Mar 5, 1994 - X-Gal (35 pLg/ml) and isopropyl-3-D-thiogalactopyranoside. (IPTG) (8 ,ug/ml) as needed .... BLAST Network Service at the National Center for Biotech- ... output of 100% (Vibra Cell, Sonic and Material, Inc., Dan- bury, Conn.) ...
Vol. 176, No. 10
OF BACrERIOLOGY, May 1994, p. 2930-2937 0021-9193/94/$04.00 + 0 Copyright ©D 1994, American Society for Microbiology
JOURNAL
Role of Phosphoglucomutase in Lipooligosaccharide Biosynthesis in Neisseria gonorrhoeae ROBIN C. SANDLINt AND DANIEL C. STEIN*
Department of Microbiology, University of Maryland, College Park, Maryland 20742 Received 26 August 1993/Accepted 5 March 1994
A region of pSG30 that complements the pyocin-derived gonococcal lipooligosaccharide (LOS) mutants 1291d and 1291e was characterized by DNA sequence analysis and an open reading frame of 1,380 bases was identified that is 89% similar and 56% identical over 452 amino acids to the algC gene product from Pseudomonas aeruginosa that encodes phosphomannomutase. Enzymatic analysis of gonococcal crude protein extracts demonstrated that pSG30 encodes phosphoglucomutase (PGM) and phosphomannomutase activity. This activity is absent in 1291d and 1291e but is restored upon introduction of pSG30. PGM encoded by pSG34, a subclone of pSG30, was able to complement Escherichia coli PGM1, a strain deficient in PGM, as determined by bacteriophage C21 plaque formation. A revertant of 1291d that binds monoclonal antibody 2-1-L8 (specific for a 3.6-kDa LOS component) was isolated. The construction of a site-specific deletion of this region in the chromosome of 1291 confirms the role of this open reading frame in LOS biosynthesis.
The lipooligosaccharide (LOS) of Neisseria gonorrhoeae is an important virulence factor that is involved both in the disease process and in generating a host immune response (3, 6, 22). Previous studies have focused on characterizing LOS by physical and chemical means (2, 7, 11, 14, 23). Progress in the genetic study of LOS biosynthesis has been slower because of the limited availability of gonococcal LOS mutants. Resistance to the killing action of pyocin has been used to identify gonococci with altered LOS profiles (5, 8). LOS structural analysis of N. gonorrhoeae 1291 and its pyocinresistant derivatives, 1291d and 1291e, indicated that 1291d and 1291C fail to add the first glucose to their LOS chains (11). This suggested that the genetic defects present in these strains would be in either glucose metabolism or the attachment of glucose to the LOS chain. We previously cloned a DNA fragment capable of complementing the pyocin-derived mutants 1291d and 1291c (20). Deletion analysis of this clone (pSG30) identified a 750-bp region that could by homologous recombination repair the genetic defects of both 1291d and 1291l. In this article, we report on the sequence characterization of pSG30, the identification of the relevant gene involved in LOS biosynthesis, and demonstrate the functionality of the gene product by enzymatic analysis and complementation of a similar Escherichia coli mutant. The role of this region in LOS biosynthesis was confirmed by the construction of a sitespecific deletion in the chromosome of 1291.
containing ampicillin (30 ,ug/ml), kanamycin (30 ,ug/ml), erythromycin (300 pg/ml), and 5-bromo-4-chloro-3-indolyl-p3-D-galactopyranoside (X-Gal) (35 jLg/ml) as needed. M13 vectors were propagated on E. coli JM101 on YT plates containing X-Gal (35 pLg/ml) and isopropyl-3-D-thiogalactopyranoside (IPTG) (8 ,ug/ml) as needed (18). E. coli JM101 was maintained on M9 media (18). N. gonorrhoeae strains 1291, 1291d, and 1291e were obtained from Michael Apicella (University of Iowa, Iowa City) and have been previously described (5). N. gonorrhoeae strains were propagated on gonococcal medium base supplemented with Kellogg's solution (GCK) agar containing erythromycin (2 ig/ml) as needed and in GCP broth supplemented with Kellogg's solution (25) and 0.042% sodium bicarbonate. Chemicals, reagents, and enzymes. Restriction enzymes and T4 DNA ligase were purchased from New England Biolabs (Beverly, Mass.). Chemicals and other enzymes used were reagent grade or better. Monoclonal antibody (MAb) 2-1-L8 was generously provided by Wendell Zollinger, Walter Reed Army Institute of Research, Washington, D.C. MAb 3F1 1 was a generous gift from Michael Apicella. DNA manipulations. Plasmid DNA was isolated by the alkaline lysis procedure (4). Template DNA for sequencing reactions was isolated according to the Bethesda Research Laboratories M13 Cloning/Dideoxy Sequencing Instruction Manual (BRL, Gaithersburg, Md.). All enzymatic manipulations of DNA were done as recommended by the manufacturer. E. coli strains JM101 and DH5zMCR were made competent by using the CaCl2 procedure (18). The recombinant plasmids used in this study are shown in Fig. 1. DNA sequencing. DNA sequencing reactions were performed by the Sanger dideoxy method (15, 21) using the Sequenase version II sequencing kit (United States Biochemicals, Cleveland, Ohio) and ot-35S-dATP (New England Nuclear, DuPont, Boston, Mass.). Sequencing products were separated on a 4% acrylamide gel (55 cm by 0.2 mm; 7 M urea) with a 0.6-mm wedge in the last 10 cm. The sequencing buffer was TBE (100 mM Tris, 83 mM boric acid, 1 mM EDTA). The majority of the sequence information was generated by using single-stranded (M13) overlapping templates. Some sequence information was verified by using specific primers and a double-stranded template. The following primers were used in
MATERIALS AND METHODS Culture conditions. Strains and plasmids constructed or used in this study are listed in Table 1 and Table 2, respectively. E. coli DH5axMCR and E. coli JM101 were obtained from Bethesda Research Laboratories (Bethesda, Md.). E. coli PGM1 was obtained from Barbara Bachmann, E. coli Genetic Stock Center, New Haven, Conn. (1). E. coli DH5otMCR was grown in L broth, Luria-Bertani agar, or MacConkey agar * Corresponding author. Phone: (301) 504-5448. Fax: (301) 3149489. t Present address: Department of Microbiology and Immunology, Uniformed Services, University of the Health Sciences, Bethesda, MD 20814.
2930
LOS BIOSYNTHESIS IN N. GONORRHOEAE
VOL. 176, 1994
TABLE 1. Relevant characteristics of strains used or constructed in this study Source
Strain
N. gonorrhoeae 1291
1291d 1291e E48
RS1501-1llb RS159c RS152d RS132LU
M. Apicella M. Apicella M. Apicella Sandlin et al. (19) This study This study This study This study
Plasmid
pSG30 pSG516
3F11
-
+
-
-
-
-
-
-
-
+
-
+
E. coli
RE18f RS349
This study This study
pK18 pSG34
aThe ability to bind MAbs is designated with a +; the inability to bind MAbs is designated with a -. b Site-specific deletion of ORF 1. c RS15OA1-1 transformed with pSG30. d RS15OA1-1 transformed with pSG516. 'Spontaneous MAb 2-1-L8-binding strain of 1291d* f PGM1 transformed with pK18. g PGM1 transformed with pSG34.
this study (the location 5' to 3' in the sequence is indicated in parentheses): RS1, CGTCAACGAATGCGGCGGCAGCGG (416 to 439); RS2, GA(A)TTCCAAGAGGCGTGC (1195 to 1179) (a mismatch indicated by parentheses changed a T to an A to give an EcoRI site); RS3, TAATCGGGCGGATTGTG GCTGCCGG (475 to 451); RS5, GT[GCCCGCGTC GATGG (683 to 667); and RS7, GGCGAAATGAGCGGAC (1110 to 1125). Primers used for sequence analysis were prepared at the University of Maryland Protein and Nucleic Acids Facility. Sequence analysis. Sequence data were analyzed with GE NEPRO (Riverside Scientific, Seattle, Wash.) and PC Gene (Intelligenetics, Mountain View, Calif) software. Searches for sequence homologies were performed with the GENINFO BLAST Network Service at the National Center for Biotechnology Information. LFAST amino acid alignments were performed at the National Center for Biotechnology Information. Preparation of crude lysate. Crude lysates were prepared according to the method of Koplin et al. (12). Strains to be tested for mutase activity were grown to late logarithmic phase TABLE 2. Specific activities of PGM and PMM in crude extracts of gonococcal strains Strain
Source
Specific activity (mU/mg)" PGM
1291
RS15OA1-1 RS159 RS152
RS15OA1-1 pSG30' RS15OA1-1 pSG516d
E48
1291e 1291d
RS132L
32.4 0.7 129.4 1.1 0.6 0.4 1.0 0.9
10.6 0.3 12.8 0.7 0.2 0.5 0.5 ± 0.5 ± ± ± ± ± ± ±
PMM
1.6 ± 0.1
NDb 7.9 0.1 0.3 0.1
0.8 0.06 0.1 0.06 ND ND ± ± ± ±
a One milliunit is defined as the conversion of 1 nmol of substrate per min. Activity is expressed as means ± standard deviations.
bND, not determined. C
d
pSG30 is the clone containing pgm. pSG516 is the vector containing a similar-sized insert.
ORF I
MAb 3F1 1 Reactivity
MAb reactivity"
2-1-L8
2931
pSG30
E P S
Hc Hc
'
I
i
Hc H Hc E
I
I
S
+
H
pSG34
+
Pv
pSG56
' '
I
Pv
+
FIG. 1. Restriction map of pSG30 and relevant subclones. Each plasmid was tested for the ability to transform N. gonorrhoeae 1291e and 1291d to reactivity with MAb 3F11. A plus sign indicates that the plasmid was able to transform MAb reactivity. Construction of pSG30 and pSG34 was as previously described (20). pSG56 was cloned from a previously described exonuclease III deletion derivative, pSG44 (20), into pUC19, using PvuII sites flanking the MCS of M13mpl9. Restriction sites are abbreviated as follows: E, EcoRI; P, PstI; Hc, Hincll; H, HindIl; S, SphI.
in 20 ml of GCP broth supplemented with 0.042% sodium bicarbonate, Kellogg's solution (25), and erythromycin (2 ,ug/ml) as needed. The cells were pelleted at 8,000 rpm (SS34 rotor) for 5 min and were washed with 10 ml of sonication buffer (50 mM morpholinepropanesulfonic acid [MOPS], pH 7.0, 1 mM dithiothreitol, 3 mM EDTA). The cells were pelleted and resuspended in 2 ml of sonication buffer and were frozen at - 90°C. The thawed suspensions were sonicated at an output of 100% (Vibra Cell, Sonic and Material, Inc., Danbury, Conn.) for 10-s bursts with 30-s rests between bursts for a total of six bursts. The cell debris in the sonicated suspension was removed by centrifugation at 18,000 rpm (SS34 rotor) for 20 min. The resulting supernatant was stored at - 90°C and is the crude lysate. Protein concentrations in the crude lysates were measured by using the Bio-Rad protein assay dye reagent concentrate (Richmond, Calif.). Bovine serum albumin was used to generate the standard curve for determination of protein concentrations. The A595 was measured with the Gilford Response spectrophotometer. Protein concentration assays were performed three times. PGM assay. Phosphoglucomutase (PGM) activity was measured as described by Koplin et al. (12) except that the MgCl2 concentration was increased threefold. The 300-,ul reaction mixture for PGM contained 100 mM MOPS buffer (pH 7.6), 0.075 mM glucose 1,6-diphosphate, 5.1 mM MgCl2, 0.2 mM NADP, 0.5 U of glucose 6-phosphate dehydrogenase, and 50 ,ul of cell extract. The addition of glucose 1-phosphate to a final concentration of 3 mM started the reaction. The increase in optical density at 340 nm at 25°C was measured for 15 min. Controls included reaction mixture lacking substrate or cofactor. The positive control was the standard reaction components with PGM purified from rabbit muscle (Sigma). The enzymespecific activities were expressed as milliunits per milligram of protein, where 1 U is defined as the amount of activity required to reduce 1 p.mol of substrate (NADP) to product (NADPH) in 1 min. The amount of NADPH produced was calculated with the formula (24) CNADPH = A340/(am)(l) where am = molar absorption coefficient = 6,220 and 1 = length of cuvette = 1 cm. Each lysate was tested at least three times. Samples from two independent lysates were tested. PMM assay. Phosphomannomutase (PMM) activity was measured as described by Koplin et al. (12) except that the MgCl2 concentration was increased threefold. The 300-,ul PMM reaction mixture contained 100 mM MOPS buffer (pH 7.6), 30 mM MgCl2, 0.075 mM glucose 1,6-diphosphate, 1 mM
SANDLIN AND STEIN
2932
J. BACrERIOL.
GCCTGCATAA AAGGCTGCTT TTATTTCAGA CGGCATGTTA TTTGATGTAA CCGGGCTATG -3 5 -10 CGGCATACCG CTATG CGCCCGT
CCGTCTGAIA ACCGGACGCA AGC RBS
TTTTCCTAAA
CGGACACAAOSAACCTTA
TGGCAAGCAT CACCCGCGAC ATCTTCAAAG 180 A S I T R D I F K
M
CCTACGACAT CCGTGGCATC GTCGGCAAAA CCCTGACCGA CGATGCCGCT TATTTCATCG 240 A Y D I R G I V G K T L T D D A A Y F I GCAGGGCCAT CGCCGCCAAA GCCGCCGAAA AAGGTATCGC CCGCATCGCG CTCGGACGCG 300 G R A I A A K A A E K G I A R I A L G R
ACGGACGCTT GAGCGGCCCC GAACTGATGG AGCACATCCA ACGCGGCCTG ACCGACAGCG 360 D
G
R
L
S
G
P
E
L
M
E
H
I
Q
R
G
L
T
D
S
GTATCAGCGT ACTCAATGTC GGCATGGTTA CCACTCCTAT GCTCTACTTC GCAGCCGTCA 420 G
I
S
V
L
N
V
M
G
V
T
T
P
M
L
Y
F
A
A
V
ACGAATGCGG CGGCAGCGGA GTGATGATTA CCGGCAGCCA CAATCCGCCC GATTACAACG 480 N
E
C
G
G
S
G
V
M
I
T
G
S
H
N
P
P
D
Y
N
GTTTCAAAAT GATGCTCGGC GGCGACACGC TCGCAGGCGA AGCCATTCAA GAACTTTTAG 540 L A G E A I Q G F K E L L MH L G D T
CTATTGTTGA GAAAGACGGT TTTGTTGCCG CCGACAAACA AGGCAGCGTA ACCGAAAAAG 600 A I V E K D G F V A A D K Q T E K G S V ACATCTCCGG CGCATACCAC GACCACATCG TCGGACACGT CAAACTCAAA CGCCCGATAA 660 D I S G A Y H D H I V G H V K L K R P I
ACATCGCCAT CGACGCGGGC AACGGCGTGG GCGGCGCGTT TGCCGGCAAA CTCTACAAAG 720 N
I
A
I
D
A
G
G
N
V
G
G
A
F
A
G
K
L
Y
K
GTTTGGGCAA CGAAGTGACC GAACTTTTCT GCGAAGTGGA CGGCAATTTC CCTAATCACC 780 E V T E L F C E V D G L G N P N H G N F ACCCTGATCC TTCCAAACCG GAAAACCTGC AAGATTTGAT TGCCGCGCTG AAAAACGGCG 840 H P D P S K P E N L Q D L I A A L K N G
ATGCCGAAAT CGGCTTGGCG TTTGACGGCG ATGCCGACCG CTTGGGCGTG GTTACCAAAG 900 D A E I F D G D A D R L G V V T K G L A
ACGGCAACAT TATTTATCCC GACCGCCAAC TGATGCTGTT CGCCCAAGAC GTTTTGAACC 960 D G N I I Y P D R Q L M L F A Q D V L N GCAATCCCGG CGCGAAAGTC ATTTTCGATG TCAAATCCAC ACGCCTGCTT GCCCCGTGGA 1020 R N P G A K V I F D V K S T R L L A P W TTAAAGAACA CGGCGGAGAA GCCATAATGG AAAAAACCGG CCACAGCTTC ATCAAATCCG 1080 I K E H A I M I K S G G E E K T G H S F
CTATGAAAAA AACCGGTGCA CTGGTTGCCG GCGAAATGAG CGGACACGTT TTCTTTAAAG 1140 G E M S F F K L V A G H V T G A M K K
A
AACGCTGGTT CGGCTTCGAC GACGGCCTGT ATGCCGGCGC ACGCCTCTTG GAAATCCTGT 1200 D G L Y A G A R L L E I L G F D R W F
E
CCGCCTCCGA CAATCCGTCC GAAGTGTTGG ACAACCTGCC GCAAAGCATT TCCACGCCCG 1260 E V L D N L P S T P Q S I N P S S A S D
AACTCAACAT CTCCCTGCCC GAAGGCAGCA ACGGGCATCA AGTTATCGAA GAACTCGCCG 1320 N G H Q V I E E E G S L A S L P E L N I CCAAAGCCGA ATTTGAAGGC GCAACCGAAA TCATCACCAT CGACGGCCTG CGCGTTGAAT 1380 I I T I D G L R V E A T E F E G K A E
A
TTCCCGACGG CTTCGGTCTG ATGCGTGCTT CCAATACCAC GCCGATTTTG GTGTTGCGTT 1440 P I L V L R S N T T N R A F P D G F G L
TTGAAGCGGA TACGCAAGCA GCCATCGAGC GCATTCAAAA CCGATTCAAA GCCGTCATCG 1500 R I Q N R F K A V I A I E F E A D T Q A AAAGCAATCC GCATTTAATC TGGCCTCTGT AAAAATAGAA AAAATGCCOT CTGAAACTTT 1560 E S N P W P L H L I
FIG. 2. The DNA sequence of ORF 1 and flanking regions. The amino acid sequence encoded by ORF 1 is shown below the DNA sequence. A putative ribosome binding site is indicated by double underlines. The putative 10 and 35 regions are indicated by El and are indicated above the sequence. The stop codon of ORF 1 is indicated by a -. Gonococcal DNA uptake sequences are shown in boldface and occur at nucleotide positions 25, 59, 85, and 1546. -
-
NADP, 1 MM mannose-1-phosphate, 0.7 U of glucose 6-phosphate dehydrogenase, 1.1 U of phosphoglucose isomerase, and 0.9 U of phosphomannose isomerase. The reaction was started by the addition of 30 ,ul of cell extract. The reaction was monitored for 15 min at 25°C and the increase in A340 was measured. Enzyme activity was calculated as in the PGM assay. Standard deviations of all measurements were calculated by using the Excel program (Microsoft Corp.). Complementation of E. coli PGM1. pSG34 was introduced into E. coli PGM1 by transformation. One transformant, RE34, was chosen for further analysis. pSG34 is a subclone of pSG30 and was chosen for this analysis because of its stability in E. coli strains. The parent, pSG30, is unstable in E. coli. The vector, pK18, was introduced into E. coli PGM1 by transfor-
mation. One transformant, RE18, was chosen for further analysis. E. coli JM101 was used as the negative control; PGM1 was used as the positive control. Sensitivity to bacteriophage C21 was determined as described below (1). C21 was plaque purified two times on E. coli PGM1, and a phage stock was prepared. Two hundred microliters of an overnight culture of RE34, PGM1, RE18, and JM101 was added to 3 ml of soft agar containing 0.4% glucose and required antibiotics and was poured onto YT plates containing 0.4% glucose. Ten microliters of various dilutions of C21 (100 to 10 -4) was spotted onto the lawn, and the plates were allowed to incubate overnight. Infection was scored by the presence of plaques in the overlay. SDS-PAGE analysis. Proteinase K-treated whole-cell lysates were prepared from 18 to 20-h cultures by the procedure of Hitchcock and Brown (8). Approximately 1 ,Ig of LOS was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on duplicate 13% isocratic gels in a Tris-glycine buffer (0.025 M Tris, 0.192 M glycine, 0.1% SDS [pH 8.3]) at a constant current of 30 mA per gel for 4 h at 100C. One gel of each pair was fixed overnight in 40% ethanol-5% acetic acid. Periodate-oxidized LOS was visualized by silver staining (24). The other gel was electroblotted onto nitrocellulose overnight in a Tris-glycine methanol buffer (0.025 M Tris, 0.192 M glycine, 20% methanol) at a constant voltage of 30 V (25). Western blot (immunoblot) analysis of the LOS transferred to nitrocellulose was performed as described below. Immunological techniques. Binding of MAbs for Western blot analysis and for colony blots of individual colonies was performed as described below. Colonies were transferred to nitrocellulose filters by placing the filters on the surfaces of the agar plates and allowing the colonies to adhere for 1 min. The filters were blocked in filler solution (2% casein, 0.2% NaN3, and 0.002% phenol red, in 100 mM Na2HPO4 [pH 7.5]) for 30 min at room temperature. Incubation in primary antibody (MAb 2-1-L8, an immunoglobulin G [IgG] class mouse antibody, or MAb 3F11, an IgM class mouse antibody) at a dilution of 1:1,000 (2-1-L8) or 1:50 (3F11) in filler solution was done for 1 h. The filters were washed three times for 10-min intervals in 100 mM Na2HPO4. Binding of the primary antibody was detected by binding of the secondary antibody (goat antimouse IgG horseradish peroxidase to detect binding of MAb 2-1-L8 and goat anti-mouse IgM alkaline phosphatase to detect binding of MAb 3F11) for 2 h at a dilution of 1:3,000 in filler solution. Horseradish peroxidase secondary antibody was detected by a colorimetric assay with the following developing solution: 10 mg of 4-chloro-1-naphthol per ml in 0.86% hydrogen peroxide and 50 mM Tris hydrochloride (pH 8.0). Alkaline phosphatase secondary antibody was detected by a colorimetric assay with the following developing solution: 45 [lI of nitroblue tetrazolium salt (75 mg/ml in 70% dimethylformamide) and 35 ,ul 5-bromo-4-chloro-3-indolyl phosphate (50 mg/ml in dimethylformamide) in 10 ml of buffer C (100 mM Tris-HCl, 100 mM NaCl, 50 mM MgCl2 [pH 9.5]). Isolation of 1291d revertant. A broth culture of N. gonorrhoeae 1291d was grown to mid-log phase and various dilutions were plated on GCK plates. After overnight growth, these colonies were replica plated using 82.5 mm by 0.45 jim nitrocellulose disks (Schleicher and Schuell, Keene, N.H.) onto fresh GCK plates. The residual colony debris bound to this filter was used in an immunoassay using MAb 2-1-L8. The resulting MAb 2-1-L8-reactive colony was identified by aligning the filter with the replica plate. Colonies in the vicinity of the reactive colony were picked and then were analyzed by immunoassays. In this manner a reactive colony, RS132L, was identified and then single colony purified.
VOL. 176, 1994
LOS BIOSYNTHESIS IN N. GONORRHOEAE
PTLPASIFRAYDIRRVVGDTLTAETAYWIGRAIGSESLARGEPCVAVGR .
oooo00 0
000000.oo00000o
000
0
00000
ASITRDIFKAYDIRGIVGKTLTDDAAYFIGRAIAAKAAEKGIARIALGR *---SGPEUEHOflGLTDSG SVTMVT0
LYFAvNEc*
0000
DGRLSGPELMF3HIQ-RGLTDSGISVLNVGN{VTTPMLYFAAVNECGGSGVM TGONPPY NGFKI VYAGET-L NEQIQ LREREDIAG--VGSVE ITGSHNPPDYNGFKMMLGGDTLAGEAIQELLAIVEKDGFVAADKQGSVT QVDILPRYFKQIRDDIAMAKPMKVVVDCGNGVAGVIAPQLIEALGCSVI a
*-
*
oo0
0oo
o
oooooo0
**00000
0*
00o
0o
EKDISGAYHDHIVGHVILKRPINIAIDAGNGVGGAFAGKLYKGLGNEVT
PLYCEVDGNFPNHHPDPGKPENLKDLIAKVKAENADLGLAFDGDGDRVG ELFCEVDGNFPNHHPDPSKPENLQDLIAALKNGDAEIGLAFDGDADRLG
VVTNTGTIIYPDRLLMLFAKDVVSRNPGADIIFDVKCTRRLIALISGYG VVTKDGNIIYPDRQLMLFAQDVLNRNPGAKVIFDVKSTRLLAPWIKEHG %7z%Zl 0 000
0000*0000
sr I%E..Kvvr %7Z JJLJ%Z.L I oi%lft **0000000000000000000600000000000
RLLEILSQDQRDSEHVFSAFPSDISTPEINITVTEDSK-FAIIEALQRD 00000o0000
of
oo00
0oX0
*0oooooo0
394
oooo0
RLLEILSASDNPSE-VLDNLPQSISTPELNISLPEGSNGHQVIEELAAK
392
AQW-GEGNITTLDGVRVDYPKGWGLVRASNTTPVLVLRFEADPEEELER
442
00
0000000**00*0000000000
00000000000000000oo
AEFEGATEIITIDGLRVEFPDGFGLMRASNTTPILVLRFEADTQAAIER IKTVFRNQLKA 0 o o
0 a a
441
453
o o o
IQNRFKAVIES
452
FIG. 3. Amino acid sequence homology of P. aeruginosa algC gene product, PMM, with the ORF 1 protein. The amino acid sequence of PMM is given on top, starting from the sixth amino acid of its sequence, and the amino acid sequence of ORF 1 is given on the bottom, starting from the second amino acid of its sequence. The active site is underlined, with the serine residue that binds phosphate shown in boldface. Identical amino acids are indicated by *, while the conserved amino acids are indicated by o. There is 56% identity and 89% similarity between these protein sequences in a 452-amino-acid overlap.
Construction of a site-specific mutant. To construct a sitespecific mutant in the chromosome of N. gonorrhoeae 1291, the EcoRI fragment of pSG44 (16) was subcloned into pUC19. The resulting plasmid, pSG56, was digested with NaeI and religated to form the plasmid pSG57. The gonococcal insert in this plasmid contains an internal 405-bp deletion spanned by the NaeI sites. This construct (400 ng) was introduced into N. gonorrhoeae 1291 by transformation (22). Twenty colonies were picked from the experimental transformation and were single colony purified four times. Single colony derivatives of the initial transformants were tested in a colony immunoassay using monoclonal antibody 3F11. Chromosomal DNA from colonies that did not react or reacted faintly with MAb 3F11 was isolated. PCR under the conditions described above using primers RS1 and RS2 was done to confirm the presence of the 405-bp deletion in the gonococcal chromosome. Nucleotide sequence accession number. The DNA sequence described in this paper has been submitted to GenBank and has been assigned accession number L23426. RESULTS Sequence analysis. In order to define the gene product(s) encoded by pSG30 that was involved in LOS modification in 1291d and 1291,, DNA sequence analysis of the region identified by genetic analysis was performed. Identified in this
2933
region was a 1,380-base open reading frame (ORF 1) that would encode a 460-amino acid protein with a predicted molecular weight of 49,465. A putative ribosome binding site (5/5 consensus) was found 6 bases upstream of the first in-frame ATG. A putative -10 region (5/6 consensus) 26 bases upstream of the ribosome binding site and a putative -35 region (4/6 consensus) with 17-bp spacing are found upstream of the protein start site. The DNA sequence is shown in Fig. 2. No additional ORFs with recognizable promoter elements were identified in the region involved in LOS modification. Homology studies. Searches using the amino acid sequences derived from the DNA sequence were performed on the GENINFO BLAST Network Service. Homology searches with this ORF identified extensive homology to a group of hexose mutase proteins from a variety of gram-negative species. The highest degree of homology was to Pseudomonas aeruginosa algC encoding PMM (463 amino acids total) (27). There were 56% identity and 89% similarity over a 452-amino-acid stretch between the protein encoded by ORF 1 (460 amino acids total) and the PMM encoded by algC. The amino acid sequence alignment of the algC PMM and the protein encoded by ORF 1 is shown in Fig. 3. Homology between these two proteins is conserved through the entire protein sequence with several stretches of identical residues. The putative active site that is involved in phosphate binding is shown as an underlined sequence in Fig. 3. The five amino acids that are critical in forming the active site (10) are indicated in boldface. The predicted protein sequence and the PMM protein sequence are identical in these five amino acids. Homology to Xanthomonas campestris xanA (PMM and PGM activity) (12), E. coli pgm (PGM), Salmonella typhimurium cpsG (PMM) (10), and rabbit pgml (PGM) (16) was also observed (19). Construction of a defined mutant. In order to directly demonstrate that a defect in the ORF encoded by pSG30 was responsible for the mutant LOS phenotype in 1291d and 1291, a site-specific deletion of this ORF was constructed and introduced into the chromosome of 1291. The scheme used to construct this mutant is given in Fig. 4A. A 405-bp fragment was deleted from pSG56 by digestion with NaeI and ligation. This produces an in-frame deletion of 135 amino acids (Fig. 4). This construct, pSG57, was introduced into 1291 by transformation, and transformants were identified by the loss of MAb 3F11 binding. MAb 3F11 reacts with the LOS component of 1291 but does not react with the LOS component produced by 1291d and 1291,. Loss of MAb 3F11 binding was used to screen potential site-directed mutants. Several transformants that lost the ability to bind MAb 3F11 were identified. The desired genetic event must be differentiated from the phenotypic variation that occurs in these strains. Confirmation that the correct deletion was introduced into the chromosome was made by PCR analysis of this region using RS1 and RS2. Introduction of this 405-bp deletion results in the shift of the 775-bp PCR product to a 304-bp product. This is seen in strain RS15OA1-1 (Fig. 4B). The LOS of this site-specific mutant was analyzed by SDS-PAGE as shown in Fig. 5. The LOS pattern of RS15OA1-1 appears to be like that of 1291d and 1291,. The site-specific deletion strain produces primarily an LOS component with an apparent molecular weight of 3,200, with minor components of 4,700 and 4,200. When the plasmid pSG30 was introduced into RS15OA1-1 by transformation and erythromycin-resistant transformants were selected, it complemented RS15OA1-1 and restored the ability to bind MAb 3F11, while pSG516 had no effect on LOS biosynthesis. The plasmid pSG516 contains an unrelated insert that does not affect LOS biosynthesis and was used as a negative control. This plasmid
2934
SANDLIN AND STEIN
J. BACTERIOL.
ORF 1
1
2
3
4
5
6
7
1*
Nael Nael
Nael
pSG56
III 11111.iuI 1||111|
775 bp PCR Product
RS2
RS1
Digest with Nael and raligate Nael
pSG57
304 bp PCR Product
>
4
RS1
RS2
In-trame deltioon o 471 basas (157
aa)
FIG. 5. SDS-PAGE analysis of the site-specific deletion strain
_64 bases
RS15OA1-1 and the spontaneous revertant RS132L; silver staining of proteinase K-treated lysates subjected to electrophoresis on a 13% polyacrylamide gel. The lanes correspond to lysates prepared from the following strains: lane 1, 1291; lane 2, 1291d; lane 3, 1291e; lane 4, RS150A1-1; lane 5, RS132L; lane 6, RS152; and lane 7, RS159.
A. 1
2
4
3
5
6
861 bp 738 bp 369 bp 246 bp
B. 36
FIG. 4. Construction and confirmation of site-specific deletion in the gonococcal chromosome. (A) Strategy used to construct a sitespecific deletion in the gonococcal chromosome. ORF 1 is shown as a thick arrow. The region targeted for deletion is indicated in more detail below the chromosome map and is drawn to indicate the overall relation of this region to ORF 1. The region to be deleted is indicated by a hatched line between the two outer NaeI sites. The locations of the primers, RS1 and RS2, are indicated by arrows. The size of the PCR product encoded by the intact region and the deletion construct is also indicated. (B) Confirmation of the introduction of a site-specific deletion in the chromosome of N. gonorrhoeae 1291. Chromosomal DNA was isolated from 1291 and the deletion strain, RS15OA1-1, and was subjected to PCR with primers RS1 and RS2. These primers should amplify a 775-bp product in the wild-type chromosome and a 304-bp product if the deletion has been incorporated into the chromosome. The plasmid used to construct the deletion, pSG56, and its deletion derivative, pSG57, were included as controls. The PCR products were subjected to electrophoresis on a 1.0% TBE agarose gel. Lane 1, 123-bp ladder size standards; lane 2, pSG56; lane 3, pSG57; lane 4, 1291; lane 5, RS150A1-1; lane 6, 123-bp ladder size standards.
has been used previously (20) as a transformation control. From these data, we concluded that defects in the ORF in question are responsible for the altered LOS phenotype seen in 1291d and 1291e.
Isolation and characterization of revertant. When we examined the LOS made by 1291d and 1291e on SDS-PAGE, we observed a small amount of LOS with mobilities similar to the mobility of the parent, 1291. Both 1291d and 1291e have been shown to revert to reactivity with MAb 2-1-L8, at a frequency of about 10-3 per cell per generation. Therefore, it is possible that these bands represent contaminating LOS from revertants. In order to determine if reversion involved repair of the identified defect or occurred at a second site, a revertant of 1291d that bound MAb 2-1-L8, RS132L, was isolated. The LOS of this strain was subjected to SDS-PAGE analysis (Fig. 5) and Western blot analysis (data not shown). The predominant component produced by this strain was the 3,200-Da band. However, this strain gained the ability to produce the 3,600-Da band that binds MAb 2-1-L8. This strain also retained the ability to produce small amounts of the 4,700-Da and 4,200-Da components that are characteristically seen with 1291d. These data indicate that the LOS phenotype of cells that can bind MAb 2-1-L8 does not resemble that of the parent or that of the mutant. This suggests that the high-molecular-weight bands seen in 1291d and 1291e are made via an alternate biosynthetic pathway. Enzymatic analysis. Structural analysis suggested that 1291d and 1291e would be defective in glucose biosynthesis or the addition of glucose. DNA sequence analysis suggested that the gene involved was PGM. In order to determine if 1291d and 1291e were deficient in PGM activity, we analyzed crude protein extracts for the presence of PGM and PMM activity. Crude extracts were prepared from a variety of gonococcal strains including both wild-type and mutant strains. PGM activity was measured indirectly as the amount of NADPH produced in a coupled reaction with glucose 6-phosphate dehydrogenase (12). The results of this assay, recorded as specific activity, are shown in Table 2. The wild-type strain 1291 had an average specific activity of 32.4 mU/mg. The highest amount of PGM activity (129.4 mU/mg) was seen with RS159, a derivative of RS15OA1-1 that contains pSG30. RS15OA1-1 contains a deletion in the region of ORF 1 and displays an LOS phenotype similar to that of 1291, and 1291d. The mutant strains, E48, RS150OA-1, 1291d, 1291e, RS152 (RS15OA1-1 pSG516), and RS132L, had similar values of less than 2 mU/mg. These values represent the limits of the detection system. These data indicate that pSG30 encodes a PGM activity. Since ORF 1 had the highest homology with PMM, the lysates described above were tested for the presence of PMM activity to determine if the ORF 1-encoded protein
VOL. 176, 1994
also possessed PMM activity. PMM was measured indirectly in a coupled reaction with phosphomannose isomerase, phosphoglucose isomerase, and glucose 6-phosphate dehydrogenase. The amount of PMM present in the crude lysates was indirectly measured as the amount of NADPH formed. The results of the PMM assay recorded as specific activity are also shown in Table 2. RS159 gave a specific activity of 7.9 mU/mg. The wild-type strain, 1291, had a specific activity of 1.6 mU/mg. The mutant strains 1291e, RS152, and E48 had no detectable PMM activity. Because PMM activity was present only in wild-type strains or mutant strains with pSG30, these results suggest that pSG30 encodes a PMM activity in addition to the PGM activity. However, the specific activity is higher when glucose is the substrate. Complementation of E. coli PGM1. Bacteriophage C21 does not adsorb to wild-type E. coli K-12 but can adsorb and form plaques on mutants that produce a rough lipopolysaccharide (LPS) structure that does not contain glucose or galactose (1). Mutations in UDP galactose epimerase, UDP glucose pyrophosphorylase, and PGM result in this phenotype (1). The ability of the ORF encoded by pSG34 to complement PGM1 was determined by testing sensitivity to C21 in the presence and absence of pSG34. The results of this assay are shown in Table 3. PGM1 and RE18 supported the growth of C21, while JM101 and RE34 did not. This analysis indicates that the introduction of pSG34 results in the ability to produce an LPS molecule that masks the binding site of C21. The production of PGM complements PGM1. DISCUSSION Our previous genetic analysis indicated that the defects in 1291d and 1291e were independent but closely linked (20). The defects present in these strains were named Isi-4 and Isi-5 (20). Structural analysis of 1291d and 1291, suggested that these strains were defective either in the ability to produce the appropriate glucose precursor for incorporation into the LOS component or in the ability to attach glucose to the growing LOS chain (11). Sequence analysis of the region involved in LOS modification identified an ORF that encodes a protein with significant homology to PMM (algC) of P. aeruginosa and to PGM (pgm) of E. coli. There is extensive homology between PGM and PMM both within and between species, with the highest degree of homology centered around the active site that is involved in phosphate binding (16, 27). It has been shown in several different systems that defects in sugar biosynthetic enzymes result in LPS modification (1, 12, 13, 27). On the basis of the biochemical properties of E. coli clones expressing this ORF, we concluded that this ORF encodes a PGM and have named the genepgm. From these data, we have concluded that lsi-4 and Isi-5 are really the same gene and should now be referred to as pgm. Sequence analysis of the regions upstream and downstream of pgm did not identify any genes that are potentially involved in gonococcal LOS biosynthesis. Directly 5' to pgm is an ORF that encodes a putative protein with 52.6% identity and 83.1% similarity in a 95-amino-acid overlap with E. coli peptidylprolyl-cis-trans-isomerase b (19). This protein is involved in protein folding (9). Downstream of pgm is an ORF that encodes a putative protein with 49.2% identity and 82.4% similarity in a 244-amino-acid overlap with Bacillus stearothermophilus glnQ (19). This protein is involved in nitrogen transport (26). This demonstrates that pgm is not located in a LOS biosynthetic operon. In E. coli, pgm does not map to the rfa or rjb regions of the chromosome that are associated with LPS biosynthesis (1). PMM from P. aeruginosa does not map in
LOS BIOSYNTHESIS IN N. GONORRHOEAE
2935
a cluster and is separate from other mannose biosynthetic enzymes (e.g., algD) (27). However, in X. campestris, the genes encoding PGM-PMM and GDP-mannose pyrophosphorylase are linked and are located in a cluster of genes involved in xanthan gum biosynthesis (12). In order to confirm that the region identified by genetic analysis was responsible for the phenotype of 1291d and 1291", an in-frame site-specific deletion was introduced into the chromosome of 1291. This mutant produced an LOS phenotype similar to those of 1291d and 1291,. Colonies of 1291d and 1291e that bind MAb 2-1-L8 can be isolated at a frequency of about 1/4,000. Analysis of one of these MAb 2-1-L8-positive colonies derived from 1291d' RS132L, demonstrated a complex LOS phenotype that consisted of a mixture of LOS components ranging in size from a 3.2-kDa component (characteristic of 1291d) to a 4.7-kDa component (characteristic of 1291). The genetic basis of this revertant is not known. RS132L does not possess PGM activity, indicating that a compensating mutation in PGM is not responsible for the phenotype of RS132L. It is intriguing that this revertant can express a MAb 2-1-L8-binding component in the absence of PGM activity. This suggests either that the gonococcus may have multiple pathways to make glucose-iphosphate, but that under typical culture conditions, these pathways are suppressed, or that the gonococcus can substitute other sugars when elongating its LOS. This phenotype could also result from the ability to transport glucose-i-phosphate into the cell from the culture media. The ability to produce higher-molecular-weight components indicates that the block in the formation of precursor sugars has been circumvented. However, this mutation cannot completely compensate for the defect in PGM as evidenced by the inability to produce the parental LOS component in normal amounts. Enzymatic analysis of crude extracts prepared from both wild-type and mutant gonococcal strains demonstrates that PGM and PMM activities are present in strains that produce the 4,700-Da LOS component and absent in strains that produce primarily the 3,200-Da component. Introduction of pSG30 into RS150A1-1 restored mutase activity. The specific activity of PGM seen in the RS15OA1-1 genetic background was much higher than that seen in the genetic background of 1291. This higher activity is likely due to an increase in gene dosage. The plasmid pSG30 encoded both PGM and PMM activities, although the activity seen for glucose was 16 times higher than that seen for mannose. The reaction conditions in both the PGM assay and the PMM assay were performed at saturating substrate concentrations (12). The active site that is responsible for phosphate binding in PGM has been determined. However, the region of the protein that imparts substrate specificity has not been identified. Glucose and mannose are epimers of each other, differing at the position of the hydroxyl group on carbon 2 (28). Most systems have a specific mutase for the metabolism of glucose and mannose. However, the xanA gene product ofX campestris encodes both PGM and PMM activity in a bifunctional enzyme (12). Our gene product has the highest degree of homology with PMM enzymes, although the LOS structural analysis of 1291d and 1291e would indicate that it is PGM activity that is deficient in these strains. The results of enzymatic analysis demonstrate that the carbohydrate binding region of the pgm gene product can act on both glucose and mannose. Adhya and Schwartz isolated PGM1, a strain of E. coli K-12 defective in PGM activity and sensitive to C21, and demonstrated that the defect was most likely in the structural gene for PGM (1). The ability to complement E. coli PGM1 demonstrates that pSG34 produces a functional PGM activity in E.
2936
SANDLIN AND STEIN
J. BACIrERIOL.
TABLE 3. Complementation of PGM1 as assessed by sensitivity to bacteriophage C21 Strain
Plasmid
PGM1 RE18 RE34 JM1O1
None pK18 pSG34 None
Sensitivity to C21' S S R R
a Determined by the ability of C21 to produce plaques on the indicated strains. S indicates sensitivity to C21; R indicates resistance.
coli. Complementation of the PGM defect in PGM1 was assessed on the basis of the inability to support the growth of bacteriophage C21 (1). This complementation was specific to pSG34, as the introduction of the vector alone did not alter the ability to bind C21. This is biological evidence that pSG34 encodes a PGM activity. In our studies of 1291d and 1291e, we observed on silverstained SDS-PAGE gels minor amounts of 4.7- and 4.2-kDa LOS components, in addition to the predominant 3.2-kDa LOS component (20). These higher-molecular-weight bands were also observed in the site-specific deletion strain RS150A1-1. Colony blot analysis of several hundred individual colonies of 1291d and 1291, demonstrated that the colonies bound MAb 3F11, the 4.7-kDa LOS-component-specific antibody that is bound by the parent, 1291 (unpublished observations). This suggested that a small amount of the parental LOS component that could be the result of a leaky mutation, was produced by these strains. Enzymatic analysis indicates that these strains do not possess detectable PGM activity. However, the presence of intermediate LOS components in RS132L and the parental size component, coupled with the faint 3F11 reactivity, suggests that the gonococcus possesses an alternate mechanism to produce higher-molecular-weight components in the absence of PGM activity. Structural analysis demonstrates that 1291d and 1291e produce a minor component that may contain glucose (11). In E. coli, glucose-1-phosphate is converted to UDP-glucose. UDP-galactose is produced from UDP-glucose. The sugar nucleotide form of glucose and galactose is the activated form that is added to the growing LPS chain (13, 28). Because the gonococcus cannot transport galactose (17), the ability to produce UDP-galactose would be dependent on the ability to produce glucose-i-phosphate, unless another mechanism to produce galactose is present in the gonococcus. Following the addition of glucose to the LOS core, the following sugars are added in 1291: galactose, N-acetylglucosamine (GlcNAc), and galactose (terminal sugar) (11). Therefore, 1291 contains two galactose residues for each glucose residue. If glucose-iphosphate is limiting, the production of UDP-galactose would also be affected, and the production of LOS components would be modified. If the structure of the highest-molecular-weight component produced in 1291d and 1291e is like the 4,700-Da component of 1291, the presence of the 4,200-Da component in 1291d and 1291e suggests that the LOS component is truncated at the GlcNAc residue and cannot add the terminal galactose. This incomplete component becomes displayed on the exterior of the cell before the terminal galactose residue is added. The appearance of the 4,200-Da component would be expected if the level of UDP-galactose was limiting or the activity of the galactosyl transferase was modified. The presence of glucose in these components suggests that a glucose-i-phosphate can be produced and added to the LOS core in the absence of PGM, that an alternate form of glucose
can be activated and added to the LOS core, that an alternate sugar can be added to the core, or that an alternate LOS structure exists. The minor amounts and the presence of the 4,200-Da component suggest that the process is inefficient. We have isolated a MAb 2-1-L8-binding variant of 1291d, and analysis of this strain suggests that an alternate pathway exists in the gonococcus for the production of LOS components. ACKNOWLEDGMENTS This work was supported by grant A124452 from the National Institutes of Health to D.C.S. R.S. was supported by the Program in Cell and Molecular Biology at the University of Maryland. REFERENCES 1. Adhya, S., and M. Schwartz. 1971. Phosphoglucomutase mutants of Escherichia coli K-12. J. Bacteriol. 108:621-626. 2. Apicella, M. A., M. Shero, G. A. Jarvis, J. M. Griffiss, R. E. Mandrell, and H. Schneider. 1987. Phenotypic variation in epitope expression of the Neisseria gonorrhoeae lipooligosaccharide. Infect. Immun. 55:1755-1761. 3. Apicella, M. A., M. A. J. Westerink, S. A. Morse, H. Schneider, P. A. Rise, and J. M. Griffiss. 1986. Bactericidal antibody response of normal human serum to the lipooligosaccharide of N. gonorrhoeae. J. Infect. Dis. 153:520-526. 4. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. 5. Dudas, K. C., and M. A. Apicella. 1988. Selection and immunochemical analysis of lipooligosaccharide mutants of N. gonorrhoeae. Infect. Immun. 56:499-504. 6. Gregg, C. R., A. P. Johnson, D. Taylor-Robinson, M. A. Melly, and Z. A. McGee. 1981. Host species-specific damage to oviduct mucosa by N. gonorrhoeae lipopolysaccharide. Infect. Immun. 34:1056-1058. 7. Griliss, J. M., J. P. O'Brien, R. Yamasaki, G. D. Williams, P. A. Rice, and H. Schneider. 1987. Physical heterogeneity of Neisserial lipooligosaccharides reflects oligosaccharides that differ in apparent molecular weight, chemical composition, and antigenic expression. Infect. Immun. 55:1792-1800. 8. Guymon, L. F., M. Esser, and W. M. Shafer. 1982. Pyocin-resistant lipopolysaccharide mutants of N. gonorrhoeae: alterations in sensitivity to normal human serum and polymyxin B. Infect. Immun. 36:541-547. 9. Hayano, T., N. Takahashi, S. Kato, N. Maki, and M. Suzuki. 1991. Two distinct forms of peptidylprolyl-cis-trans-isomerase are expressed separately in periplasmic and cytoplasmic compartments of Eschenichia coli cells. Biochemistry 30:3041-3048. 10. Jiang, X. M., B. Neal, F. Santiago, S. J. Lee, L. K. Romana, and P. R. Reeves. 1991. Structure and sequence of the rjb (O antigen) gene cluster of Salmonella serovar typhimurium (strain LT2). Mol. Microbiol. 5:695-713. 11. John, C. M., J. M. Griffiss, M. A. Apicella, R. E. Mandrell, and B. W. Gibson. 1991. The structural basis for pyocin resistance in Neisseria gonorrhoeae lipooligosaccharides. J. Biol. Chem. 266: 19303-19311. 12. Koplin, R., W. Arnold, B. Hotte, R. Simon, G. Wang, and A. Piihler. 1992. Genetics of xanthan production in Xanthomonas campestris: the xanA and xanB genes are involved in UDP-glucose and GDP-mannose biosynthesis. J. Bacteriol. 174:191-199. 13. Makela, P. H., and B. A. D. Stocker. 1984. Genetics of lipopolysaccharide, p. 59-137. In E. T. Rietschel (ed.), Handbook of endotoxin, vol. I. Chemistry of endotoxin. Elsevier Science Publishers, New York. 14. Mandrell, R., H. Schneider, M. Apicella, W. ZoUinger, P. A. Rice, and J. M. Grilliss. 1986. Antigenic and physical diversity of Neisseria gonorrhoeae lipooligosaccharides. Infect. Immun. 54:6369. 15. Messing, J. 1979. A multipurpose cloning system based on singlestranded bacteriophage M13. Recomb. DNA Tech. Bull. 2:43-48. 16. Ray, W. J., Jr., M. A. Hermodson, J. M. Puvathingal, and W. C. Mahoney. 1983. The complete amino acid sequence of rabbit
VOL. 176, 1994 muscle phosphoglucomutase. J. Biol. Chem. 258:9166-9174. 17. Robertson, B. D., M. Frosch, and J. P. M. VanPutten. 1993. The role of galE in the biosynthesis and function of gonococcal lipopolysaccharide. Mol. Microbiol. 8:891-901. 18. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 19. Sandlin, R. C. Ph.D. thesis. University of Maryland, College Park. 20. Sandlin, R. C., M. A. Apicella, and D. C. Stein. 1993. Cloning of a gonococcal DNA sequence that complements the lipooligosaccharide defects of Neisseria gonorrhoeae 1291d and 1291e. Infect. Immun. 61:3360-3368. 21. Sanger, F., S. Nicklen, and A. R Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467. 22. Schneider, H., J. M. Grilliss, J. W. Boslego, P. J. Hitchcock, K. M. Zahos, and M. A. Apicella. 1991. Expression of paragloboside-like lipooligosaccharides may be a necessary component of gonococcal pathogenesis in men. J. Exp. Med. 174:1601-1605.
LOS BIOSYNTHESIS IN N. GONORRHOEAE
2937
23. Schneider, H., C. A. Hammack, M. A. Apicella, and J. M. Grifliss. 1988. Instability of expression of lipooligosaccharides and their epitopes in Neisseria gonorrhoeae. Infect. Immun. 56:942-946. 24. Segel, I. H. 1976. Biochemical calculations, 2nd ed., p. 324-346. John Wiley & Sons, Inc., New York. 25. White, L. A., and D. S. Kellogg, Jr. 1965. N. gonorrhoeae identification in direct smears by fluorescent antibody-counterstain method. Appl. Microbiol. 13:171-174. 26. Wu, L., and N. E. Welker. 1991. Cloning and characterization of a glutamine transport operon of Bacillus stearothermophilus NUB36: effect of temperature on regulation of transcription. J. Bacteriol. 173:4877-4888. 27. Zielinkski, N. A., A. M. Chakrabarty, and A. Berry. 1991. Characterization and regulation of the Pseudomonas aeruginosa algC gene encoding phosphomannomutase. J. Biol. Chem. 266:97549763. 28. Zubay, G. 1988. Biochemistry, 2nd ed., p. 647-668. MacMillan Publishing Co., New York.
