was unexpected as substantial similarities exist between E. japonicum and chicken (McClung gt gt., 1987), mouse, and human (my unpublished results) and among the various ALA synthase sequences of eukaryotic origin (Bawden gt gt., 1987). However, the E. ggtt'gggg gene exhibits extensive similarities with the DNA and amino acid sequences of the cloned E. typhimurium Eggt gene (Elliott, 1989). The predicted amino acid sequences share 94.38 identical residues over the entire length of the polypeptides and the amino acids are 96.93 conserved (data not shown). This lack of similarity to cloned ALA synthase sequences, coupled with the lack of detectable ALA synthase activity in E. ggtt cell extracts, suggested that the primary route of ALA synthesis in E. coli might not be via ALA synthase. Interestingly, the E. japonicum hemA gene cloned into pUC9 complemented the E. coli hemA strain SASX41B. This result implies that ALA synthase can function in E. coli but does not require that this is the 48 usual means of ALA biosynthesis in this organism. Two recent reports (Avissar and Beale, 1989; Li _t _t., 1989b) present convincing evidence that the 5-carbon glutamate pathway operates in E. ggtt. By following the fate of labeled precursor compounds, both groups demonstrated that glutamate, but not glycine, is incorporated into ALA by E. ggtt_cells. ALA formation in tg ytttg assays displayed characteristics typical of the 5-carbon pathway, requiring glutamate and reduced pyridine nucleotide and being sensitive to RNase and gabaculine, an inhibitor of the aminotransferase (Hoober gt gt., 1988). Furthermore, Avissar and Beale (1989) have shown that the enzyme activity absent in the Eggg strain SASX41B appears to be that of the dehydrogenase. These data are consistent with my results and suggest that the Eggg gene codes for the dehydrogenase. In summary, I have presented evidence that the Eggg gene encodes a polypeptide which functions in ALA synthesis, but which shows no similarity to cloned ALAS sequences. I was unsuccessful in detecting ALAS activity in E. ggtt extracts. Both of these results agree with evidence discussed above which indicates that E. ggtt synthesizes ALA by the 5-carbon glutamate pathway. It is likely that the Eggg gene codes for the dehydrogenase component of this reaction sequence, which converts the tRNA-activated glutamate to GSA. CHAPTER 3 Fine Structure Analysis of the Escherichia coli hemA Transcriptional Unit Introduction This chapter describes experiments performed to characterize the transcriptional unit of the Eggg gene with the eventual goal of determining whether its expression is regulated at the level of transcription initiation, and, if so, by what controlling elements and cellular signals. As noted in Chapter 1, the amount of heme present in E. ggtt cells appears to be regulated and can vary under different growth conditions (Ishida and Hino, 1972). Similarly, in E. typhimurium, the synthesis of cytochromes is reduced under conditions of anaerobic growth on glucose as compared to aerobic growth (Elliott and Roth, 1989). With the exception of the activation and/or stabilization of PBG D by the availability of its substrate, PBG, the molecular mechanisms that effect regulation of heme synthesis have not been characterized. E. coli cells grown in media with an excess of exogenous ALA are reported to accumulate and excrete porphyrin compounds (McConville and Charles, 1979a), suggesting that the first step in heme biosynthesis, ALA formation, may be a rate-limiting step for the overall pathway. 49 50 The biosynthesis of ALA from glutamate confers additional complexity to the question of regulation because at least three enzymes and a tRNA species are requisite components of this three-step reaction sequence. To date, regulation of the 5-carbon glutamate pathway has been investigated only in plants, where experiments with dark-grown seedlings suggest that ALA formation in leaf plastids controls tetrapyrrole biosynthesis (Kannangara gt _t., 1988). When supplied with radiolabeled ALA, such leaves convert the exogenous ALA into protochlorophyllide, indicating that the other enzymes in the pathway are present in the plastids and that ALA synthesis is the reaction inhibited in the dark. The molecular mechanism(s) that regulate the ALA synthesizing activity in plastids remains unresolved. The regulation of ALA synthesis in E. ggtt has not yet been investigated and may be very different from regulation of ALA formation in plants. It will be interesting to determine if the very different biological systems, i.e., a prokaryotic chemoheterotroph and a photosynthetic eukaryotic organelle, share common controlling elements. As described in Chapter 2, the E. ggtt Eggg gene is required for ALA synthesis and probably encodes the dehydrogenase component of the 5-carbon pathway. As an initial step toward the characterization of expression of this gene, I have localized the transcriptional initiation sites for the hemA message using 81 nuclease protection 51 procedures. An upstream transcript with opposite polarity to Eggg was also revealed during the course of these experiments. Having established a detailed map of the region containing these initiation sites, it will be possible to use the cloned sequences as probes to determine whether the level of transcription of Eggg is regulated by the cell and to correlate the level of its expression with the amount of ALA and heme synthesized by the cell in response to different growth environments. Materials and Methods E. coli Strains and Growth Conditions. The E. ggtt strains used are described in Chapter 1, Table 1. Routine plasmid construction and maintenance was done in HBlOl and M13 manipulations were performed using JM103. The RNA samples used in 81 nuclease protection experiments were prepared from 500 m1 cultures of CR63 grown under aerobic or anaerobic conditions (as described in Chapter 2) in M9 medium containing glucose or succinate and harvested at the mid-log phase of growth. RNA Isolation. RNA was isolated by a CsCl gradient method as described (Adams and Chelm, 1984). 81 Nuclease Mapping. The method of Berk and Sharp (1977) as modified by Adams and Chelm (1984) was used to 52 map the 5’ transcriptional initiation sites of the Eggg mRNA. A 447-bp EgtI-EgtNI fragment that includes 144 bp of the Eggg coding sequence and 303 bp of upstream sequence was used in initial 81 nuclease protection experiments (Figure 1). This fragment was isolated from a polyacrylamide gel, treated with calf intestinal alkaline phosphatase, and then 5’-end-labeled with [Y32PJATP using T4 polynucleotide kinase. Single-stranded labeled fragments were separated on an 8% denaturing polyacrylamide gel, isolated, and hybridized to 20 ug of RNA according to procedures referenced above. To localize the 5’ ends of the transcripts, each of the SI nuclease—protected fragments was size-fractionated by electrophoresis adjacent to the DNA sequence ladder of its corresponding full-sized probe. For better resolution of the 5’ end of the Eggt transcript, a EggE-specific single-stranded DNA probe was synthesized by a primer extension method described by Adams and Chelm (1988). A synthetic oligonucleotide, 5’-ACGATACACGTTCTCGC (Figure l) was labeled at its 5’ end and annealed to a single-stranded template prepared from a recombinant Ml3mp19 phage containing the 675-bp Egtl fragment shown in Figure 1. The hybridized primer was extended, permitting the synthesis of an end-labeled DNA strand that was complementary to the hemA message. This labeled probe was purified and used in 81 nuclease protection experiments as described above. The same 53 oligonucleotide primer and single-stranded DNA template were used in a dideoxy chain termination reaction to generate a sequence ladder. Results Determination gt the Transcriptiongt_Start Site gt the Eggt Eggg. The initiation sites of transcripts which originated in the region upstream of the Eggt gene were mapped by $1 nuclease protection methods using the Egtl-EgtNI probe described in Materials and Methods and depicted in Figure 1. Each end-labeled strand of probe was hybridized to RNA prepared from cultures of E. ggtt strain CR63 grown under several different conditions, selected to represent various energy-generating modes of metabolism. The results of one set of experiments are presented in Figure 2. Two RNA species of different sizes protected the probe corresponding to one of the DNA strands (lanes 6, 7, and 8) and one RNA species protected the probe corresponding to the opposite strand (lanes 3 and 4; a band in lane 2 is present, but faint). These results indicated that within the DNA encompassed by the EgtI-EgtNI fragment there were two divergent transcriptional units. Electrophoresis of each of the $1 nuclease protected products adjacent to a Maxam Gilbert sequence ladder generated from its corresponding full-length probe enabled identification of the 5’ end points of the presumptive 54 Figure 1. Nucleotide sequence of the 5’ end and the flanking upstream region of the E. coli hemA gene. The open arrows on the schematic diagram of this region indicate the beginning of the hemA and ORFD coding region as predicted by the DNA sequence shown below. The SalI/BstNI fragment that spans the intergenic region was used in initial 81 nuclease protection experiments. The dashed lines and arrows below represent the RNA transcripts that initiate in this region. The nucleotide sequence with the reverse complement of the synthetic oligonucleotide used in high resolution 81 nuclease mapping of the hemA transcript is in brackets. The asterisks indicate the site of transcriptional initiation for hemA and the arrowheads indicate the approximate location of the 5’ end of the ORFD transcript. Nucleotides that display homology to the E. coli promoter consensus sequences are underlined for the hemA transcripts and are indicated by lowercase letters for the ORFD transcript. The broken lines underscore the two different sequences with homology to the -35 consensus sequence for hemAi. 8*" I *oaso] BstI 55 11—1 GTAGCAGGCG CCTTGAGATA -2oo gcagat TATCGTCTAT ~1so CAATTACTCC -lOO ATGACCCTTT @CGAGAACGT GCCTGCTTGC GATAAGACGA CGTTGCAGTT AAATCGGGCA VVV ATAACCCTTA ORFD+ gta GGGGCATAGT tacgat ATGCTAGCGT m2 I GTTCAAGTTG AAAAGGGGGC was t CAGACTAACC TAGCACTCGG orircoflrrr GCAGCCGATG TCTTAATTGC CAGAATCTAA GCTCTCTttt_ATIGATCTTA CTATCAACGT TATCAACCAT CGCCGGATAA GTGCAGGGCG TGGTATTATT AAAACGGCAC GCTCGATCAG GCGTGGTGCT GATGACAAGT TACCGTCCGC CGGCTTTCGG CGCATCCTGt TCCCGCAGAC CTGTATCGCT GCGCTTGACA GTCGACGTGC 56 Figure 2. 81 nuclease protection analysis. The probes used in 81 nuclease protection experiments were each of the 5’ end-labeled strands of the 447-bp EgtNI-EgtI fragment shown in Figure 1. Each probe was hybridized to the following RNA types: aerobic M9-succinate (lanes 2 and 6); aerobic M9-glucose (lanes 3 and 7); anaerobic M9 glucose (lanes 4 and 8). Control lanes: lanes m1 and ma contain the probes alone; lanes 1 and 5 contain the labeled probes hybridized with salmon sperm DNA. ORFD -- 57 58 transcriptional start sites (results for ORFD are shown in Figure 3). Because the nucleotides which corresponded to the initiation sites for the Eggg message could not be determined unambiguously using this probe, a Eggg-specific oligonucleotide probe was used for high resolution mapping of this transcript and these data are presented in Figure 4. The nucleotides which correspond to the 5’ ends of the Eggt message and the divergent ORFD transcript are indicated in Figure 1. The 5’ ends of the Eggg transcript are separated by 92 nucleotides and are evident in each of four different RNA preparations and in experiments using several different probes (Figure 2, 4, and data not shown). It is therefore unlikely that they are due to artifacts associated with the 81 nuclease digestion or RNA isolation procedure. These results suggested that there may be two functional transcriptional start sites for the Eggg message or that a RNA processing (cleavage) event produces the two different RNA species. The apparent start sites of the distal Eggg transcript and the ORFD transcript do not overlap, but are separated by approximately 45 nucleotides (Figure 1). Whether these sequences play a role in the expression of either transcript has not yet been addressed experimentally. Nucleotides which comprise possible promoter sequences for each of the transcripts were identified upstream of each of the transcriptional start site. These are depicted 59 Figure 3. $1 nuclease protection mapping of the ORFD transcriptional start site. The product of an 81 nuclease protection experiment that was carried out using the 447-bp EgtI/EgtNI fragment that spans the hemA/ORFD intergenic region (Figure 1) was electrophoresed in lane 1, adjacent to the Maxam and Gilbert sequence reactions performed on the labeled ORFD coding strand. The nucleotide sequence depicted on the left is that of the ORFD coding strand and arrows indicate the approximate location of the ORFD transcriptional start site. -I-IQO>0-14> 60 G A>C T+C C o -J’ 61 Figure 4. High resolution 81 nuclease mapping of the hemA transcript. A single-stranded DNA probe was generated by primer extension of the hemA—specific synthetic oligonucleotide (see Figure 1) and hybridized to 20 ug of RNA. The protected DNA fragments of 81 nuclease digestion were size-fractionated next to a DNA sequencing ladder which had been produced by primer extension of the same oligonucleotide. s = probe hybridized to RNA extracted from E. coli cultures grown aerobically on M9 medium plus succinate; g = probe hybridized to RNA extracted from E. cultures grown aerobically on M9 medium plus glucose; h = probe hybridized to an equal amount of heterologous denatured salmon sperm DNA; and p = probe not subjected to 81 nuclease digestion. Nucleotides depicted on the left correspond to the sequence of the hemA coding strand. Nucleotides which correspond to the most likely 5’ ends of hemA transcripts are indicated by arrowheads. 62 sghp _cATc -/ GTTCAACAGA/A/ p 2 A m s h ..__...._.._..._........._. c r . \. TCGTCTGAT horns, : 63 in Figure 5 in a comparison with the canonical E. coli 07° consensus sequences (Hawley and McClure, 1983). Discussion The results of 81 nuclease protection experiments suggest that transcription of the Eggg message may initiate at two different sites, or, alternatively, that there may be a processing event at the 5’ non-translated end of the message. Two protected fragments of different sizes were apparent when experiments were performed with RNA samples isolated from cells grown under several different cultural conditions with respect to carbon substrate and anaerobicity. The experiments shown in Figure 2 suggest that the steady state level of the Egggi transcipt is higher in cells grown on glucose as compared to cells grown on succinate. Anaerobiosis had no apparent effect on the level of this transcript and the level of the Egggz transcript was approximately equal for all of the RNA samples tested. The ORFD transcript was present at levels comparable to those seen for Egggi. It will now be possible to perform a thorough investigation into the question of whether Eggg expression plays a role in the regulation of heme biosynthesis. The intergenic region upstream of Eggg is likely to contain controlling elements that affect the expression of ORFD as well as hemA. Nucleotides upstream of each of the g. ggtt ...TTGACA.. consensus Egggi ...TTtAtt... Eggti ...TTGAtc... £2252 ...TTaAtg... ORFD ...TaGACg... 64 .17 bp.... .16 bp.... .12 bp.... .17 bp.... .17 bp.... TATAAT TATgAT TATgAT. TATch TAgcAT Figure 5. Promoter sequence comparisons. denote homology between the E. ..5-9 bp...* ...7 bp....g ..7 bp....g ...12 bp...g ..7-9 bp...t or a Uppercase letters the proposed promoter sequences for hemA and ORFD. initiation of transcription is indicated by an asterisk. coli consensus sequences and The 65 transcriptional start sites were identified that display similarity to the canonical -10 and -35 sequences of E. oli 7° consensus promoter elements (Figure 5). For 0 Egggl, two different hexanucleotides are denoted in Figure 5 that show similarity to the -35 promoter element. It should be noted that although the sequence proximal to Eggtl is better matched to the consensus, it is not spaced appropriately with respect to the -10 element, and thus, may not function with the designated -10 element as a RNA polymerase binding site. Whether these or other sequences comprise functional promoters has not been addressed in this study. Regions of divergent transcription are common in E. ggtt, and in many cases, the product of one transcript plays a regulatory role in the level of expression of transcripts that originate in the region (Beck and Warren, 1988). The possibility that ORFD could play a role in the level of Eggg mRNA or vice versa therefore warrants investigation, particularly since the putative RNA polymerase binding sites for ORFD and EggEZ overlap. A 4l-bp region, detailed in Figure 6, separates the stop codon of the Eggg open reading frame from the start codon of the pttg gene. Because the evidence presented in Chapter 2 suggests that pttt expression is dependent upon Eggg, it is reasonable to hypothesize that this region of nucleotide sequence may contain control signals that affect prfA expression. The stop codon of the hemA gene is an 66 amber codon, which is recognized specifically by RFl 1., 1985). Whether this observation is of any (Craigen gt functional significance has not been determined experimentally, but a model for autoregulation of RFI has been suggested (Elliott, 1989). A phenomenon termed translational readthrough has been documented in E. ggtt, whereby, instead of terminating protein synthesis, an amino acid is inserted for a termination codon by the translational machinery and protein synthesis proceeds until another in-frame termination codon is reached (Hershey, 1987). If protein synthesis continued through the Eggg amber termination codon (facilitated by low abundance of RFl), it could continue to an in-frame RF2-specific UGA termination codon, the last two nucleotides of which overlap the AUG initiation codon of pgtg (Figure 6). Because there is not a nucleotide sequence with homology to the Shine-Dalgarno consensus sequence upstream of pttg, the translation of pgtg may be dependent upon reinitiation allowed by translational readthrough of the Eggg gene. According to this hypothesis, under conditions of sufficient or excess RFl protein in the cell, translation would be terminated at the amber codon, preventing pgtg translation. A putative promoter sequence for the pttt gene was identified by Craigen gt gt. (1985) (see Figure 6) in what has now been identified as the COOH-terminal region of the hemA coding sequence. There has been no demonstration that 67 hemA cggflgtgmgoo GAU AAC GAA CGC CUG AAEHAQQEQEG CGC R D G D N E R L N I L R GAC AGC CUC GGG CUG GAG UAG CAG ”AC AUC AUU UUC L DSLG EEnd has; ... ... . +1 prfA UUU UUU UAC AGG GUG CAU UUA CGC CU A AG CCU P Figure 6. Nucleotide sequence of the hemA-prfA intergenic region. The potential stem-loop structure in the RNA transcript is indicated with arrows and the run of uridine residues is denoted by dots. The promoter sequences proposed for the prfA gene are identified by dashed lines and the downstream hemA in-frame UGA termination codon discussed in the text is underlined twice. 68 the sequence functions as a promoter for transcriptional initiation, but 81 nuclease protection experiments should clarify this point. Another noteworthy feature of this intergenic region is the presence of sequences which resemble a tEg—independent terminator, having a possible hairpin structure followed by a series of U’s (Figure 6) (Yager and von Hippel, 1987). Whether this sequence terminates Eggg transcription has not been determined. With these cloned genes available, it will be possible to test the pttg translational reinitiation hypothesis and also to perform further experiments to study the transcriptional control of this putative operon. The proposed AUG codon of the open reading frame downstream of pttt overlaps the termination codon of pgtg (Chapter 2, Figure 3), which makes it a likely candidate for a third member of this genetic unit. The 5-carbon glutamate pathway for ALA biosynthesis appears to share features with protein biosynthesis, with the involvement of a tRNA and a glutamyl tRNA synthetase enzyme. Results of this study indicate that the Eggg gene probably encodes a second enzyme in the pathway, the dehydrogenase, which is genetically linked to an enzyme that is a component of the translational machinery of the cell. Further studies into the possible ties between these two fundamental cellular processes could provide informative data on overall metabolic regulation in E. coli. Chapter 4 Summary and Conclusions At the outset of this thesis project, it had been shown that the the E. ggtt,Eggt gene is required for ALA synthesis, but the initial genetic characterization of this locus did not allow a definitive assignment for its gene product. It was generally assumed that the ALA auxotrophy associated with the Eggg mutation was due to the absence of ALA synthase (Bachmann, 1983), the enzyme thought to catalyze ALA synthesis in bacteria. Data on ALA synthase activity in E. ggtt have been contradictory; some investigators have been unable to detect enzyme activity (McConville and Charles, 1979b; my results) while other laboratories have reported a disparate range of enzyme activities (Ishida and Hino, 1972; Schoenhart _t gt., 1986; Tai gt gt., 1988). These ambiguities could be explained if there were inherent difficulties in assaying the enzyme under the conditions of the tg yttgg assay, possibly due to its instability, requirement for activation, or the presence of inhibitors in bacterial extracts (Jacobs, 1978). Although it is rare for a fundamental metabolic precursor such as ALA to be synthesized via two independent biosynthetic pathways, Beale and Castelfranco (1974) discovered that a second enzymatic route exists in higher 69 70 plants that uses glutamate as the starting substrate for ALA formation. A scheme for the multi—enzymatic reaction sequence has since been proposed by several laboratories (Huang et a1., 1984; Huang and Wang, 1986; Bruyant and Hannagara, 1987; Weinstein gt gt., 1987). This route was originally thought to be limited to plants and algae, but recent studies indicate that it also operates in some eubacteria (Oh-hama gt gt., 1986; Oh-hama gt gt., 1988; Rieble and Beale, 1988; Avissar gt gt., 1989) and a species of archaebacteria (Friedman and Thauer, 1986). One of the specific goals of this thesis project was to evaluate the role of the Eggg gene product in ALA synthesis in E. ggtt. Chapter 2 details the molecular genetic techniques employed for the isolation, characterization, and DNA sequencing of the Eggg gene from E. ggtt. If the Eggg gene encoded ALA synthase, I expected that its nucleotide sequence would display significant similarity to other cloned ALA synthase genes, given that these sequences showed similarity across broad evolutionary boundaries. Unexpectedly, the E. ggtt Eggg gene bore no resemblance to the ALA synthase genes from other organisms. Furthermore, ALA synthase activity was not detectable in E. ggtt extracts from cells grown under a variety of culture conditions. Activity was also absent in cells which appeared to overproduce porphyrin compounds after being transformed with recombinant multicopy plasmids carrying the cloned hemA gene. A number of explanations could be 71 put forward to account for these results, the most satisfactory being that an alternate pathway for ALA synthesis normally operates in E. ggtt. Two laboratories have recently presented evidence that E. ggtt uses glutamate as the precursor for ALA formation via the 5-carbon pathway, rather than using glycine and succinyl-CoA as substrates for ALA synthesis via the ALA synthase reaction (Avissar and Beale, 1989; Li gt gt, 1989b). Avissar and Beale (1989) extended these studies to show that the enzyme activity absent in extracts of the Eggg strain SASX4lB is that of the dehydrogenase, which catalyzes the second reaction in the 5-carbon pathway, i.e., the reduction of activated glutamate to GSA. Extracts prepared from the Eggg strain aminoacylated tRNA61“ that was capable of acting as substrate for ALA formation and converted GSA to ALA by the aminotransferase reaction, demonstrating that the other enzyme activities in the reaction sequence were present. Contrary to previous assumptions, this biochemical characterization of the Eggg mutation and the results presented in Chapter 2 argue strongly that the E. ggtt Eggg gene encodes the dehydrogenase component of the 5-carbon pathway, rather than ALA synthase. However, the possibility that the Eggg gene product participates in the activation of the dehydrogenase or the regulation of its expression has not been excluded by myself or others. 72 The discovery of the 5-carbon pathway for ALA synthesis in E. ggtt has resolved the nature of the Eggt_mutation, first isolated twenty years ago, and has also opened many avenues for biochemical and genetic investigation. A thorough genetic analysis aimed toward the isolation and characterization of additional ALA auxotrophs is warranted because of the likelihood that there are other loci on the E. ggtt genome that affect this multi-enzymatic process. The png mutation which results in ALA auxotrophy, but maps to the 4 min region of the chromosome, requires further investigation in order to resolve the basis for the associated block in ALA synthesis. A mutation analogous to the png mutation has been isolated in E. typhimurium (Elliott and Roth, 1989) but, as in E. ggtt, the nature of the lesion has not yet been characterized. The identification of genes that affect the other components of the 5-carbon pathway might be complicated by the possibility that their actions may not be restricted to the ALA biosynthetic process. For example, if the same glutamyl-tRNA synthetase and tRNA‘1° participate in both ALA synthesis and protein translation, then mutations in these genes will not be specific for ALA synthesis. Whether the glutamyl-tRNA synthetase and the tRNAclu species that activates glutamate for its conversion to ALA are identical to those which participate in protein synthesis needs to be determined. There are four loci in E. coli (gltT, 90 min; gltU, 84.5 min; gltV, 90.5 min; and 73 gt_E, 57 min) that encode glutamyl-tRNA, all of which have the UUC anticodon (Fournier and Ozeki, 1985), and three loci (gttE, 81 min; gt_E, 43 min; gttt, 52 min) which affect the activity of glutamyl-tRNA synthetase (see references in Bachmann, 1983). Characterization of the tRNAGlu required for ALA synthesis has been investigated in other systems. In a study of the tRNA that participates in ALA formation in barley, where there is only one chloroplast-encoded gene for tRNAG1“, it was shown that the tRNA which supports ALA formation contains a highly modified UUC glutamate anticodon (Schon gt gt., 1986). Other glutamate-accepting tRNA species were identified but were shown to have glutamine anticodons (Schon _t _t., 1988). The glutamate on these mis-charged tRNA species is subsequently converted to glutamine through an amidotransferase reaction, with glutamine or asparagine as the amido donor. These glutamate-accepting tRNA species are not capable of participating in ALA synthesis, suggesting that the anticodon may be an important structural component of tRNA species that support ALA formation. Although the UUC anticodon is required, it is apparently not sufficient for activity, as demonstrated by the results of heterologous reconstitution experiments in which E ggtt tRNA°1“<""°) was capable of supporting ALA synthesis in isolated enzyme fractions prepared from Chlam domonas, but could not support ALA formation in extracts prepared from 74 Synechocystis, Chlorella or Euglena, even though these latter extracts could charge the tRNA with glutamate (Kannangara gt gt., 1988; O’Neill gt gt., 1988). Furthermore, studies using aplastidic mutants of Euglena indicate that plastid tRNAGI“, but not other cellular tRNA species, is effective in supporting ALA biosynthesis (Mayer _t _t., 1987). There must therefore be other constraints which determine whether the tRNA¢1° species can function in ALA synthesis, perhaps dictated by the ability of the dehydrogenase to recognize the glutamyl tRNA. In the blue—green alga Synechocygttg sp 6803, two fractions of tRNA°1“(“"°> have been resolved by HPLC that differ in their effectiveness to support ALA formation. Both were also capable of participating in protein synthesis, as determined by tg yttgg translation experiments (Schneegurt _t _t., 1988). This raises the question of how the charged tRNA‘“u is divided between the two processes of protein synthesis and tetrapyrrole synthesis and whether diversion of available glutamyl-tRNA into ALA biosynthesis is regulated. If the tRNA which supports ALA formation were processed post-transcriptionally by specific modifying enzymes, then the regulation of these enzyme activities could be involved in the control of overall tetrapyrrole synthesis. Further biochemical and genetic experiments are require before these question can be addressed in E. coli. 75 The glutamyl--tRNA‘”u formed in the first step of the reaction sequence is the substrate for the dehydrogenase, which catalyzes the reduction of the activated carboxyl group to yield GSA. The evidence that the Eggg gene in E. ggtt encodes this enzyme was discussed above. If the level of activity of this enzyme controlled the overall pathway, then an overproduction of this enzyme would be predicted to increase the formation of ALA in a cell. Li, gt gt. (1989b) have shown that an E. ggtt strain transformed by a multicopy plasmid containing the cloned Eggg gene excretes ALA into the medium. I have observed an accumulation of porphyrin compounds in similarly-transformed strains. In light of these results, the possibility that the regulation of dehydrogenase activity could be a control point for ALA and overall tetrapyrrole synthesis warrants investigation. This control could be exerted at several different levels, including activity of the enzyme (possibly by feedback inhibition by the end product or intermediate products in the pathway), translational regulation of the Eggg gene product, or control of transcription of the Eggg gene. As an initial step in characterizing Eggg expression, a detailed analysis of transcripts which originate in the region upstream of the Eggg gene was performed and results of these studies were presented in Chapter 3. It appears that the Eggg mRNA has two different 5’ ends and that an ORF of opposite polarity is transcribed upstream of the distal hemA transcriptional start site. 76 With this information available, it will be possible to perform a thorough quantitative analysis of Eggg transcription utilizing Sl nuclease or gene fusion techniques in order to determine if expression of Eggg is controlled by transcriptional regulation. One intriguing aspect of the characterization of the Eggg gene is its location directly upstream of pttg and the possibility that there is coordinate regulation of the two genes. This genetic link between tetrapyrrole biosynthesis and protein synthesis raises the question of whether regulation of these two fundamental cellular processes share controlling elements. Two other components of the translational machinery of the cell, tRNAGlu and glutamyl-tRNA synthetase also participate in ALA biosynthesis. Future investigation in this area may supply additional insight into the global regulation of cellular metabolism. LIST OF REFERENCES LIST OF REFERENCES Adams, T.H., and H.E. Chelm. 1988. Effects of oxygen levels on the transcription of nif and gln genes in Eradyrhtgobigg japonicum. J. Gen. Microbiol. 134:611-618. Adams, T.H., and B.E. Chelm. 1984. The nifH and nifDE promoter regions from Rhizobium japonicum share structural homologies with each other and with nitrogen-regulated promoters from other organisms. J. Mol. Appl. Genet. E:392-405. Adams, T.H., C.R. McClung, and B.K. Chelm. 1984. Physical organization of the Bradyrhizobium japonicum nitrogenase gene region. J. Bacteriol. 159:857-962. Aldea, M., V.F. Maples, and S.R. Kushner. 1988. Generation of a detailed physical and genetic map of the ilv—metE-udp region of the Eggherichig coli chromosome. J. Mol. Biol. 200:427-438. Alefounder, P.R., C. Abell, and A.R. Battersby. 1988. The sequence of hemC, hemD and two additional E. coli genes. Nucleic Acids Res. th9871. Avissar, Y.J., and S.I. Beale. 1989. Identification of the enzymatic basis for 5-aminolevulinic acid auxotrophy in a hemA mutant of Escherichia coli. J. Bacteriol. 171:2919-2924. Avissar, Y.J., J.G. Ormerod, and S.I. Beale. 1989. Distribution of 5-aminolevu1inic acid biosynthetic pathways among phototrophic bacterial groups. Arch Microbiol. 151:513-519. Bachmann, B.J. 1983. Linkage map of Escherichia coli K-12, Edition 7. Microbiol. Rev. 11:180-230. Bachmann, B.J. 1987. Derivations and genotypes of some mutant derivatives of Escherichia coli E-12, p.1190-1224. tg F.C. Neidhardt, J.L. Ingraham, H.E. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 2. American Society for Microbiology, Washington, D.C. 77 78 Bawden, M.J., I.A. Borthwick, H.M. Healy, C.P. Morris, B.E. May, and W.H. Elliott. 1987. Sequence of human 5-aminolevulinate synthase cDNA. Nucleic Acids Res. tE:8563. Beale, S.I. 1976. The biosynthesis of 5-aminolevulinic acid in plants. Phil. Trans. R. Soc. London B 273:99-108. Beale, S.I., and P. Castelfranco. 1974. The biosynthesis of 5-aminolevulinic acid in higher plants. Plant Physiol. EE:291-296. Beale, S.I., S.P. Cough, and S. Granick. 1975. Biosynthesis of 5-aminolevu1inic acid from the intact carbon skeleton of glutamic acid in greening barley. Proc. Natl. Acad. Sci. USA 1E:27l9-2723. Beale, S.I., T. Foley, and V. Dzelzkalns. 1981. Aminolevulinic acid synthase from Euglena gracilis. Proc. Natl. Acad. Sci. USA 1E:1666-1669. Beck, C.F., and R.A.J. Warren. 1988. Divergent promoters, a common form of gene organization. Microbiol. Rev. EE:3l8-326. Berk, A.J., and P.A. Sharp. 1977. Sizing and mapping of early adenovirus mRNAs by gel electrophoresis of 81 endonuclease-digested hybrids. Cell tE:721-732. Biel, S.W., M.S. Wright, and A.J. Biel. 1988. Cloning of the Rhodobacter capsulatug hemA gene. J. Bacteriol. 170:4382-4384. Borthwick, I.A., G. Srivastava, A.R. Day, B.A. Pirola, M.A. Snoswell, B.K. May, and W.H. Elliott. 1985. Complete nucleotide sequence of hepatic 5-aminolevulinate synthase precursor. Eur. J. Biochem. tE93481-484. Bottomley, 8.8., and G.A. Smithee. 1968. Characterization and measurement of 5-aminolevulinate synthetase in bone marrow cell mitochondria. Biochim. Biophys. Acta 159:27-37. Boyer, H.W., and D. Roulland-Dussoix. 1969. A complementation analysis of the restriction and modification of DNA in Escherichia coli. J. Mol. Biol. tt:459-472. Bruyant, P., and C.G. Eannangara. 1987. Biosynthesis of 5-aminolevulinate in greening barley leaves. VIII. Purification and characterization of the glutamate-tRNA ligase. Carlsberg Res. Commun. EE:99-109. 79 Burnham, B.F. 1970. 5-Aminolevulinic acid synthase. Methods Enzymol. t1:l95-204. Burnham, B.F., and J. Lascelles. 1963. Control of porphyrin biosynthesis through a negative-feedback mechanism. Biochem. J. Et:462-472. Casadaban, M.J., and S.N. Cohen. 1980. Analysis of gene control signals by DNA fusion and cloning in Escherichia coli. J. Mol. Biol. 138:179-207. Chopra, I. anle. Ball. Transport of antibiotics into bacteria. 1982. Advances in microbial physiology. EE:183-217. Clewell, D.B., and D.R. Helinski. 1972. Effect of growth conditions on the formation of the relaxed complex of supercoiled ColEl deoxyribonucleic acid and protein in Escherichia coli. J. Bacteriol. tttgll35-ll46. Cox, R. and H.P. Charles. 1973. Porphyrin-accumulating mutants of Escherichia coli. J. Bacteriol. 113:122-132. Craigen, W.J., R.G. Cook, W.P. Tate, and C.T. Caskey. 1985. Bacterial peptide chain release factors: conserved primary structure and possible frameshift regulation of release factor 2. Proc. Natl. Acad. Sci. USA EE:3616-3620. Davis, R.W., D. Botstein, and J.R. Roth. 1980. Advanced bacterial genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Echelard, Y., J. Dymetryszyn, M. Drolet, and A. Sasarman. 1988. Nucleotide sequence of the hemB gene of Escherichia coli H12. Mol. Gen. Genet. 214:503-508. Elhai, J., and C.P. Wolk. 1988. A versatile class of positive-selection vectors based on the nonviability of palindrome-containing plasmids that allows cloning into long polylinkers. Gene EE:119-138. Elliott, T. 1989. Cloning, genetic characterization, and nucleotide sequence of the hemA-p fA operon of Salmonella typhimurium. J. Bacteriol. 171:394873960. Elliott, T. and J.R. Roth. 1989. Heme-deficient mutants of Salmonella typhimurium: Two genes required for ALA synthesis. Mol. Gen. Genet. 216:303-314. Fournier, M.J., and H. Ozeki. 1985. Structure and organization of the transfer ribonucleic acid genes of Escherichia coli K-12. Microbiol. Rev. EE:379-397. 80 Friedmann, 8.0., and H.E. Thauer. 1986. Ribonuclease-sensitive 5-aminolevulinic acid formation from glutamate in cell extracts of MethangEgcterigg thermoautotrophicum. FEBS Lett. EEZ:84-88. Gillis, H., R. Jaenchen, and H.E. Thauer. 1983. Biosynthesis of 5-aminolevu1inic acid in Methanobacterium thermoautotrophicgg. Arch. Microbiol. 135:237-240. Gottschalk, G. 1986. Bacterial metabolism, second edition. Springer-Verlag, New York, New York. Gough, S.P., and C.G. Hannangara. 1977. Synthesis of 5-aminolevulinate by a chloroplast stroma preparation from greening barley leaves. Carlsberg Res. Commun. EE:459-464. Granick, S. and S.I. Beale. 1978. Hemes, chlorophylls, and related compounds: biosynthesis and metabolic regulation. Adv. Enzymol. EE;33-203. Guerinot, M.L., and B.K. Chelm. 1986. Bacterial 5-aminolevu1inic acid synthase activity is not essential for leghemoglobin formation in the soybean/Brady- rhizobium japonicum symbiosis. Proc. Natl. Acad. Sci. USA EE:1837-1841. Hadar, R., A. Slonim, and J. Kuhn. 1976. Role of o-tryptophan oxidase in o-tryptophan utilization by Eggherichtg coli. J. Bacteriol. 125:1096-1104. Hanahan, D. 1985. Techniques for transformation of E. coli, p.109-135. tg D.M. Glover (ed.), DNA Cloning, Vol.I, IRL Press, Oxford. Hawley, D.R., and W.H. McClure. 1983. Compilation and analysis of Escherichia coli promoter DNA sequences. Nucleic Acids Res. tt:2237-2255. Herrero, A., J. Elhai, B. Hohn, and C.P. Walk. 1984. Infrequent cleavage of cloned Anabaena variabilis DNA by restriction endonuclease from E. variabilis. J. Bacteriol. 160:781-784. Hershey, J.W.B. 1987. Protein synthesis, p.613-647. tg F.C. Neidhardt, J.L. Ingraham, H.E. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 1. American Society for Microbiology, Washington, D.C. Holmes, D.S., and M. Guigley. 1981. A rapid boiling method for the preparation of bacterial plasmids. Anal. Biochem. 114:193-197. 81 Hoober, J.R., A. Eahn, D.E. Ash, 8. Gough, and C.G. Eannangara. 1988. Biosynthesis of.Aaminolevulinate in greening barley leaves. IX. Structure of the substrate, mode of gabaculine inhibition, and the catalytic mechanism of glutamate-l-semialdehyde aminotransferase. Carlsberg Res. Commun. EEzll-25. Huang, D., and W. Wang. 1986. Chlorophyll biosynthesis in Chlamydomonas starts with the formation of glutamyl-tRNA. J. Biol. Chem. 261:13451-13455. Ishida, A., and S. Hino. 1972. Effect of oxygen on cytochrome pattern and heme synthesis in Escherichia coli. T J. Gen. Appl. Microbiol. Tokyo tE:225-237. Jacobs, N.J. 1974. Biosynthesis of heme, p. 125-148. tE_ g J.B. Nielands (ed.), Microbial iron metabolism. A ' comprehensive treatise. Academic Press, Inc., New York. * Jordan, P.M., B.I.A. Mgbeje, A.F. Alwan, and S.D. Thomas. 1987. Nucleotide sequence of hemD, the second gene in the hem operon of Escherichtg coli K-12. Nucleic Acids Res. tE:10583. Jordan, P.M., B.I.A. Mgbeje, S.D. Thomas, and A.F. Alwan. 1988. Nucleotide sequence for the hemD gene of Escherichia coli encoding uroporphyrinogen III synthase and initial evidence for a Egg operon. Biochem. J. EEE:613-616. Jordan, P.M., L. Marshall, and S. Thomas. 1986. The enzymes of uroporphyrinogen III biosynthesis, p. 83-93. tg Y. Nordmann (ed.), Porphyrins and porphyrias. Colloque INSERM/John Libbey Eurotext Ltd. Kannangara, C.G., S.P. Gough, P. Bruyant, J.R. Hoober, A. Hahn, and D. von Wettstein. 1988. tRNA‘“u as a cofactor in 5-aminolevulinate biosynthesis: steps that regulate chlorophyll synthesis. TIBS thl39-l43. Hannangara, C.G., S.P. Gough, H.P. Oliver, and S.R. Rasmussen. 1984. Biosynthesis of 5-aminolevulinate in greening barley leaves. VI. Activation of glutamate by ligation to RNA. Carlsberg Res. Commun. 32:417-437. Hawakami, H., I. Toshifumi, and Y. Nakamura. 1988. Conditionally lethal and recessive UGA-suppressor mutations in the prfB gene encoding peptide chain release factor 2 of Escherichia coli. J. Bacteriol. t1E:5378-5381. Heng, T., E. Alani, and L. Guarante. 1986. The nine amino-terminal residues of 5-aminolevulinate synthase direct B-galactosidase into the mitochondrial matrix. Mol. Cell. Biol. E:355-364. 82 thara, T., E. Akiyama, and H. Isono. 1987. The physical map of the whole E. coli chromosome: application of a new strategy for rapid analysis and sorting of a large genomic library. Cell EE:495-508. Eredich. N.M. 1987. Biosynthesis of cysteine, p. 412-418. tg F.C. Neidhardt, J.L. Ingraham, E.B. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 1. American Society for Microbiology, Washington, D.C. Lascelles, J. 1964. Tetrapyrrole biosynthesis and regulation. W.A. Benjamin Inc., New York. Lee, C.C., Y. Kohara, H. Akiyama, C.L. Smith, W.J. Craigen, and C.T. Caskey. 1988. Rapid and precise mapping of the . Escherichia coli release factor genes by two physical approaches. J. Bacteriol. tZE:4537-4541. Leong, S.A., G.S. Ditta, and D.R. Helinski. 1982. Heme biosynthesis in Rhizobium. Identification of a cloned gene coding for 5-aminolevulinic acid synthetase from Rhizobium meliloti. J. Biol. Chem. EE1:8724-8730. Li, J., C.S. Russell, and S.D. Cosloy. 1989a. The structure of the Egcherichia coli hemB gene. Gene 1E3177-184. Li, J., H. Umanoff, R. Proenca, C.S. Russell, and S.D. Cosloy. 1988. Cloning of the Egchgrichig coli E-12 hemB gene. J. Bacteriol. 170:1021-1025. Li, J., O. Brathwaite, S.D. Cosloy, and 0.8. Russell. 1989b. 5-Aminolevulinic acid synthesis in Escherichia coli. J. Bacteriol. 171:2547-2552. Lin, E.C.C., and D.R. Kuritzkes. 1987. Pathways for anaerobic electron transport, p. 201-221. tg F.C. Neidhardt, J.L. Ingraham, H.E. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 1. American Society for Microbiology, Washington, D.C. Lowry, C.R., N.J. Rosebrough, A.L. Farr, and R.J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Maguire, D.J., A.R. Day, I.A. Borthwick, G. Srivastava, P.L. Wigley, B.K. May and W.J. Elliott. 1986. Nucleotide sequence of the chicken 5-aminolevu1inic acid synthase gene. Nucleic Acids Res. t1:1379-1391. 83 Maniatis, T., E.F. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory. Cold Spring Harbor, N.Y. Marmur, J., and P. Doty. 1962. Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J. Mol. Biol. E:109-118. Marver, H.S., A. Collins, D.P. Tschudy, and J. Rechcigl. 1966. 5-Amin01evulinic acid synthetase. II. Induction in rat liver. J. Biol. Chem. 241:4323-4329. Maxam, A.M., and W. Gilbert. 1980. Sequencing end-labeled DNA with base-specific chemical cleavages. Methods Enzymol. EE:499-560. Mayer, S.M., J.D. Weinstein, and S.I. Beale. 1987. Enzymatic conversion of glutamate to 5-aminolevulinate in soluble extracts of Euglena gracilis. J. Biol. Chem. EEE:12541-12549. McClung, C.R., J.E. Somerville, M.L. Guerinot, and D.R. Chelm. 1987. Structure of the Bradyrhizobium japonicum gene hemA encoding 5-aminolevulinic acid synthase. Gene E33133-139. McConville, M.L., and H.P. Charles. 1979a. Isolation of haemin-requiring mutants of Escherichia coli E12. J. Gen. Microbiol. 113:155-164. McConville, M.L., and H.P. Charles. 1979b. Mutants of Escherichia coli E12 accumulating porphobilinogen: a new locus, hemC. J. Gen. Microbiol. 111:193-200. McConville, M.L., and H.P. Charles. 19790. Mutants of Escherichia coli E12 permeable to haemin. J. Gen. Microbiol. 113:165-168. Messing, J. 1983. New M13 vectors for cloning. Methods Enzymol. 101:20-78. Messing, J., R. Crea, and P.A. Seeburg. 1981. A system for shotgun DNA sequencing. Nucleic Acids Res. E:309-321. Miller, J.R. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Oh-hama, T., H. Seto, and S. Miyachi. 1986. 13C-NMR evidence of bacteriochlorophyll a formation by the Cs pathway in Chromatium. Arch. Biochem. Biophys. 246:192-198. Oh-hama, T., N.J. Stolowich, and A.I. Scott. 1988. 5-Aminolevulinic acid formation from glutamate via the C5 84 pathway in Clostridium thermoaceticum. FEBS Lett. 228:89-93. O’Neill, G.P., D.M. Peterson, A. Schon, M. Chen, and D. 8011. 1988. Formation of the chlorophyll precursor 5-aminolevulinic acid in cyanobacteria requires aminoacylation of a tRNA°‘“ species. J. Bacteriol. t1E:3810-3816. Poncz, M., D. Solowiejczyk, M. Ballantine, E. Schwartz, and S. Surrey. 1982. "Nonrandom" DNA sequence analysis in bacteriophage M13 by the dideoxy chain-termination method. Proc. Natl. Acad. Sci. USA tE:4298-4302. Poole, H.E., and W.J. Ingledew. 1987. Pathways of electrons to oxygen, p. 170-200. tg F.C. Neidhardt, J.L. Ingraham, H.E. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 1. American Society for Microbiology, Washington, D.C. Poulson, R., N.J. Whitlow, and W.J. Polglase. 1976. Catabolite repression of protoporphyrin IX biosynthesis in Escherichia coli E-12. FEBS Lett. EE:351-353. Powell, K.A., R. Cox, M. McConville, and H.P. Charles. 1973. Mutations affecting porphyrin biosynthesis in Escherichia coli. Enzyme tE:65-73. Prentki, P., and H.M. Erisch. 1984. In vitro insertional mutagenesis with a selectable DNA fragment. Gene EE;303-3l3. Pustell, J. and F.C. Kafatos. 1984. A convenient and adaptable package of computer programs for DNA and protein sequence management, analysis and homology determination. Nucleic Acids Res. tE:643-655. Rieble, 8., and S.I. Beale. 1988. Transformation of glutamate to 5-aminolevulinic acid by soluble extracts of Synechocystis sp. PC06803 and other oxygenic prokaryotes. J. Biol. Chem. EEE:8864-8871. Ryden, M., J. Murphy, R. Martin, L. Isaksson, and J. Gallant. 1986. Mapping and complementation studies of the gene for release factor 1. J. Bacteriol. 168:1066—1069. Sancar, A., A.M. Hack, and W.D. Rupp. 1979. Simple method for identification of plasmid-coded proteins. J. Bacteriol. 137:692-693. 85 Sanger, F., S. Nicklen, and A.R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 11:5463-5467. Sasarman, A., A. Nepveu, Y. Echelard, J. Dymetryszyn, M. Drolet, and C. Goyer. 1987. Molecular cloning and sequencing of the hemD gene of Excherichia coli E-12 and preliminary data on the Uro operon. tEE:4257-4262. Sasarman, A., M. Surdeanu, G. Szegli, T. Horodniceanu, V. Greceanu, and A. Dumitrescu. 1968a. Hemin-deficient mutants of Escherichia coli E-12. J. Bacteriol. EE:570-572. Sasarman, A., M. Surdeanu, and T. Horodniceanu._1968b. Locus determining the synthesis of 5-aminolevu1inic acid in Escherichia coli H-12. J. Bacteriol. EE:1882-1884. Sasarman, A., P. Chartrand, R. Proschek, M. Desrochers, D. Tardif, and C. Lapointe. 1975. Uroporphyrin-accumulating mutant of Escherichia coli E-12. J. Bacteriol. tgg;1205—1212. Sasarman, A., and T. Horodniceanu. 1967. Locus determining normal colony formation on the chromosome of Escherichia coli K-12. J. Bacteriol. 21:1268-1269. Sasarman, A., Y. Echelard, J. Letowski, D. Tardif, and M. Drolet. 1988. Nucleotide sequence of the hemX gene, the third member of the Uro operon of Escherichia coli E12. Nucleic Acids Res. tE;ll835. Schneegurt, M.A., and S.I. Beale. 1988. Characterization of the RNA required for biosynthesis of 5-aminolevulinic acid from glutamate. Plant Physiol. EE:497-504. Schneegurt, M.A. , S. Rieble, and S. I. Beale. 1988. The tRNA required for in vitro 5- aminolevulinic acid formation from glutamate in Synechocystis extracts. Plant Physiol. 88. 1358- 1366. Schoenhaut, D.S., and P.J. Curtis. 1986. Nucleotide sequence of mouse 5-aminolevulinic acid synthase cDNA and expression of its gene in hepatic and erythroid tissues. Gene gE:55-63. Schon, A., C.G. Eannangara, S. Gough, and D. 8011. 1988. Protein biosynthesis in organelles requires misaminoacylation of tRNA. Nature. 331:187-190. Schon, A., G. Krupp, S. Gough, S. Berry-Lowe, C.G. Eannangara, and D. 8011. 1986. The RNA required in the first step of chlorophyll biosynthesis is a chloroplast glutamate tRNA. Nature (London) EEE:28l-284. 86 Southern, E. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. EE:503. Spears, P.A., D. Schauer, and P.E. Orndorff. 1986. Metastable regulation of type 1 piliation in Escherichia coli and isolation and characterization of a phenotypically stable mutant. J. Bacteriol. tEE:179-185. Tai, T., M.D. Moore, and S. Eaplan. 1988. Cloning and characterization of the 5-aminolevu1inate synthase gene(s) from Rhodobggter sphaeroides. Gene 1E:l39-l5l. Thomas, S.D., and P.M. Jordan. 1986. Nucleotide sequence of the hemC locus encoding porphobilinogen deaminase of Escherichia coli K12. Nucleic Acids Res. tt:6215. Umanoff, H., C.S. Russell, and S.D. Cosloy. 1988. Availability of porphobilinogen controls appearance of porphobilinogen deaminase activity in Escherichia coli K-12. J. Bacteriol. 170:4969-4971. Wang, W., D.D. Huang, D. Stachon, S.P. Gough, and C.G. Hannangara. 1984. Purification, characterization and fractionation of the 5-aminolevulinic acid-synthesizing enzymes from light-grown Chlamydomonas reinhardtii cells. Plant Physiol. 13:569-575. Wang, W., S.P. Gough, and C.G. Kannangara. 1981. Biosynthesis of 5-aminolevulinate in greening barley leaves. IV. Isolation of three soluble enzymes required for the conversion of glutamate to 5-aminolevulinate. Carlsberg Res. Commun. EE:243-257. Weinstein, J.D., and S.I. Beale. 1983. Enzymatic conversion of glutamate to 5-aminolevulinic acid in soluble extracts of Euglena gracilis. J. Biol. Chem. 258:6799-6807. Weinstein, J.D., and S.I. Beale. 1985. Enzymatic conversion of glutamate to 5-aminolevulinate in soluble extracts of the unicellular green alga, Chlorella vulgaris. Arch. Biochem. Biophys. EE1:454-464. Weinstein, J.D., S.M. Mayer, and S.I. Beale. 1987. Formation of 5-aminolevu1inic acid from glutamic acid in algal extracts. Plant Physiol. Eg:244-250. Winans, S.C., S.J. Elledge, J.R. Erueger, and G.C. Walker. 1985. Site-directed insertion and deletion mutagenesis with cloned fragments in Escherichia coli. J. Bacteriol. 161:1219-1221. 87 Yager, T.D., and P.H. von Hippel. 1987. Transcript elongation and termination in Escherichia coli, p.1241-1275. tg F.C. Neidhardt, J.L. Ingraham, E.B. Low, B. Magasanik, M. Schaechter, and H.E. Umbarger, (ed.), Escherichia coli and Salmonella typhimurium: cellular and molecular biology, vol. 2. American Society for Microbiology, Washington, D.C. Yamamoto, M., N.S. Yew, M. Federspiel, J.D. Dodgson, N. Hayashi, and J.D. Engel. 1985. Isolation of recombinant cDNAs encoding chicken erythroid 5-aminolevulinate synthase. Proc. Natl. Acad. Sci. USA EE:3702-3706. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the Ml3mpl8 and pUCl9 vectors. Gene EE:103-119. HICHIGQN STRTE UNIV. LIBRQRIES llll lllll 1 7 312930078 5 43