Vol. 174, No. 8JOURNAL OF BACTERIOLOGY, Apr. 1992, p. 2694-27010021-9193/92/082694-08$02.00/0Copyright X) 1992, American Society for Microbiology

Sequence Analysis and Complementation Studies of the argJ GeneEncoding Ornithine Acetyltransferase from Neisseria gonorrhoeae

PAUL R. MARTIN AND MARTHA H. MULKS*

Department ofMicrobiology and Public Health, Michigan State University,East Lansing, Michigan 48824-1101

Received 3 September 1991/Accepted 10 February 1992

Clinical isolates of Neisseria gonorrhoeae frequently are deficient in arginine biosynthesis. These auxotrophsoften have defects in the fifth step of the arginine biosynthetic pathway, the conversion of acetylornithine toornithine. This reaction is catalyzed by the enzyme ornithine acetyltransferase, which is a product of the argigene. We have cloned and sequenced the gonococcal argi gene and found that it contains an open reading frameof 1,218 nucleotides and encodes a peptide with a deduced M, of 42,879. This predicted size was supported byminicell analysis. This gene was capable of complementing both Escherichia coli argE and argA mutations andof transforming an ArgJ- strain of N. gonorrhoeae to Arg+. Southern blots were able to detect bands thatspecifically hybridized to the gonococcal argJ gene in genomic DNA from Pseudomonas aeruginosa but not E.coli, a result that reflects the divergent nature of the arginine biosynthetic pathway in these organisms.

Arginine biosynthesis in prokaryotes occurs by an eight-step pathway (6) (Fig. 1). The first four steps of the pathwayinvolve N-acetylated intermediates, beginning with the acet-ylation of glutamate and ending with acetylornithine.N-Acetylomithine is converted to ornithine in the fifth stepof the pathway, and two separate enzymes have evolved tocatalyze this reaction. In members of the family Enterobac-teriaceae, N-acetylomithine is hydrolyzed to ornithine andacetate by the ArgE enzyme N-acetylornithinase (EC3.5.1.16) (35). In other bacteria, including the methanogens,cyanobacteria, pseudomonads, and Neisseria gonorrhoeae,the acetyl group of N-acetylornithine is transferred to gluta-mate by the ArgJ enzyme ornithine acetyltransferase(OATase, EC 2.3.1.35) (13, 19, 28, 34). The N-acetylgluta-mate produced from this reaction can be cycled back into thearginine biosynthetic pathway, bypassing step 1. N-Acetyl-ornithinase has been purified from Escherichia coli andappears to function as a monomer with an Mr of 62,000 (6).In contrast, no OATase from any source has been charac-terized, and little is known of the physical properties of thisenzyme. While the biosynthesis of arginine by E. coli hasbeen extensively studied as a model of gene regulation, fewof the genes encoding the enzymes involved in this pathwayhave been sequenced. Nucleotide sequence data are avail-able for E. coli argA, argF, argI, and carAB, for theregulatory region of the argECBH operon, and for Pseudo-monas aeruginosa argF, arcB, and argA. Neither E. coliargE nor P. aeruginosa argJ has been sequenced.Although N. gonorrhoeae possesses all the enzymes re-

quired for arginine biosynthesis, arginine auxotrophs arecommonly encountered among clinical isolates of this patho-gen (3). In a survey of 212 arginine auxotrophs, 69% werefound to have blocks at the fifth step of the pathway (4).Gonococcal strains with combined auxotrophy for arginine,hypoxanthine, and uracil (AHU strains) are of particularinterest because of their prevalence among isolates fromasymptomatic and disseminated gonococcal infections (4,14). AHU strains have repeatedly been shown to be a highlyhomogeneous group and are believed to be derived clonally

* Corresponding author.

(15, 18, 21). All AHU strains tested were found to be argJmutants (28), and AHU strains were shown to have identicalor overlapping argJ mutations in that DNA from one AHUstrain could not restore any other strain to Arg+ in pairedtransformation experiments (4, 18).At least four of the genes of the arginine biosynthetic

pathway (argB, argF, argG, and argJ), plus the genes forcarbamoylphosphate synthetase (carA and carB), have beencloned from N. gonorrhoeae (17, 23). Picard and Dillon (23)found that a single lambda clone containing the gonococcalargJ gene could complement both E. coli argE and argAmutations. They proposed that the gonococcal OATase maybe able to complement the argA mutation through thefeedback pathway, although it is also possible that thegonococcal argA gene was present on the same clone. Onlythe carA and carB genes were shown to be closely linked inthat study (23).

In this paper, we report the cloning and sequencing of thegonococcal argJ gene and investigate its potential to com-plement E. coli argE and argA mutations. This representsthe first reported sequence of an OATase from any source.Along with providing new information on the properties ofthe OATase enzyme, this study will provide a foundation forfurther investigations into the molecular basis of arginineauxotrophy and the regulation of gene expression in N.gonorrhoeae.

MATERIALS AND METHODS

Bacterial strains. E. coli GM2163 (mcrA mcrB ara-14leuB6 tonA13 lacYl tsx-78 supE44 galK2 dcm-6 hisG4rpsL136 dam-13::Tn9 xyl-5 mtl-l thi-1 hsdR2) was used forthe initial construction of the genomic library. The argE E.coli strain AT2538 [thr-1 ara-14 leuB6 A(gpt-proA)62 lacYlsupE44 galK2 A- hisG4 rpsL31 xyl-5 mtl-1 pyrE60 argE3thi-1] and the argA E. coli strain W3421 [argA21 galT23IN(rrnD-rrnE)1 X-] were used for cloning and complemen-tation studies. The minicell-producing E. coli strain DS410(minAl minB2 rpsLI35xyl-7 mtl-2 thi-1) was used for mini-cell isolation and protein labelling experiments. E. coli DH1(Arg+), E. coli N134 [A(gpt-lac)5 reLA1 spoT] thi-1 argI68],and P. aeruginosa ATCC 27853 were used for Southern

2694

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

N. GONORRHOEAE argJ GENE 2695

GLUTAMATEACETYL CoA (argA)

ACETYLGLUTAMATE

ACETYLGLUTAMYL PHOSPHATE

ACETYLGLUTAMATE SEMIALDEHYDE

ACETYLORNITHNE

GLUTAMATEIOrnThAMcE --,, AcetylornithinaseOrnithine Acetyl (argE)transferase (argJ)

ACETATE

ORNITHINE

CARBAMOYLPHOSPHATE

CITRULLINE

ARGINOSUCCINATE

ARGININEFIG. 1. Arginine biosynthesis, showing the alternative pathways

for the conversion of N-acetylornithine to ornithine.

blots. N. gonorrhoeae CDC50 (Pro-) was the source of theargJ gene. N. gonorrhoeae NRL30465 (ArgJ- Hyx- Ura-)was used as a transformation recipient.Media and growth conditions. All E. coli strains were

grown on Luria-Bertani (LB) medium; when needed, ampi-cillin was added to 50 ,ug/ml for plasmid selection. All N.gonorrhoeae strains were routinely grown on GC basemedium (Difco) plus 1% Kellogg's supplement (39) at 36°Cunder 5% CO2. NEDA defined medium for gonococci (4),either as a broth or with 1.5% agarose (Baker), was used forgrowth of all bacteria in auxotyping or complementationexperiments.DNA manipulations. Restriction enzymes and DNA ligase

were purchased from Boehringer Mannheim Biochemicalsor Bethesda Research Laboratories and used according tothe manufacturer's instructions. Genomic and plasmid DNApreparations, gel electrophoresis, and E. coli transformationwere all performed by conventional methods (26).

Cloning and sequencing of the gonococcal argJ gene. Agenomic library of N. gonorrhoeae CDC50 was constructedby partially digesting genomic DNA with Sau3A and ligatingfragments of 1.5 to 8 kb into the BamHI site of the plasmidvector pUC12. A recombinant plasmid containing the gono-coccal argJ gene was isolated from this library by comple-mentation of the argE E. coli strain AT2538. This isolate wassubcloned into M13mpl9, and nested deletions were con-

structed by using the Cyclone deletion kit from InternationalBiotechnologies, Inc. These deletions were sequenced withthe dideoxy chain termination reagents and Sequenase 2.0enzyme from U.S. Biochemicals by the method of Sanger etal. (27). The sequence was analyzed by using GENEPROsoftware (Riverside Scientific Enterprises), and these data

were used to determine the probable argJ open readingframe (ORF). Second-strand sequencing was done fromidentified restriction sites in M13 or pUC constructs.Computer analysis. GenBank (release 68, dated June 1991)

and EMBL (release 27, dated May 1991) data bases weresearched for sequences homologous to argJ by using theFASTA and TFASTA programs of the Genetics ComputerGroup (GCG) Sequence Analysis Software Package, version7.0 (7), using a VAX 8650 computer. GCG GAP and BEST-FIT programs and the GENEPRO Align program were usedto compare nucleic acid and amino acid sequence homology.GCG PEPTIDESTRUCTURE and PLOTSTRUCTUREprograms and PARCOR Protein Structure Comparison Soft-ware (Mulks Micro, Williamston, Mich.) were used to ana-lyze similarities in predicted protein secondary structure.Codon usage tables were compared by using the GCGCORRESPOND program.OATase assays. OATase activity was assayed by measur-

ing the production of ornithine from acetylornithine andglutamate, using the ninhydrin reaction as described byShinners and Catlin (28).

Gonococcal transformation. Transformation of N. gonor-rhoeae NRL30465 was done as previously described (1),using recombinant plasmid DNA as the donor. Positivetransformation was scored as growth of NRL30465 onNEDA plates lacking arginine after 48 h.

Minicell analysis. The minicell-producing E. coli strainDS410, transformed with pUC19 or pRGE clones, wasgrown to high density in 1-liter LB broth cultures, andminicells were harvested by differential centrifugation fol-lowed by sucrose density gradient centrifugation as de-scribed by Dougan and Kehoe (8). The minicells wereresuspended in 1 ml of NEDA broth lacking methionine; 50,uCi of [35S]methionine (Amersham) was added, and theminicells were labelled at 37°C for 1 h. Labelled proteinswere separated by sodium dodecyl sulfate-polyacrylamidegel electrophoresis (SDS-PAGE) and visualized by autora-diography.

Southern blots. Genomic DNA from N. gonorrhoeae(CDC50 and NRL30465), E. coli (DH1, N134, and AT2538),and P. aeruginosa (ATCC 27853) was digested to completionwith either ClaI or EcoRI, separated on an 0.8% agarose gel,and transferred to nitrocellulose by the method of Southern(31). Blots were probed with either the entire 3.4-kb insert ofpRGE5 or with the 0.6-kb HincII fragment of pRGEBlabelled with digoxigenin by using the Genius kit fromBoehringer Mannheim Biochemicals. Hybridizations wereperformed at both low and high stringency. High-stringencyconditions included hybridization at 600C in 5 x SSC (lxSSC is 0.15 M NaCl plus 0.015 M sodium citrate) followed bywashes in 0.lx SSC plus 0.1% SDS at 60°C. Low-stringencyconditions included hybridization at 50°C in 5 x SSC fol-lowed by washes in 2x SSC plus 0.1% SDS at 50°C.

Nucleotide sequence accession number. The nucleotide andamino acid sequences for the N. gonorrhoeae argJ genehave been submitted to GenBank and assigned the accessionnumber M65216.

RESULTS

Cloning of the N. gonorrhoeae argJ gene. A genomic libraryof N. gonorrhoeae CDC50 DNA was constructed in pUC12and used to transform the argE E. coli strain AT2538 toarginine prototrophy. A recombinant plasmid, designatedpRGE5, was isolated from an Arg+ transformant. Restric-tion enzyme analysis revealed that pRGE5 contained a

VOL. 174, 1992

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

2696 MARTIN AND MULKS

_ (1)

H CpRGE5

pRGEA

pRGEC

pRGEH

pRGE 172 ! I

pRGES I

pRGEB

(2) argJ

(4)

B HA C AH S AI1 I I

COMPLEMENT TRANSFOR4 OATaseArgE ArgA AHJ to Arg+ ACTIVITY

+ + + 683

- - - <0.1

- - + * <0.1

I~ - - - <0.1

+ + + 47.7- - + <0.1

+ + + 54.7

I I I I I0.0 0.5 1.0 1.5 2.0 2.5 3.0 kb

FIG. 2. Restriction map and analysis of deletions of the gonococcal argJ gene. Restriction sites shown are as follows: A, AvaI; B, BsmI;C, ClaI; H, HincII; S, SacII. Arrows represent ORFs. Complementation ofargE was tested in E. coli AT2538. Complementation of argA wastested in E. coli W3421. Gonococcal transformation to Arg+ was tested in strain NRL30465. OATase activity is given in nanomoles ofornithine produced per minute per milligram of protein in the enzyme sample. The asterisk indicates an approximately 10-fold lowertransformation frequency relative to those of the other positive samples.

3.4-kb insert of gonococcal DNA (Fig. 2). OATase assays ofE. coli AT2538 cells containing pRGE5 confirmed that thisplasmid encoded active OATase enzyme. Deletion mutantswere constructed by usingAval, ClaI, and HincII restrictionsites identified in pRGE5; these were tested for argE com-plementation and for production of OATase (Fig. 2). Allthree deletion mutants were negative for OATase activityand were unable to complement E. coli argE, indicating thatat least part of the gonococcal argJ gene is encoded by the0.19-kb AvaI-ClaI fragment deleted in all three constructs(Fig. 2).

Sequencing and identification of the argj gene. The entire3.4-kb insert in pRGE5 was subcloned into M13mpl9 forDNA sequencing. Nested deletions of the 3.4-kb insert weregenerated by using the 3'-5' exonuclease activity of T4polymerase provided in the Cyclone deletion kit from Inter-national Biotechnologies, Inc. These deletions were se-quenced and provided a 3.01-kb stretch of continuous nucle-otide sequence which when analyzed revealed four largeORFs (Fig. 2). Comparison of the codon usage and thirdcodon bias in these four ORFs to codon frequency tables forhighly expressed E. coli genes (10), weakly expressed E. coligenes (11), N. gonorrhoeae chromosomal genes (37), and P.aeruginosa chromosomal genes (38) suggests that ORFs 1and 2 are real and that ORFs 3 and 4 probably do not encodeactual proteins. This is supported by the fact that ORFs 1and 2 have good potential ribosome-binding sites (25), whileORFs 3 and 4 do not.

It was clear that ORF 1 could not encode argJ, sincepRGEA could neither complement argE nor specify produc-tion of active OATase. To determine which of the threeremaining ORFs encoded argJ, three more deletion mutantswere constructed. In pRGE172 (Fig. 2), the nested deletionclosest to the 3' end of ORF 2 was subcloned into pUC19.This plasmid was trimmed to the BsmI site to constructpRGEB and to the SacIl site to construct pRGES. pRGEBcontained two complete overlapping ORFs, ORF 2 and ORF3, as well as a truncated ORF 4, and it was able tocomplement argE and produced active OATase. The abilityto complement E. coli argE, which requires an intact argJgene, was lost in pRGES, a deletion that truncates ORF 2

but does not enter ORF 3. Enzyme assays showed nodetectable OATase production by pRGES. These data indi-cate that argJ is encoded by ORF 2.

Minicell analysis. Plasmid pRGE5 and its deletions weretransformed into the minicell-producing E. coli strain DS410,and the protein products of these plasmids were labelledwith [35S]methionine, separated by SDS-PAGE, and visual-ized by autoradiography (Fig. 3). Compared with the pUC19vector, the plasmids containing the complete ORF 2 codedfor the production of an additional peptide with an apparentMr of 46,000. When pRGES was compared with pRGE5, this

66.2-*_

FIG. 3. SDS-PAGE ofIS-labelled peptides produced in E. coliDS410 minicells. Minicells'contained the following: lane 1, pUC19;lane 2, pRGE5; lane 3, pRGEA; lane 4, pRGEB; lane 5, pRGES;lane 6, pRGEH. Molecular mass standards (in kilodaltons) areindicated on the left. Arrowheads on the right indicate peptidesproduced by recombinant plasmids: 46-kDa ArgJ peptide, 38-kDatruncated ArgJ peptide, and 23-kDa peptide.

J. BACTrERIOL.

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

N. GONORRHOEAE argJ GENE 2697

GCATACGCGGGGATGGGGCTGACGTTCAACGGCGTGCTGACCGCGCTGATTGCGCCGCTGCTTATCCCCGTTTTG 75

GGGTTCTGAACCCGTTTCAGACGGCATTTCAACCCATGCCGTCTGAACGCCCGACACACTCGCAAGGAGAACCGTTRBS

ATGGCTGTCAACCTGACCGAAAAAACCGCCGAACAACTGCCCGACATCGACGGCATTGCCCTCTACACCGCCCAAM A V N L T E K T A E Q L P D I D G I A L Y T A Q

A G V K K P G H T D L T L I A V A A G S T V G A V

TTCACGACCAACCGTTTCTGTGCCGCGCCCGTCCACATCGCCAAATCGCACCTTTTCGACGAAGACGGCGTGCGCF T T N R F C A A P V H I A K S H L F D E D G V R

150

225

300

375

ATCGATGCTTTGGCAGTGTGT 450A L V I N T G N A N A G T G A Q G R I D A L A V C

GCCGCCGCCGCCCGGCAAATCGGCTGCAAACCGAACCAGGTGATGCCCTTCTCCACCGCCGTGATTCTCGAACCGA A A A R Q I G C K P N Q V M P F S T G V I L E P

CTCCCCCCAGACAAAATCATCGCCGCCCTGCCCAAAATGCAGCCTGCCTTCTGGAACGAAGCGGCACGCGCCATCL P A D K I I A A L P K M Q P A F W N E A A R A I

ATGACCACCGACACCGTTCCCAAAGCCGCCTCGCGCGAAGGCAAGGTCGGCGACCAACACACCGTCCGCGCCACAM T T D T V P K A A S R E G K V G D Q H T V R A T

GGCAT1TGUAAAGUGTGGUGCATGATTCATCCCAATATGGCGACCATGCTCGGTTTCATCGCCACCGATGCCAAAG I A K G S G M I H P N M A T M L G F I A T D A K

525

600

675

750

GTTTCCCAACCCGTCCTCCAACTGATGACGCAGGAAATCGCCGACGAAACCTTCAACACCATCACCGTTGACGGC 825V S Q P V L Q L M T Q E I A D E T F N T I T V D G

GACACUAGUACUAACGACAGCTTCGTCATCATCGCCACCGGCAAAAACAGCCAAAGCGAAATCGACAACATCGCCD T S T N D S F V I I A T G K N S Q S E I D N I A

900

GACCCGCGTTACGCCCAACTCAAAGAATTGTTGTGCAGCCTTGCGCTCGAACTCGCCCAAGCCATCGTCCGCGAC 975D P R Y A Q L K E L L C S L A L E L A Q A I V R D

GGCGAGGGCGCGACCAAGTTCATCACCGTCCGCGTCGAAAACGCCAAAACCTGCGACGAAGCCCGCCAAGCCGCC 1050G E G A T K F I T V R V E N A K T C D E A R Q A A

TACGCCGCGGCACGTTCGCCACTGGTCAAAACCGCCTTTTTCGCCTCCGACCCCAACCTCGGCAAGCGTCTCGCC 1125Y A A A R S P L V K T A F F A S D P N L G K R L A

GCCATCGGTTATGCCGACGTTGCCGACCTCGATACCGACCTCGTGGAAATGTATCTCGACGATATTTTGGTTGCC 1200A I G Y A D V A D L D T D L V E M Y L D D I L V A

GAACACGGCGGACGCGCCGCAAGCTACACCGAAGCACAAGGGCAGGCGGTGATGTCGAAGGACGAAATCACCGTCE H G G R A A S Y T E A Q G Q A V M S K D E I T V

CUGATUCAGTGCATCGCGGACAAGCCGCCGCCACCGTCTATACCTGCGACCTGTCGCACGGATACGTTTCCATCR I K L H R G Q A A A T V Y T C D L S H G Y V S I

1275

1350

AACGCCGACTACCGTTCCTGACCCGACACGGCTTCAGACGGCATACATAAATGCCGTCTGAACCGTCGGACAAC 1425N A D Y R S ***

ATACCATGACCTCCACATTCCCCCGCCGCCTCGCCCGCAAAATCCGCCAAACCCGCCGCCTGTCGCGCAAAAGCA 1500

FIG. 4. Nucleotide sequence of the gonococcal argJ gene and flanking regions. The predicted amino acid sequence is shown below thecoding region. Inverted repeats are underlined, as is the putative ribosome-binding site (RBS). Neisserial DNA uptake sequences areunderlined in boldface.

46-kDa peptide was lost, and a new unique peptide with anapparent Mr of 38,000 appeared; this is consistent with theconclusion that the 46-kDa peptide is encoded by the argJORF, ORF 2, and is truncated in pRGES. A smaller peptideof Mr 23,000 was also produced from plasmids pRGE5,pRGEB, and pRGES. Although this peptide was producedonly by plasmids containing the argJ ORF, it is unlikely torepresent the intact ArgJ peptide, since it appeared unalteredin pRGES, a plasmid in which ArgJ activity was lost. Wewere unable to identify peptides that correlated with eitherORF 1 or ORF 4.

Nucleotide sequence analysis. Second-strand sequencing

was performed to confirm the sequence of the argJ gene andits flanking regions. The nucleotide sequence and corre-sponding amino acid sequence of argJ are shown in Fig. 4.The gonococcal argJ gene consists of an ORF of 1,218 bpencoding a peptide with a predicted molecular weight of42,879. This is in excellent agreement with the Mr of 46,000for the ArgJ enzyme determined by minicell analysis. TheargJ ORF is preceded by a potential ribosome-binding sitecontaining a 6-bp sequence, AAGGAG, complementary tothe 3' end of the gonococcal 16S rRNA (25). No consensuspromoter sequence of either the -35/-10 (12) or -24/-12(33) type was found upstream of argJ or ORF 1. It is possible

VOL. 174, 1992

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

2698 MARTIN AND MULKS

TABLE 1. Codon usage of the N. gonorrhoeae argJ gene

Aio No. of Frac- Amino No f Frac- Aio No. of Fa Aio No. of FrcCodon occurrences tia Codon occurrences tion Codon acid occurrences Codon acid occurrences tionacd in argJ acid argi in argi in argJ

TTlT Phe 1 0.07 TCT Ser 0 0.00 TAT Tyr 3 0.32 TGT Cys 2 0.33TTC Phe 10 0.93 TCC Ser 5 0.27 TAC Tyr 3 0.68 TGC Cys 4 0.67TTA Leu 0 0.00 TCA Ser 0 0.00 TAA ***b 0 TGA * 1TTG Leu 4 0.14 TCG Ser 6 0.34 TAG *** 0 TGG Trp 1 1.00

C-TT Leu 2 0.07 CCT Pro 1 0.06 CAT His 2 0.37 CGT Arg 4 0.24CTC Leu 13 0.43 CCC Pro 10 0.69 CAC His 5 0.63 CGC Arg 10 0.61CTA Leu 0 0.00 CCA Pro 1 0.06 CAA Gln 12 0.71 CGA Arg 0 0.00CTG Leu 11 0.36 CCG Pro 3 0.19 CAG Gln 5 0.29 CGG Arg 1 0.05

ATT Ile 6 0.22 ACT Thr 0 0.00 AAT Asn 1 0.05 AGT Ser 0 0.00ATC Ile 21 0.78 ACC Thr 28 0.82 AAC Asn 14 0.95 AGC Ser 7 0.39ATA Ile 0 0.00 ACA Thr 2 0.06 AAA Lys 12 0.67 AGA Arg 1 0.05AT.G Met 10 1.00 ACG Thr 4 0.12 AAG Lys 6 0.33 AGG Arg 1 0.05

GTT Val 6 0.21 GCT Ala 2 0.03 GAT Asp 4 0.15 GGT Gly 4 0.14GTC Val 15 0.52 GCC Ala 47 0.71 GAC Asp 24 0.85 GGC Gly 20 0.69GTA Val 1 0.03 GCA Ala 9 0.13 GAA Glu 17 0.95 GGA Gly 3 0.10GTG Val 7 0.24 GCG Ala 8 0.12 GAG Glu 1 0.05 GGG Gly 2 0.07

a The ratio of the number of occurrences of a specific codon to the number of occurrences of all codons in the same synonymous codon group. A rare codonis defined as one with a fraction of less than 0.10.

b ***, nonsense codon.

that this gene is cotranscribed from sequences upstream ofORF 1, although the fact that pRGEB, which has theupstream portion of ORF 1 deleted, can complement E. coliargE argues against this assumption. In addition, argJ isexpressed in either orientation in pUC12, strongly suggest-ing that it is being expressed from its own promoter ratherthan the lac promoter. We also found no evidence of an "argbox" operator sequence, the binding site for the argininerepressor that has been described for E. coli (5) and Bacillussubtilis (29), upstream of the argJ ORF.Codon usage for the gonococcal argJ gene is listed in

Table 1. We compared our codon usage pattern for this genewith reference codon usage data for N. gonorrhoeae (37), P.aeruginosa (38), and both highly and weakly expressed E.coli genes (10, 11). While codon usage for argJ most closelyresembled that seen with N. gonorrhoeae, there was also astrong resemblance to that of P. aeruginosa. There was adistinct bias in the choice of nucleotide in the wobble (third)position in degenerate codons, with 58.9% containing cy-tosine and 76.1% containing either cytosine or guanine inthis position; the overall guanine-plus-cytosine ratio for argJis 59.2%. This is similar to the data for P. aeruginosa, wherethe 52.6% cytosine and 85.6% cytosine or guanine in thethird position reflects an overall guanine-plus-cytosine ratioof 67%. In contrast, the N. gonorrhoeae reference tableindicates that 34.3% of codons contain cytosine and 52.1%contain cytosine or guanine in the third position, whichwould be expected of an organism with a 49.5% guanine-plus-cytosine ratio.The 10-bp neisserial DNA uptake sequence, 5'-GCCGTC

TGAA-3' (9), occurs four times in this sequence, but only aspart of a larger 12-bp sequence, 5'-ATGCCGTCTGAA-3', inclosely spaced inverted repeats in the flanking regions of thegene and not in the coding region. The downstream invertedrepeats could form a stable stem-and-loop structure with aAG of -23.4 kcal/mol (ca. -97.9 kJ/mol) that may act as atranscriptional terminator. The upstream inverted repeatscould also form a stable stem-and-loop structure with a AG

of -22.4 kcal/mol (ca. -93.7 kJ/mol). This structure followsthe unidentified ORF 1 upstream of the argJ ORF andoverlaps the region where a promoter for argJ would beexpected to be located.We reviewed 18 different N. gonorrhoeae DNA sequences

available in the GenBank and EMBL data bases. Of these,12 contained at least one copy of the gonococcal uptakesequence, either singly or in inverted repeats. When' wealigned these uptake sequences and their flanking sequences,we found that 16 of 22 copies (12 pairs of inverted repeatsand 10 single copies) of the 10-bp uptake sequence alsocontained the 5'-AT extension. This occurred more fre-quently in the inverted repeats than in single copies, suggest-ing that this extension may contribute to the proposedterminator function of these repeats. We also found this12-bp sequence in thrpe Neisseria meningitidis genes, bothin single copies within the genes and in inverted repeats atthe ends of the ORFs-.

Analysis of the predicted Argj protein. The argJ ORFencodes a predicted protein with' a molecular weight of42,879 and an isoelectric point of 5.18. A hydrophobicityplot, using the algorithm of Kyte and Doolittle (16), indicatedno pronounced hydrophobic regions in the argJ gene prod-uct, suggesting that this protein is soluble rather than mem-brane bound. We could find no recognizable signal peptidesequence at the N terminus (36), suggesting that this proteinis not secreted. We therefore conclude that'ArgJ is a solublecytoplasmic protein, as would be expected for an enzymeinvolved in amino acid biosynthesis.

Southern blots. Genomic DNA from E. coli and P. aerug-inosa was tested for homology to the cloned gonococcal argJgene by Southern blot hybridization (Fig. 5). Under high-stringency conditions, neither E. coli nor P. aeruginosaDNA hybridized with either the 0.6-kb (Fig. SC) or the3.4-kb (data not shown) gonococcal argJ probe. Underlow-stringency conditions, specifically hybridizing bandswere apparent above background levels in N. gonorrhoeae

J. BACTERIOL.

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

N. GONORRHOEAE argJ GENE 2699

A1 2 3 4 5 6

B C2 3

N\11.2-6.1 -

4.0-

2.0-J1.

.I

FIG. 5. Southern blot of gonococcal, E. coli, and P. aeruginosaDNA probed with the gonococcal argJ gene. DNA size markers are

indicated in kilobases. Genomic DNA used was from the followingsources: lane 1, N. gonorrhoeae NRL30465; lane 2, N. gonorrhoeaeCDC50; lane 3, E. coli DH1; lane 4, E. coli AT2538; lane 5, E. coliN134; lane 6, P. aeruginosa ATCC 27853 digested with EcoRI; lane6a, P. aeruginosa ATCC 27853 digested with ClaI. (A) EcoRI-digested DNA probed with the 3.4-kb insert from pRGE5 underlow-stringency conditions (see Materials and Methods). (B and C)ClaI-digested genomic DNA (except lanes 6) probed with the 0.6-kbHincIl fragment from pRGEB under low-stringency conditions (B)and high-stringency conditions (C). Solid circles mark P. aeruginosabands that hybridize to the gonococcal probes under low-stringencyconditions. Asterisks mark N. gonorrhoeae bands that hybridizeunder high-stringency conditions.

and P. aeruginosa lanes, but not in E. coli lanes, with boththe 0.6- and 3.4-kb argJ probes (Fig. 5A and B).Homology to other sequenced genes. Comparison of the

argJ nucleotide and predicted amino acid sequences with theGenBank and EMBL data bases yielded no listed DNA or

protein sequences with significant homology to the gonococ-cal argJ gene.We carefully compared the gonococcal argJ gene and its

product with E. coli argA (2), N. gonorrhoeae argF (17), anda gene from Leptospira biflexa which complements argE inE. coli (40), using a variety of nucleotide and amino acidsequence analysis and protein structure analysis programs,and found no significant sequence or structural similaritybetween argJ and any of these genes. We were also able tocompare our sequence with the proposed sequence of theputative argJ gene from Bacillus subtilis (22). Significantidentity was found between the predicted amino acid se-

quences of these two genes, including a region with 23 of 26identical amino acid residues (88.5% homology). We alsoused this conserved 26-amino-acid sequence, and a larger84-amino-acid conserved sequence containing this region, tosearch the GenBank and EMBL data bases, but we found no

other published sequences with significant homology to thissequence.E. coli complementation and N. gonorrhoeae transforma-

tion. The original pRGE5 clone and subsequent deletionswere tested for their abilities to complement argE and argAmutations in E. coli and to transform the argJ N. gonor-rhoeae strain NRL30465 to Arg+ (Fig. 2). These experi-

ments demonstrated that the argE and argA complementingactivities were inseparable and dependent on the presence ofan intact argJ ORF. In contrast, transformation ofNRL30465 to Arg+ was not dependent on an intact argJORF. Both plasmid pRGES, which lacks the argJ genedownstream of the SacII site, and plasmid pRGEC, whichlacks the 5' end of the argJ gene, were able to transformNRL30465 to Arg+, while plasmids pRGEA and pRGEHcould not. These data imply that the lesion in the argJ geneof NRL30465 lies at least partially within the 0.39-kb ClaI-HinclI fragment in the center of the argJ ORF and entirelywithin the 0.63-kb ClaI-SacII fragment (Fig. 2).

DISCUSSION

In this paper, we report the cloning and complete nucleo-tide sequence of the N. gonorrhoeae argJ gene, whichencodes OATase.Four ORFs were found in this sequence. The ORF encod-

ing argJ was identified by the construction of deletionmutants and analysis of their abilities to produce activeOATase and to complement E. coli and N. gonorrhoeaemutants. There is a good ribosome-binding site in the appro-priate position upstream from the putative start codon.However, we were unable to identify a sequence homolo-gous to either the -35/-10 or -24/-12 E. coli consensuspromoters. This was also the case when we studied thegonococcal argF gene (17). While E. coli-type promotershave been found to be associated with some neisserial genes,such as the opaEl and IgAl protease genes (24, 32), little isknown about promoter structure in Neissenra spp., particu-larly for housekeeping genes such as those involved inarginine biosynthesis.

In both E. coli and B. subtilis, many of the genes involvedin arginine biosynthesis are repressed by arginine (6, 30). InE. coli, this repression occurs when the arginine repressorprotein binds to the arg box operator sequence overlappingthe promoter of arginine-repressible genes (5). One or twocopies of the arg box consensus sequence have been foundupstream of several E. coli arginine biosynthetic genes (5) aswell as the B. subtilis argC promoter (29). We found noevidence of an arg box sequence upstream of the gonococcalargJ gene. Previous studies in Catlin's laboratory (28), aswell as preliminary studies in this laboratory (20), havefound no evidence of repression of the gonococcal argJ orargF gene by arginine.The N. gonorrhoeae argJ ORF is closely flanked by

inverted repeats of the neisserial DNA uptake sequence. Intheir study that elucidated the uptake sequence (9), Good-man and Scocca found that this sequence was often presentwithin inverted repeats at the end of undefined ORFs andproposed that the uptake sequence may also be involved intranscription termination. The presence of the uptake se-quence in inverted repeats at the end of the argJ ORFsupports this hypothesis. This inverted repeat was alsofound upstream of the argJ ORF. Whether this sequence isinvolved in termination of a gene upstream of argJ or inregulation of argJ expression has not yet been determined.An interesting observation of the uptake sequences in theargJ gene is that they occur only within the longer 12-bpsequence 5'-ATGCCGTCTGAA-3'. This 12-bp extendedsequence is also present in 16 of the 22 occurrences of theuptake sequence that we analyzed in 12 different gonococcalgenes, as well as in 3 meningococcal genes, and is morefrequently observed in copies of the uptake sequence foundas inverted repeats at the ends of ORFs than in single copies

VOL. 174, 1992

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

2700 MARTIN AND MULKS

of the sequence. These data suggest that the extendeduptake sequence may contribute to a more stable stem-and-loop terminator structure.

This is the first OATase sequence from any source to bereported. Although we were unable to compare our se-quence with either E. coli argE or P. aeruginosa argJdirectly, Southern blot analysis showed that P. aeruginosabut not E. coli genomic DNA contained sequences that couldhybridize to the cloned gonococcal argJ gene. The sequenceof a gene from the spirochete L. biflexa complementing an E.coli argE mutation has been previously reported, but thenature of the enzyme activity encoded by that gene, if any,was not defined (40). We compared this sequence with thegonococcal argJ gene and found no significant similarity ateither the DNA or the predicted amino acid sequence level,suggesting that these two genes are unrelated. We were alsoable to compare our sequence with the putative argJ genefrom B. subtilis (22). Significant identity was found betweenthe predicted amino acid sequences of these two genes. Thiscomparison revealed that these genes appear to be distantlyrelated, with several regions being highly conserved.

Picard and Dillon (23) found that their gonococcal argJlambda clone could complement both E. coli argE and argAmutations. They proposed that this may be the result of thegonococcal ArgJ enzyme producing the required acetylglu-tamate for the pathway, but because of the size of theirclone, they could not rule out the possibility that thegonococcal argA gene was also present. Our findingsstrongly support the former theory. We have shown thatargA complementation is dependent on an intact argJ geneand that no extra DNA outside the argJ gene is required forargA complementation. In addition, we have isolated a clonewhich complements argA in E. coli without complementingargE. On the basis of restriction mapping, this clone has nohomology to pRGE5 and likely contains the gonococcalargA gene (20).Our cloned argJ gene was capable of transforming the

gonococcal AHU strain 30465 to Arg+. These transformantsremained Ura- Hyx- as expected, since previous cotrans-formation studies have shown that these markers are notclosely linked (1, 18). These previous studies also reportedthat naturally occurring Arg+ revertants of AHU strains doarise, but at a very low frequency. The occurrence ofrevertants makes it unlikely that the argJ gene has beendeleted in AHU strains. In this study, it was found that N.gonorrhoeae NRL30465 contains an apparently intact argJgene as seen on Southern blots. The ability of deletions ofthe argJ gene to transform NRL30465 to Arg+ revealed thatthe argJ mutation in this strain is apparently located betweenthe internal ClaI and SacII sites. Experiments are in prog-ress to determine the nature of the argJ mutation in AHUstrains of N. gonorrhoeae. The data reported here will laythe groundwork for future studies on arginine biosynthesisand on the regulation of housekeeping genes in N. gonor-rhoeae.

ACKNOWLEDGMENT

This research was supported by Public Health Service grantA1-21264 from the National Institute of Allergy and InfectiousDiseases.

REFERENCES1. Atkinson, B. A., and M. H. Mulks. 1988. Linkage of genetic

markers in AHU strains of Neisseria gonorrhoeae, p. 223-230.In J. T. Poolman et al. (ed.), Gonococci and meningococci.Kluwer Academic Publishers, Dordrecht, The Netherlands.

2. Brown, K., P. W. Finch, P. T. Emmerson, and I. D. Hickson.1987. Complete nucleotide sequence of the Escherichia coliargA gene. Nucleic Acids Res. 15:10586.

3. Carifo, K., and B. W. Catlin. 1973. Neissena gonorrhoeaeauxotyping: differentiation of clinical isolates based on growthresponses on chemically defined media. Appl. Microbiol. 26:223-230.

4. Catlin, B. W., and E. H. Nash. 1978. Arginine biosynthesis ingonococci isolated from patients, p. 1-8. In G. F. Brooks, E. C.Gotschlich, K. K. Holmes, W. D. Sawyer, and F. E. Young(ed.), Immunobiology of Neisseria gonorrhoeae. American So-ciety for Microbiology, Washington, D.C.

5. Cunin, R., T. Eckhardt, J. Piette, A. Boyen, A. Pierard, and N.Glansdorff. 1983. Molecular basis for modulated regulation ofgene expression in the arginine regulon of Escherichia coli.Nucleic Acids Res. 11:5007-5019.

6. Cunin, R., N. Glansdorf, A. Pierard, and V. Stalon. 1986.Biosynthesis and metabolism of arginine in bacteria. Microbiol.Rev. 50:314-352.

7. Devereux, J., P. Haeberli, and 0. Smithies. 1984. A comprehen-sive set of sequence analysis programs for the VAX. NucleicAcids Res. 12:387-395.

8. Dougan, G., and M. Kehoe. 1984. The minicell system as amethod for studying expression from plasmid DNA. MethodsMicrobiol. 17:233-258.

9. Goodman, S. D., and J. J. Scocca. 1988. Identification andarrangement of the DNA sequence recognized in specific trans-formation of Neissenia gonorrhoeae. Proc. Natl. Acad. Sci.USA 85:6982-6986.

10. Gouy, M., and C. Gauthier. 1982. Codon usage in bacteria:correlation with gene expressivity. Nucleic Acids Res. 10:7055-7074.

11. Grosjean, H., and W. Fiers. 1982. Preferential codon usage inprokaryotic genes: the optimal codon-anticodon interactionenergy and the selective codon usage in efficiently expressedgenes. Gene 18:199-209.

12. Hawley, D. K., and W. R. McClure. 1983. Compilation andanalysis of Escherichia coli promotor DNA sequences. NucleicAcids Res. 11:2237-2255.

13. Hoare, D. S., and S. L. Hoare. 1966. Feedback regulation ofarginine biosynthesis in blue-green algae and photosyntheticbacteria. J. Bacteriol. 92:375-379.

14. Knapp, J. S., and K. K. Holmes. 1975. Correlation of auxotypeand disseminated infection among clinical isolates of Neisseriagonorrhoeae. J. Infect. Dis. 132:204-208.

15. Knapp, J. S., M. R. Tam, R. C. Norwinski, K. K. Holmes, andE. G. Sandstrom. 1984. Serological classification of Neisseriagonorrhoeae with use of monoclonal antibodies to gonococcalouter membrane protein I. J. Infect. Dis. 150:44-48.

16. Kyte, J., and R. F. Doolittle. 1982. A simple method fordisplaying the hydropathic character of a protein. J. Mol. Biol.157:105-132.

17. Martin, P. R., J. W. Cooperider, and M. H. Mulks. 1990.Sequence of the argF gene encoding ornithine transcarbamoyl-ase from Neissena gonorrhoeae. Gene 94:139-140.

18. Mayer, L. W., G. K. Schoolnik, and S. Falkow. 1977. Geneticstudies on Neisseria gonorrhoeae from disseminated gonococ-cal infections. Infect. Immun. 18:165-172.

19. Meile, L., and T. Leisinger. 1984. Enzymes of arginine biosyn-thesis in methanogenic bacteria. Experientia 40:899-900.

20. Mulks, M. H., and P. R. Martin. Unpublished data.21. Mulks, M. H., D. A. Simpson, and R. J. Shoberg. 1988.

Restriction site polymorphism in genes encoding type 2 but nottype 1 gonococcal IgAl proteases, p. 419-426. In J. T. Poolmanet al. (ed.), Gonococci and meningococci. Kluwer AcademicPublishers, Dordrecht, The Netherlands.

22. O'Reilly, M., and K. Devine (Trinity College, Dublin, Ireland).1991. Personal communication.

23. Picard, F. J., and J. R. Dillon. 1989. Cloning and organization ofseven arginine biosynthetic genes from Neisseria gonorrhoeae.J. Bacteriol. 171:1644-1651.

24. Pohlner, J., R. Halter, K. Beyreuther, and T. F. Meyer. 1987.Gene structure and extracellular secretion of Neisseria gonor-

J. BACTERIOL.

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

N. GONORRHOEAE argJ GENE 2701

rhoeae IgA protease. Nature (London) 325:458-462.25. Rossau, R., L. Heyndrickx, and H. Van Heluverswyn. 1989.

Nucleotide sequence of a 16S ribosomal RNA gene from Neis-seria gonorrhoeae. Nucleic Acids Res. 16:6227.

26. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1990. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

27. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

28. Shinners, E. N., and B. W. Catlin. 1978. Arginine biosynthesis inNeisseria gonorrhoeae: enzymes catalyzing the formation ofornithine and citrulline. J. Bacteriol. 136:131-135.

29. Smith, M. C. M., L. Czaplewski, A. K. North, S. Baumberg, andP. G. Stockely. 1989. Sequences required for regulation ofarginine biosynthesis promotors are conserved between Bacil-lus subtilis and Escherichia coli. Mol. Microbiol. 3:23-28.

30. Smith, M. C. M., A. Mountain, and S. Baumberg. 1986. Cloningin Escherichia coli of a Bacillus subtilis arginine repressor genethrough its ability to confer structural stability on a fragmentcarrying genes of arginine biosynthesis. Mol. Gen. Genet.205:176-182.

31. Southern, E. M. 1975. Detection of specific sequences amongDNA fragments separated by gel electrophoresis. J. Mol. Biol.98:503-517.

32. Stern, A., M. Brown, P. Nickel, and T. F. Meyer. 1986. Opacitygenes in Neisseria gonorrhoeae: control of phase and antigenicvariation. Cell 47:61-71.

33. Thony, B., and H. Hennecke. 1989. The -24/-12 promotercomes of age. FEMS Microbiol. Lett. 63:341-358.

34. Udaka, S. 1966. Pathway-specific pattern of control of argininebiosynthesis in bacteria. J. Bacteriol. 91:617-621.

35. Vogel, H. J., and D. M. Bonner. 1956. Acetylornithinase ofEscherichia coli: partial purification and some properties. J.Biol. Chem. 218:97-106.

36. Watson, M. E. E. 1984. Compilation of published signal se-quences. Nucleic Acids Res. 12:5145-5164.

37. West, S. E. H., and V. L. Clark. 1989. Genetic loci and linkageassociations in Neisseria gonorrhoeae and Neisseria meningiti-dis. Clin. Microbiol. Rev. 2(Suppl.):S92-S103.

38. West, S. E. H., and B. H. Iglewski. 1988. Codon usage inPseudomonas aeruginosa. Nucleic Acids Res. 16:9323-9335.

39. White, L. A., and D. S. Kellogg, Jr. 1965. Neisseria gonor-rhoeae identification in direct smears by a fluorescent antibody-counterstain method. Appl. Microbiol. 13:171-174.

40. Zuerner, R. L., and N. W. Charon. 1988. Nucleotide sequenceanalysis of a gene cloned from Leptospira biflexa serovarpatocwhich complements an argE defect in Escherichia coli. J.Bacteriol. 170:4548-4554.

VOL. 174, 1992

on May 28, 2018 by guest

http://jb.asm.org/

Dow

nloaded from