Gene.86 (1990) 107-111
107
Elsevier G E N E 03375
Comparison of proC and other housekeeping genes of P s m t d o m o u s a e m g i w s a with their c o u t m p a r t s in EschericMa coil (Recombinant DNA; ~l'-pyrroline 5-carboxylate reductase; proline; nucleotide sequence; evolution; 16S rRNA; enzyme purification)
Armand Savioz% David J. Jeenes"*, Hans P. Kocher b and Dieter Haas" ° Mikrobiologbches institut, Eidgen~sbche Technbche Hochschule, CH-8092 Z ~ c h (S~t:erland), and b Sando: AG, CH-4002 Base/ ($wit:erland) Tel. (061)J2411 ! ! Received by J. Davison: 13 June 1989 Revised: 20 September 1989 Accepted: 22 September 1989
SUMMARY
In a comparative study of housekeeping genes of Pseudomonas aemginosa and Escherichta colt, the nucleotide sequence of a proline biosynthetic gene, pvoC, of P. aemginosahas been determined. The subunit molecular mass ( ~ 29 kDa) and the N-terminal amino acid sequence of purified ,Jt-pyrroline $-carboxylate reductase, the proC gene product, were in agreement with the proC nucleotide sequence. A survey of pairs of isofunctional genes from P. aeruginosaand E. coltreveals that within each pair, translated genes (including proC) have diverged more strongly than have untranslated genes specifying ribosomal or transfer RNAs. The translated genes, but not the untranslated ones, have a O + C content that is typical of the respective genomic O + C contents.
INTRODUCTION
Pseudomonasaemginosaand Esckerlchlacoil, members of the ? subgroup of'purple' bacteria, differ in a large number of metabolic properties, as well as in G + C content (Woese et al., 1985; Palleroni, 1986; Woese, 1987). We have begun to compare biosynthetic genes (argF,proC, argA)from both organisms (Jeenes et al., 1986). These genes, when carried by multi-copy plasmids, complement the corresponding Correspondence to: Dr. D. Haas, MikrobioloBisches lnstitut, ETHZentrum, CH-8092 Ztlrich (Switzerland) Tel. (01)256 3315; Fax (01)2529613. * Present address: AFRC Institute of Food Research, Colney Lane, Norwich, NR4 7UA (U.K.) Tel. 0603-56122. Abbreviations: an, amino acid(s); bp, base pair(s); kb, kilobase(s) or 1000 bp; Km, kanamycin; at, nucleotide(s); ORF, open reading frame; PSC, At-pyrroline 5-carboxylate; proC, gene encoding P$C reductase; a, resistant; RBS, ribosome-binding site; rRHA, ribosomal RNA; UWGCG, University of Wisconsin Genetic Computer Group; wt, wild type; [ ], denotes plasmid-earrier state. 0378-1119/901503.50© 1990Elsevier SciencePublishersB.V. (BiomedicalDivision)
mutations in the other bacterium. The level at which these genes are expressed in the heterologous host largely depends on the functioning of their promoters (Jeenes et al., 1986; Soldati et ai., 1987; ltoh et al., 1988). The ornithine carbamoyitransferase genes ofP. aemginosa(argF) and E. colt (argl) have 45% nt sequence identity (ltoh et al., 1988). This value is considerably lower than the identity of the 16S rRNAs from both bacteria, viz. 85% (Brosius et al., 1981; Toschka et al., 1988; Table I). In order to obtain a general measure of sequence conservation between P. aemginosa and E. colt, we have decided to investigate the degree of similarity between housekeeping genes having the same function in both organisms, with special reference to the different G + C contents of translated and untranslated genes. The specific aim ofthe present study was to determine the ntsequence of a proline biosynthetic gene, proC, of P. aeruginosa, so that this gene could be included in the survey of comparable E. colt and P. aeruginosa genes. Furthermore, the published method (Krishna et al., 1979)
108 m
EXPERIMENTAL A N D D I S C U S S I O N
(a) Nueleotide sequence of the Pseudomoaas aerugiaosa
pvoC gene
'KmR )-..----,e.
'
~kb
P. aeruglnosa PAO969, proC mutant, is complemented by pME58 (Fig. 1; Jeenes et al., 1986). This plasmid contains a 2.2-kb EcoRI-PmlI chromosomal fragment from the wt P. aerug~nosa PAOI in the broad-host-range vector pKT240 and allows high-level expression of PSC reductase in P. aerug~nosa (Jeenes et al., 1986). A pME58 derivative, pME74 (Fig. 1), carries the proC structural gene but not the proC promoter(s) (Jeenes et al., 1986; Soldati et al., 1987). The nt sequence of the 1.2-kb XhoI-Pmll insert of pME74 has now been determined (Fig. 2) on both strands, mostly by the chain-termination method as previously described (Baur et al., 1987; 1989; Itoh et al., 1988) and, in part, by the chemical degradation method (Maxam and Gilbert, 1980). The proC ORF has a coding capacity
O0.6kb)
:
Fig. I. Physicalmap ofthe proCgcneregionofP. aeruginosa.The restrictionfragmentsshownby a heav~lineare derivedfromthe PAO1chromosomeand insertedintopKT240(Jeeneset al., 1986).Gene proCis shown by an arrow. ~ indicates the Xhol deletion producing pME74 from pME58.Thisdeletioninactivatesthe Kma determinant.The sequenceof' the Xhol.Pvull fragmentis shownin Fig.2. X and Y are parts of ORFs with -__-k-ownfunctions (see Fig.2). for the purification of P5C reductase (EC 1.5.1.2), the proC gene product, w~s improved to provide an enzyme preparation that was suitable for N-terminal aa analysis.
XhoI
KpnI
~S
c c A Qc ccATca cG G T c c T Tcc c Tc TAcc ccT TTcQQ acccGc Ac Ac c cGc G cTTc TG a Tccc L
E
k
M ~
A
A
Z
S
l
E
G
~
D P
E
I
G
~
G A
A
~
E
T
i
~
W
~
R
V
~
A
R
~
K
l
~
I
G
T
M A
A
k
A
L
S
F
L
G
Z
A
R
G G
D
L
¥
R
G
A
A
P
A
G V
S
P
*
A
A
Z
R
(33)
G E
F
A
I
D
V
V
E
S
N
A
E
A
V
A
D
A
O V
V
G
H T
k
k
I
¥
I
L
9
T
L
G
S
L
¥
L
L
V
3oo (6?)
us
A E
L
G
* Ps~ l
M I
L
I
TGcTGcTGcGcTT:&T~G~&~T~GGTc~G~&cT;~&~ccC~c>Tc&TcGT~GcG~c~GCcGCTGc~G~cCcGcTGcG L
L
g
F
I
L
L
V
R
A
D
F
¥
.
p
L
S
.
F
i
V
R
A
T
K
p
L
L
N
p
L
1100
R
. vL~ll
R
I
I
P
G F
G G
l
D L
k
S
F~. 1 ~ e nt s ~ u e n ~ and d ~ u c e d aa s ~ u e n c e o f ~ e S t m i n ~ a t e ~ e stop c ~ o n s . N ~ r s in brackets ~ r
L
V
L
k
I
L
l
9
L
p~C ~ n e o f ~ aerugin~a. ~ e aa ~ s ~ u e s c o n f l r m ~ ~ protein s~uencm~ are ~ d i n e d . ~ aa ~ d u . .
109 for 273 aa. The codon usage (not shown) is typical of P. aeruginosa (West and Iglewski, 1988). Upstream and downstream from the proC gene, ORFs were found (Fig. 2) whose functions are unknown. Although the proC chromosome region of strain PAO l has been mapped extensively by transduction (Rella and Haas, 1982; Fyfe and Govan, 1984), classical genetic methods have revealed no other marker and in particular no other pro gene in the immediate vii:inity of preC. Since the N-terminal sequence ofPSC reductase has been determined (section b), the translation initiation region ofthe proCgene can be localized; a 5 bp stretch preceding the first AT(} of the proC gene is complementary to the 3' end of P. aeruginosa 16S rRNA (Fig. 2). (b) Purification and N-terminal amino acid analysis of PSC reductase from Pseadomoaas aerugiaosa Previously, PSC reductase had been partially purified from the wt PAOI (Krishna et al., 1979). This enzyme was now obtained as a nearly hom*ogeneous preparation by the combined use of an overproducing P. aeruginosastrain and two additional purification steps, i.e., hydrophobic and affinity chromatography. PSC reductase activity was assayed according to Krishna et al. (1979), and protein concentrations were estimated by the Lowry method. One unit is defined as the amount of enzyme that oxidizes 1 #mol NADH/min. Strain PAO969[pME$8] (Jeenes et al., 1986) was grown with aeration at 37°C in nutrient yeast broth containing 300/Ag Km/mi to approx. 109 cells/ml. Cells (35 g wet weight) were suspended in buffer 1 (50 mM imidazole, HCI
pH 6.9/0.5 mM dithiothreitol/l mM phenylmethylsulfonyl fluoride/10% (v/v) glycerol; Krishna and Leisinger, 19"/9) and broken by sonication. This and all subsequent operations were cmT~ed out at 4°(:, and buffer I was used throughout unless specified otherwise. After centrifugation at 40000 × g for 45 mln, the crude extract (specific activity 5 units/rag) was fractionated by the addition of solid ammonium sulfate. The 40-70~ fraction was dialyzed and applied to a DEAE cellulose (Whatman) column (3 × 12 cm). ARer washing the column, the enzyme was elated by a linear 0-0.35 M KCI gradient (680 ml). The fractions containing the highest activities were pooled and transferred to a phenyl-sepharose (Pharmacia) column (2 × 8 cm). The enzyme was desorbed by a linear 0-50% (v/v) ethylene glycol gradient (200 ml), dialyzed and concentrated with a PMI0 ultrafdtration membrane (Amicon). This fraction was diluted with buffer I not containing glycerol to a f'mal glycerol concentration of 2.5 % (v/v) and applied to a Matrex Gel RedA (Amlcon) column (2 ml bed) in the same buffer. PSC reductase was elated with a linear 0-10 mM NAD ÷ gradient (40 ml) in this buffer; the most active fractions were combined and concentrated by ultraf'dtration. The specific activity ofthis preparation was about 1250 units/rag, corresponding to an overall 2$0-fold purification. During the f'mal purification step the enzyme was unstable, resulting in a low total yield (5 ~o). According to SDS-PAGE (Coomassie blue staining), the preparation contained < $Yo impurities, and the P$C reductase polypeptide comigrated with the 28.7-kDa carbonic anhydrase standard (data not shown). Thus, the predicted size (28.2 kDa) and the experimental value were in good agreement.
MER~CGNMGKAILGGLI~S~Q~LP~W~YT~SP~KV~HD~b.~MSTP~F~A~S~ZGG~RAQG-~PA~IRASDPGAEQB~KI~GF~AMNLGIZGYGNIGELLSQNZZSHD-FCNLNRLYIANRTLSKINHLKDIDP
50 50 48
P.a. M.s.
-IN~AQ.E~QI~.I~L~A.~(~G~IR~LSE~TSS~NK~S~T -~D~V~S~.~ADA~VVVLSVKPQA~KA~CQA~A~ALKPEQL~SIA~IP RISITDDNIEVAKTCEKIIISVKTPDLANVLDPLKPHITKNTQIIHTCAGTD
I01 I01 i00
P.a. M.s.
LDQ~ARALGHDI~KI~.RAMPNTPALVNA~I~SXTPNALV~PEDTAD~rLNIFRC CASL~AW~GQPR~V~RCMPNTPALLBQ~G~YAN~C~Q~GQ~A LEFKDCGLSCVIPTISSTYDEDN-PKKGVSIIMHDENVSGENREFVEKLFSR
153 153 151
E,c,
P.a. M.s. E,C,
E,C,
P.a. M.s.
E~E~E-~IA~PM~HP~E'VGVSGSSPAYVFMFIE~A~-~VLG.~AQ~YKF203 ~GI~LW~.DE~QIDAVTAVSGSG~LZQAMTDA-GEKLG~TASRL 204 FSQIKVVDSPMDLEIATIAASCMPAFIALGVDLfa*gELEEKCNLSKEETFKI 203
P.a. M.s.
AA~A..~SAK_~ETGEHP_~AL_KDMVC SPGGTT IE A _ ~ V L E E K G F R ~ I EA 255 TLQ~b.~LGA~Q~LSSEVEPAELRR~TSPNGTTEA~KSFQANGFEA_LVEQA 256 LAETLNSTAY ILKED IYSPDELINKVATKNG ITQKGLDVLDKRPARYL 251
E.c. P.a.
MTKCMEI(~.EKLSK S LNAASQRSAELAEQLGQ
E,c,
269 273
Fig. 3. The proC genes of P. aeruginosa, E. coli and M. smithU. Comparison of deduced aa sequences of PSC reductases by the Gap program (UWGCG). The aa numbers ofP. aemginosa PSC reductase relate to Fig. 2. M.s., M. smithli; E.c., £. coil; P.a., P. aerugtnosa. Bold lines under E.c. residues and above M.s. residues indicate identity with P.a. residues; thin lines signify similarity (defined previously by Baur et al., 1987) with P.a. residues.
110
In a literature search we found 14 gene pairs which satisfy the above conditions (Table I). This survey leads us to several observations. Seven pairs of untranslated genes coding for structural RNAs are more strongly conserved than are seven pairs of translated genes specifying biosynthetic enzymes or the RecA protein. The genomic G + C contents, approx. 50% in E. coil and 67% in P. aeruginosa (West and Iglewski, 1988), are roughly reflected by the translated genes but not by the untranslated genes of both bacteria. This situation is not unique and presumably results from the structural constraints in the rRNAs and tRNAs. Muto and Osawa (1987) have previously noted that in eubacteria the G + C content of protein genes is strongly correlated with the genomic G + C content; in contrast, there is only a weak correlation of the G + C content of both rRNA and tRNA genes to genomic G + C content. The proC genes and five pairs of other biosynthetic genes inspected (Table I) display less sequence identity at the aa level than at the nt level, despite the fact that the different G + C contents of P. aerufinosa and E. coli per se result in a number of 'silent' nt substitutions which do not change the aa specified. For instance, 52 silent substitutions occur in the proC pair. Within enteric bacteria (E. coli, Salmonella, Klebsieila) the situation is different in that hom*ologous biosynthetic genes show more aa than nt sequence identity (Yanofsky, 1984; Van Vliet et al., 1988). Although the number of gene pairs analysed is still relatively small, we conclude that considerable sequence
The N-terminal aa sequence of P5C reductase was determined with an Applied Biosystems 470A protein sequencer according to standard procedures (Hochuli et al., 1987). The sequence (STPRIAFIGAGN) matched that predicted from the nt sequence (Fig. 2), except for the N-terminal methionine~ which appears to be cleaved off post-translationally. (e) Comparison of the Pseudomonas aeruginosa PSC reductase with s|m__|l_ar enzymes from other microorganisms The deduced aa sequence of the P5C reductase from P. aeruginosa had 34% identity (and 56% similarity) with the P5C reductase from E. coli (Deutch et al., 1982) and 17% identity with a similar enzyme from Methanobrevibacter smithii (Hamilton and Reeve, 1985). (d) Comparison of housekeeping genes of Pseudomonas aeruginosa and Escherichia coil We have chosen to compare pairs of genes that meet the following criteria: (i) the genes within each pair have been shown or are very likely to be functionally equivalent in P. aeru~nosa and in E. coli; (ii)the genes have no established record of lateral transmission; they are not carried by plasmids, phages, or known transposable elements; (iii) the entire nt sequence has been determined; (iv)the genes within each pair are of similar length and permit sequence alignment by current computer programs.
TABLE ! Surveyof pairs of i8ot~nctionaihousekeepinggenes whose sequenceshave been determined in Pseudomonasaeruftnosa and £schertchlacolt Oene
% O + C in
% Identity within pair
References
E, coil
P. aerugfnosa
nt sequence
aa sequence
53 55 54 53
80a 70a 85b 83b
._
tRNA °~
63 60 54 53 57
$7
89"
--- "~
tRNAT M
$4
54
---
Hughes and Jones (1988)
tRNATyr recA trpA trpB trpO argF° lysA proC
59 54 48 55 55 52 53 55
50 64 69 67 67 67 67 71
86" 87" 71b
Sano and Kapyama (1987)
47 b
-71~ 34d ]
60b
52 d
53 d
$2© 32G 31d 34b
5S rRlqA 6S RNA 16S rRNA 23S rP.NA
45c 44b 48b
m
_. __
Brosius et al. (1981); Housiaux et al. (1988) Vogel et al, (1987) Brosius et al. (1981); Toschka et al. (1988) Brosius ©t al. (1981); Toschka et al. (1987)
Yanofsky et al.(1981); Hadero and Crawford (1986) Yanofsky et al. (1981); Crawford and Eberly (1986) Itoh et al. (1988) Martin et ol. (1988) Deutch et al. (1982); this study
" Obtainedby the Bestfit program (UWGCG) b Obtainedby the Gap program(UWGCG); Gap was used whenevera better fit was obtained than with Bestfit. ©Thesevalues are reported in the referencescited. d Deducedfrom published aa sequencealignment. ° The hom*ologous£. coA~g~ne is awl
111
changes have occurred in proC and other translated housekeeping genes of Pseudomonas and E. coli, whereas the primary structures of untranslated RNA genes are conserved more strongly. These fmdings concur with Stackebrandt's (1988) observation that during the aerobic phase bacteria evolved rather rapidly and acquired highly divergent phenotypes while the sequences of rRNA genes did not change dramatically. Consequently, 16S rRNA analysis tends to overestimate phylogenetic relationships of these bacteria (Stackebrandt, 1988) and additional molecular criteria may. be needed to refine taxonomic classification. ACKNOWLEDGEMENTS
We thank Hedi Strahm for competent technical assist° ance, Jean-Claude Patte, Hauke Hennecke and Jean-Pierre Bachellerie for discussion, and H61~ne Paul for secretarial assistance. This project was supported by the BonizziTheler foundation (A.S.) EMBO (D.J.J.) and the Schweizerische Nationalfonds (project 3.583-0.87). REFERENCES Baur, H., Stalon, V., Falmague, P., L0thi, E. and Haas, D.: Primary and quaternary structure of the catabolic ornithine carbamoyltransferase from Pseudomonas aer~rinosa. Eur. J. Biochem. 166 (1987) ! 11-117. Baur, H., Lttthi, E., Stalon, V., Mercenier, A. and Haas, D.: Sequence analysis and expression of the arginine-deiminase and carbamatekinase genes ofPseudomonas aerug~nosa. Eur. J. Biochem. 179 (1989) $3-60. Brosius,J.,Dull,TJ., Sleeter,D.D. and Noller,H.F,:(3eneorganization and primary structure of a ribosomal RNA operon from £sche~chla coil J. Mol. Biol, 148 (1981) 107-127, Crawford, I,P, and Eberly, L.: Structure and regulation ofthe anthranilate synthase genes in Psoudomonasaemginosa: 1. SequenceoftvpG encoding the giutamine amidotransferase subunit. Mol. Biol. Evol. 3 (1986) 436-448. Oeutch, A.H., Smith, CJ., Rushlow, K,E. and Kretschmer, PJ.: Escherichla coil dt-pyrroline-5-carboxylate reductase: gone sequence, protein overproduction and purification. Nucleic Acids Reg. 10 (1982) 7701-7714. Fyfe, J.A.M. and Govan, J.R.W.: Chromosomal loci associated with antibiotic hypersensitivity in pulmonary isolates of Pseudomonas aevuginosa. J. Gan. Microbiol. 130 (1984) 825-834. Hadero, A. and Crawford, I.P.: Nuclcotide sequence of the genes for tryptophan synthase in Pseudomonas aerugtnosa. Mol. Biol. Evol. 3 (1986) 191-204. Hamilton, P.T. and Reeve, J.N.: Structure of genes and an insertion element in the methane producing archaebacterium Methanobrevibatter smithU. Mol. Gen. Genet. 200 (1985) 47-59. Hochuli, E., Gillessen, D. and Kocher, H.P.: Specificityof the immunoadsorbent used for large-scalerecovery ofinterferon-2a. J. Chromatography 411 (1987) 371-378. Housianx, P.J., Hill, D.F. and Petersen, G.B.: Nocleotide sequence of a gone for 5S ribosomal RNA from Pseudomonas aeruginosa. Nucleic Acids Res. 16 (1988) 2722. Hughes, M.A. and Jones, D.S.: A fragment ofthe Pseudomonasaeruginosa genome contains five tRNA genes, four of which are linked to an EF-Tu gone. Nucleic Acids Res. 16 (1988) 7193.
ltoh, Y., Soldati, L, Stalon, V., Falmagne, P., Terawaki, Y., Lcisinger,T. and Haas, D.: Anabolic omithine carbamoyi-transfera~ of monas aemginosa: Nuclcotide sequence and transcriptional control of" the argF structural 8one. J. Bacteriol. 170 (1988) 2725-2734. Jcenes, DJ., Soldati, L, Baur, H., Watson, J.M., Mercenicr, A., Reimmann,C., Lcisinger,T. and Haas, D.: Expression ofbiosynthetic genes from P~,udomonm a£vugfnasa and F.sdm,id~ coil in the heterolngous host. Mol. Gen. Genet. 203 (1986) 421--429. Krishna, R.V. and Lcisinger, T.: Biosynthesisof proline in aemginosa I. Biochem J. 181 (1979) 215-222. Krishna, R.V., Beilstein, P. and Leisinger,1".: Biosynthesis ofproline in Ps~domoaas aeruginosa II. Biochem.J. 181 (1979) 223-230. Martin, C., Cami, B., Yeh, P,, Stragier, P.. Parsot, C. and Patio, J.-C.: Pseudomonas aeruginosa diaminopimelate decarboxylase: Evolutionary relationship with other amino acid decarboxylases. Mol. Biol. Evol. $ (1988) 549-5~9. Maxam, A.M. and Gilbert, W.: Sequencingend-laheled DNA with basespecific chemical cleavages. Methods Enzymol. 65 (1980)499-$60. Mute, A. and Osawa, S.: The guanine and cytosine content of genomie DNA and bacterial evolution. Prec. Natl. Acad. Sci. USA 84 (1987) 166-169. Paileroni, N.: Taxonomy of the pseudomonads. In Sokatch, J.R. and Ornston, LN. (Eds.), The Bacteria, Vol. 10. Academic Press, Orlando, FL, 1986, pp. 3-25. Rella, M. and Haas, D.: Resistance of Pseudomonas aeruginosa PAO to nalidixic acid and low levels of p.lactam antibiotics: mapping of chromosomal genes. Antimicrob. Agents Chemother. 22 (1982) 242-249. Sane, Y. and Kageyama, M.: The sequence and function ofthe recA 8ene and its protein in PseudomonasaeruginosaPAO. Mol. Gen. Genet. 208 (1987) 412-419. Soldati, L., Jeanes, DJ. and Haas, D.: Effective gone expression in Pseudomonas aeruginosa under the control of the £scherichla coli consensus promoter. FEMS Microbiol. Lett. 42 (1987) 163-167. Stackebrandt, E.: Phylogenetic relationships vs. phenotypic diversity: how to achieve a phylogeneticclassificationsystem of the eubacteria. Can. J. Microbiol. 34 (1988) $$2-$$6. Toschka, H.Y., HOpfi, P., Ludwig, W., Schleifer, K.H., Ulbrich, N. and Erdmann, V.A.: Complete nucleotide sequence of a 23S ribosomal RNA gene from Pseudomonas aeruflnosa.Nucleic Acids Res. IS (1987) 7182. Toschka, H.Y., H6pfi, P., Ludwig, W., Schleifer, K.H., Ulbrich, N. and Erdmann, V.A.: Complete nucleotide sequence of a 16S ribosomal RNA gone from Pseudomonas aemginosa. Nucleic Acids Reg. 16 (1988) 2348. Van Vliet, F., Boyen, A. and Glansdorff, N.: On interspecies gone transfer: The case of the avgF gone of £schen'chia coll. Ann. Inst. Pasteur/Microbiol. 139 (1988) 493-496. Vogel, D.W., Hartmann, R.K., Struck, J.C.R., Ulbrich, N. and Erdmann, V.A.: The sequence of the 6S RNA gone ofPseudomonm aeru~nma. Nucleic Acids Reg. 15 (1987) 4583-4591. West, S.E.H. and Igiewski, B.H.: Codon usage in Pseudomonas aeruginosa. Nucleic Acids Reg. 16 (1988) 9323-9335. Woese, C.R.: Bacterial evolution. MicrobioL Roy. 51 (1987) 221-271. Woese, C.R., Weisburg, W.G., Hahn, C.M., Paster, BJ., Zahlen, L.B., Lewis, B.J., Macke, TJ., Ludwig, W. and Stackcbrandt, E.: The phylogeny of purple bacteria: the gamma subdivision. Syst. Appl. Microbiol. 6 (1985) 25-33. Yanofsky, C.: Comparison of regulatory and structural regions of genes of tryptophan metabolism. Mol. Biol. Evol. I (1984) 143-161. Yanofsky, C., Plait, T., Crawford, I.P., Nichols, B.P., Christie, G.E., Horowitz, H., Van Clcemput, M. and Wu, A.M.: The complete nucleotide sequence of the tryptophan operon of £sche~chia coll. Nucleic Acids Res. 9 (1981) 6647-6668.