JOURNAL OF BACTERIOLOGY, Nov. 1990, p. 6581-6584 Vol. 172, No. 11 0021-9193/90/116581-04$02.00/0 Copyright X 1990, American Society for Microbiology Cloning of an aroF Allele Encoding a - Insensitive 3-Deoxy-D-arabino-Heptulosonate 7-Phosphate Synthaset LISA M. WEAVER AND KLAUS M. HERRMANN* Department ofBiochemistry, Purdue University, West Lafayette, Indiana 47907 Received 9 May 1990/Accepted 10 August 1990 In Escherichia coli, genes aroF+, aroG+, and aroH+ encode isoenzymes of 3-deoxy-D-arabino-heptulosonate 7-phosphate synthases that are feedback inhibited by tyrosine, , and , respectively. A single base pair change in aroF causes a Pro-148-to-Leu-148 substitution and results in a tyrosine-insensitive .

In bacteria and plants, the aromatic amino acids phenyl- plasmid, designated pLW22, contained an 8-kb insert that alanine, tyrosine, and tryptophan are synthesized via the hybridized to the 714-bp DdeI (Fig. 1) aroF probe (6, 11). shikimate pathway (7, 13). The first enzyme of this pathway Digestion of pLW22 with HindIII-BglII gave a 950-bp frag- is 3-deoxy-D-arabino-heptulosonate 7-phosphate (DAHP) ment that hybridized to the probe. This result was unex- synthase (EC 4.1.2.15). In Escherichia coli, the three un- pected, since the corresponding wild-type fragment is 1.8 kb linked genes aroF+, aroG+, and aroH+ (1) encode three in size. Subsequent detailed restriction analysis and hybrid- isoenzymes of DAHP synthase that are sensitive to tyrosine, ization of pLW22 with the 796-bp DdeI (Fig. 1) aroF probe phenylalanine, and tryptophan, respectively (2). Carbon (6, 11) suggested a new BglII site and revealed that only the flow through the shikimate pathway is controlled by modu- 5' end of the MK201 aroF allele had been cloned (Fig. 1). lation of DAHP synthase; although all three DAHP syn- Apparently, our selection allowed survival of plasmid-bear- thases are transcriptionally regulated (3, 6, 11), feedback ing strains that did not contain an intact aroF gene. inhibition is quantitatively the major control mechanism in To verify that the aroF allele of MK201 contained a new vivo (12). The three genes have been cloned (4, 19, 20), and BgIII site, chromosomal and plasmid DNA of wild-type E. the sequences of the encoded DAHP synthases coli W3110, mutant MK201, and plasmid pLW22 were have been obtained through combinations of amino acid and digested with BglII-EcoRI and subjected to Southern analy- nucleotide sequence analyses (8, 14, 16, 17). The three sis with the 714-bp DdeI probe. The hybridization pattern for isoenzymes are immunologically distinct (10), yet amino acid the BgIII-EcoRI digest of MK201 and pLW22 DNA showed sequence alignment shows regions of near identity inter- a 2.6-kb hybridizing fragment instead of the 3.6-kb fragment spersed with seemingly unrelated sequences (5, 16). Several found in W3110 DNA. This result indicated that the MK201 aroH alleles encoding feedback-insensitive have mutation had generated a new BglII site within the aroF been isolated and the mutations have been identified by coding region. nucleotide sequence analysis (14). We report the cloning and To obtain the 3' end of the aroF mutant allele, a partial nucleotide sequence of an aroF allele encoding a tyrosine- pKGW-derived PstI library was generated. One plasmid of insensitive DAHP synthase. A single base pair change, this library, designated pLW12 (Fig. 1), contained a 3.5-kb resulting in a Pro-to-Leu substitution, is sufficient to lead to PstI fragment that hybridized to both the 714- and 796-bp tyrosine insensitivity. DdeI probes (Fig. 1). The aroF gene from MK201 was Cloning of the MK201 aroF allele. E. coli mutants contain- reconstructed from plasmids pLW22 and pLW12 (Fig. 1). ing tyrosine-insensitive DAHP synthase were obtained by The ClaI-BgIIl fragment of pLW18 containing the wild-type selection for growth in the presence of 3-fluorotyrosine aroF gene was replaced with the CIaI-BglII fragment from (Kimpel and Herrmann, unpublished data). The aroF allele pLW22 containing the 5' end of the mutant aroF allele. The was cloned from one such mutant, designated MK201; in cell resulting plasmid pLW1812 contains a unique BglII site into extracts of strain MK201, DAHP synthase is insensitive to which the BgIII fragment of pLW12 was inserted, giving tyrosine. In wild-type E. coli, the entire aroF gene is plasmid pLW4 (Fig. 1). contained on a single BglII fragment (8). We intended to Nucleotide sequence analysis. Figure 2 shows the strategy clone the corresponding BgIII fragment from size-fraction- for sequencing the entire aroF mutant gene of pLW4, the ated MK201 DNA by employing the positive selection nucleotide sequence of the relevant PstI-DraI fragment that vector, pKGW (9) (Fig. 1). A partial pKGW-derived library covers the new BglII site, and the deduced amino acid was screened in the aroF strain YS482 plated on M9 medium sequence. A single C-to-T change at nucleotide 443 re- supplemented with phenylalanine, tryptophan, kanamycin, sulted in the new BglII site and a Pro-to-Leu change for and isopropyl-,B-D-thiogalactopyranoside. One resulting amino acid residue 148. In addition, we found an 11-bp insert, a direct repeat of nucleotides 881 to 891. The insert introduced tandem TGA codons into the aroF gene reading * Corresponding author. frame, which shortened the polypeptide by 53 amino acid t Paper no. 12,286 of the Purdue University Agricultural Exper- residues. However, immunoblots of protein extracts from E. iment Station. coli MK201 and W3110 showed no difference in the size of 6581 6582 NOTES J. BACTERIOL.

I H;ndi

In

Hind M Dde I 714 APRroF DjeI pCG2OI Bss HE 796 tyrA' I on ~~Ddel Barn HI

FIG. 1. Strategy to clone strain MK201 aroF (pLW4) and the aroF Leu-148 allele (pLW42); _, MK201 aroF sequences; 1 , wild-type aroF sequences. the DAHP synthase polypeptide that cross-reacted with proximately 6 kDa smaller than the corresponding polypep- antibodies raised against the wild-type aroF . This tides of the other aroF alleles carrying strains. Assays for result suggested that the 11-bp insert was a cloning artifact. DAHP synthase enzyme activity confirm the immunoblots. The 11-bp insert of pLW4 lies on a 1.4-kb BssHII-BamHI Strain HE628(pCG201), bearing the plasmid with the wild- fragment. Plasmid pLW42 was constructed by substituting type aroF allele, produced DAHP synthase that was feed- this BssHII-BamHI fragment of pLW4 for the corresponding back inhibited by L-tyrosine. In strain HE628(pLW42), wild-type fragment. Plasmid pLW42 does complement the bearing the plasmid that encodes the Leu-148 aroF product, aroF aroG double deletion strain HE628 (6) and therefore the' enzyme was insensitive to tyrosine. Extracts of contains a gene encoding a functional DAHP synthase. HE628(pLW4) showed no measurable DAHP synthase. The DAHP synthase encoded by plasmid pLW42. Figure 3 Leu-148 aroF product was purified and tested for tyrosine shows immunoblots of cell extracts from strain MK201 and sensitivity (Fig. 4). The enzyme was not inhibited by ty- from strain HE628 bearing either pCG201, pLW4, or rosine, even at concentrations that reduced the wild-type pLW42. An aroF gene product of the appropriate size was enzyme activity by more than 90%. In fact, this enzyme may detected in extracts of MK201 cells and of HE628 cells be slightly activated by the amino acid, a phenomenon that carrying plasmid pCG201 or pLW42. Cells transformed with we observed previously for the wild-type E. coli enzyme at pLW4 produced an immunoreactive protein that was ap- a very low concentration (15) and also for the VOL. 172, 1990 NOTES 6583

A -x "_ _. ml - CE I I0 cr l- I .1 I1. ATG .__ +AA

l00bp

360 375 390 B CTG CAG ATC GCG CGT AAA TTG CTG CTT GAG CTG GTG AAT ATG GGA CTG LEU GLN ILE ALA ARG LYS LEU LEU LEU GLU LEU VAL ASN MET GLY LEU139 BalII 405 420 T' 435 450 CCA CTG GCG ACG GAA GCG TTA GAT CCG AAT AGC CCG CAA TAC CTG GGC PRO LEU ALA THR GLU ALA LEU ASP PRO ASN SER PRO GLN TYR LEU GLY155 LEU 465 480 495 GAT CTG TTT AGC TGG TCA GCA ATT GGT GCT CGT ACA ACG GAA TCG CAA ASP LEU PHE SER TRP SER ALA ILE GLY ALA ARG THR THR GLU SER GLN171 510 525 540 ACT CAC CGT GAA ATG GCC TCC GGG CTT TCC ATG CCG GTT GGT TTT AAA THR HIS ARG GLU MET ALA SER GLY LEU SER MET PRO VAL GLY PHE LYS187 FIG. 2. Leu-148 aroF allele ofE. coli. (A) Strategy for sequencing the entire gene; the arrows indicate the direction and length of individual sequencing runs. (B) Nucleotide sequence of the PstI-DraI fragment and the deduced amino acid sequence around residue Leu-148; the overline indicates the new BglII site at nucleotides 423 through 428.

enzyme from carrots under physiologically more relevant other organisms (5, 11) revealed common structural domains conditions (18). Thus, a single amino acid residue change at that are separated by regions of largely unique sequences. position 148 is sufficient to generate an aroF product deriv- Residue 148 is in a region of low sequence identity between ative that is fully insensitive to tyrosine feedback inhibition. two structural domains. This region was previously identi- The nature of the amino acid residue change is notewor- fied as a major part of the allosteric site for the tryptophan- thy. Although residue 148 is Pro in the tyrosine-sensitive sensitive isoenzyme (14). Amino acid changes in the two DAHP synthase, the corresponding residue in the phenylal- residues flanking position 148, namely, Val-147 and Gly-149, anine- and tryptophan-sensitive DAHP synthases is Met (4, gave aroH products that were no longer feedback inhibited 14). Met and Leu residues have similar hydrophobic char- by tryptophan. Thus, residues 147 through 149 appear to acters and similar effects on protein secondary structure. represent part of an aromatic amino acid-binding pocket Thus, the mutation that results in a tyrosine-insensitive common to at least two of the three E. coli DAHP synthase DAHP synthase increases the similarity in the primary isoenzymes, and variations in the sequence of this region structure between the three enzymes. contribute critical structure to the different allosteric binding The position of the mutation within the gene deserves sites of this regulatory enzyme. comment. Sequence alignment of the three E. coli DAHP synthase isoenzymes (14, 16) and with the enzyme from 120

100

C:c N 0 3: :,?. (u IJ j UJ (I.) 2 CL CL CL I 80

110- 84 -60

47-

33--

24 -- 20

6--

FIG. 3. Immunoblot of cell extracts of strains MK201 and HE628 Q05 0.10 0.15 020 025 030 bearing the plasmids indicated above the blot; comparable amounts [TYROSINI] (mM) of protein were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted onto nitrocellulose, and the blot was FIG. 4. Tyrosine sensitivity of wild-type (C)) and Leu-148 (l) developed with anti-tyrosine-sensitive DAHP synthase. DAHP synthase. 6584 NOTES J. BACTERIOL.

We thank R. L. Somerville and H. Zalkin for bacterial strains and thase isoenzymes. J. Biol. Chem. 254:3761-3764. plasmids, M. Kimble for help in the selection of aroF mutants 11. Muday, G. K., and K. M. Herrmann. 1990. Regulation in containing the tyrosine-insensitive DAHP synthase, and M. D. Escherichia coli of the aroF gene of Salmonella typhimurium. J. Poling for technical assistance. Bacteriol. 172:2259-2266. This work was supported by a U.S. Department of Agriculture 12. Ogino, T., C. Garner, J. L. Markley, and K. M. Herrmann. Biotechnology National Needs Fellowship to L.M.W. and by Public 1982. of aromatic compounds: 13C NMR spectros- Health Service grant GM-17678 to K.M.H. from the National copy of whole Escherichia coli cells. Proc. Natl. Acad. Sci. Institutes of Health. USA 79:5828-5832. 13. Pittard, A. J. 1987. Biosynthesis of the aromatic amino acids, p. LITERATURE CITED 368-394. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. 1. Bachmann, B. J. 1987. Linkage map of Escherichia coli K-12, Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Esche- edition 7, p. 806-807. In F. C. Neidhardt, J. L. Ingraham, K. B. richia coli and Salmonella typhimurium: cellular and molecular Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), biology. American Society for Microbiology, Washington, D.C. Escherichia coli and Salmonella typhimurium: cellular and 14. Ray, J. M., C. Yanofsky, and R. Bauerle. 1988. Mutational molecular biology. American Society for Microbiology, Wash- analysis of the catalytic and feedback sites of the tryptophan- ington, D.C. sensitive 3-deoxy-D-arabinoheptulosonate 7-phosphate syn- 2. Brown, K. D. 1968. Regulation of aromatic amino acid biosyn- thase of Escherichia coli. J. Bacteriol. 170:5500-5506. thesis in Escherichia coli K-12. Genetics 60:31-48. 15. Schoner, R., and K. M. Herrmann. 1976. 3-Deoxy-D-arabino 3. Brown, K. D., and R. L. Somerville. 1971. Repression of heptulosonate 7-phosphate synthase. Purification, properties, aromatic amino acid biosynthesis in Escherichia coli K-12. J. and kinetics of the tyrosine-sensitive isoenzyme from Esche- Bacteriol. 108:386-399. richia coli. J. Biol. Chem. 251:5440-5447. 4. Davies, W. D., and B. E. Davidson. 1982. The nucleotide sequence of aroG, the gene for 3-deoxy-D-arabino heptu- 16. Shultz, J., M. A. Hermodson, C. C. Garner, and K. M. Herr- losonate 7-phosphate synthetase (phe) in Escherichia coli K-12. mann. 1984. The nucleotide sequence of the aroF gene of Nucleic Acids Res. 10:4045-4058. Escherichia coli and the amino acid sequence of the encoded 5. Dyer, W. E., L. M. Weaver, J. Zhao, D. N. Kuhn, S. C. Weller, protein, the tyrosine-sensitive 3-deoxy-D-arabino heptulosonate and K. M. Herrmann. 1990. A cDNA encoding 3-deoxy-D- 7-phosphate synthase. J. Biol. Chem. 259:9655-9661. arabino-heptulosonate 7-phosphate synthase from Solanum tu- 17. Shultz, J., M. A. Hermodson, and K. M. Herrmann. 1981. A berosum L. J. Biol. Chem. 265:1608-1614. comparison of the amino terminal sequences of 3-deoxy-D- 6. Garner, C. C., and K. M. Herrmann. 1985. Operator mutations arabino heptulosonate 7-phosphate synthase isoenzymes from of the Escherichia coli aroF gene. J. Biol. Chem. 260:3820- Escherichia coli. FEBS Lett. 131:108-110. 3825. 18. Suzich, J. A., J. F. D. Dean, and K. M. Herrmann. 1985. 7. Herrmann, K. M. 1983. The common aromatic biosynthetic 3-Deoxy-D-arabino heptulosonate 7-phosphate synthase from pathway, p. 301-322. In K. M. Herrmann and R. L. Somerville carrot (Daucus carota) is a hysteretic enzyme. Plant Physiol. (ed.), Amino acids: biosynthesis and genetic regulation. Addi- 79:765-770. son-Wesley Publishing Co., Inc., Reading, Mass. 19. Zurawski, G., K. Brown, D. Killingly, and C. Yanofsky. 1978. 8. Hudson, G. S., and B. E. Davidson. 1984. Nucleotide sequence Nucleotide sequence of the leader region of the phenylalanine and transcription of the phenylalanine and tyrosine operons of operon of Escherichia coli. Proc. Natl. Acad. Sci. USA 75: Escherichia coli K12. J. Mol. Biol. 180:1023-1051. 4271-4275. 9. Kuhn, I., F. H. Stephenson, H. W. Boyer, and P. J. Greene. 20. Zurawski, G., R. P. Gunsalus, K. D. Brown, and C. Yanofsky. 1986. Positive-selection vectors utilizing lethality of the EcoRI 1981. Structure and regulation of aroH, the structural gene endonuclease. Gene 42:253-263. for the tryptophan-repressible 3-deoxy-D-arabino-heptulosonic 10. McCandliss, R. J., and K. M. Herrmann. 1979. Immunological acid 7-phosphate synthetase of Escherichia coli. J. Mol. Biol. studies of 3-deoxy-D-arabino-heptulosonate 7-phosphate syn- 145:47-73.