Vol. 50, No. 3INFECTION AND IMMUNITY, Dec. 1985, 738-7440019-9567/85/120738-07$02.00/0Copyright © 1985, American Society for Microbiology

Effects of Human Serum on the Growth and Metabolism ofNeisseria gonorrhoeae: An Alternative View of Serum

BRADLEY E. BRITIGAN,1 YAACOV CHAI,' AND MYRON S. COHEN2*

Departments of Medicine' and Microbiology and Immunology,2 University of North Carolina, Chapel Hill,North Carolina 27514

Received 16 May 1985/Accepted 5 September 1985

Humans are the sole reservoir of Neisseria gonorrhoeae, an organism which undergoes a marked increase inmetabolic rate after exposure to a low-molecular-weight, heat-stable component(s) of human serum. Furtherstudies on the effect of serum on gonococcal metabolism were undertaken. Gonococcal broth (GCB) iscommonly used for in vitro cultivation of gonococci. Gonococci suspended in GCB plus 10% serum exhibitedoxygen consumption rates of 139% (P < 0.01) and 456% (P < 0.01) of those suspended in GCB or Hanksbalanced salt solution, respectively. A twofold increase in growth rate also resulted from the addition of 10%serum to GCB. Gonococcal '4C-labeled adenine incorporation increased threefold with 10% serum supple-mentation of Hanks balanced salt solution. Dialysis of serum in 1,000-molecular-weight exclusion tubingremoved the stimulatory factor(s). Neither correction of anion-cation concentrations altered by dialysis noraddition of substances of known importance to the metabolism of gonococci (i.e., lactate, pyruvate, cysteine,ATP, AMP, NADPH, amino acids, malate, and glutathione) to dialyzed serum reconstituted stimulatorycapacity. The effect of serum on gonococcal glucose-catabolic pathways was measured by modifiedradiospirometry. An apparent threefold increase in Entner-DoudorofT and pentose phosphate pathwayactivities was induced by 10% serum, as was the increased shunting of glucose-derived glyceraldehyde-3-phosphate into these pathways. These metabolic changes did not allow specific identification of the serumstimulatory factor(s). Acetate, the major by-product of gQnococcal glucose catabolism, inhibited gonococcaloxygen consumption as previously reported. A high-molecular-weight serum component, probably albumin,reversed acetate-mediated inhibition of gonococcal oxygen consumption, identifying a second mechanism bywhich serum increases gonococcal metabolism. These results suggest that supplementation of growth mediawith serum should be considered to provide N. gonorrhoeae with-conditions more consistent with its normalenvironment.

Humans are the only reservoir of Neisseria gonorrhoeae,an organism responsible for both mucosal and disseminatedinfection (3). Common sites of gonococcal infection mustprovide factors critical for gonococcal growth. Serum hastraditionally been viewed as an important component of hostdefense against bacterial pathogens. However, previouswork from our laboratory has shown that gonococci exposedto serum undergo an immediate increase in their metabolicrate as reflected by a fourfold increase in oxygen consump-tion and adenine incorporation as well as by a threefoldincrease in glucose utilization (6). Vaginal mucosal secre-tions and ascitic fluid show similar stimulatory capability (5).Stimulatory activity persists despite heating (56°C for 30min) or boiling of serum before use. However, serum whichhas been dialyzed overnight in 3,500-molecular-weight ex-clusion tubing against Hanks balanced salt solution (HBSS)is no longer stimulatory. Its identity and mechanism ofaction remain unclear, although KCN prevents gonococcalresponses (6).A serum factor with characteristics similar to those asso-

ciated with metabolic stimulation is able to induce an alter-ation in gonococcal outer membrane phenotype andneutrophil association (2). Other investigators have reportedthe ability of serum factors to influence the susceptibility ofN. gonorrhoeae and of Haemophilus influenzae to serum-mediated killing (1, 20, 21, 26, 27, 30). These observationssuggest that gonococcal response to local environmentalfactors could be important in pathogenesis. Much is known

* Corresponding author.

about the physiology and metabolism of N. gonorrhoeae (4,9-14, 18, 22-25, 32), facets of which are influenced in vitroby a variety of exogenous factors. In this report, we explorethe effect of serum on gonococcal growth and glucose andoxygen metabolism, further characterize the serum stimula-tory factor(s), and examine the implications of these obser-vations on the cultivation of N. gonorrhoeae for studies ofpathogenesis.

MATERIALS AND METHODS

Preparation of log-phase gonococci. N. gonorrhoeae FA19(serum resistant) was subcultured daily on gonococcal broth(GCB) agar (GC medium base; Difco Laboratories, Detroit,Mich.) containing 1% Kellogg defined supplements 1 and 2 aspreviously described (17). Opacity was determined andpiliated colony variants were identified by using the criteriaof Swanson et al. (31). A nonpiliated opaque strain of FA19was used in all experiments.Gonococci were grown to log phase by inoculation ofGCB

agar colonies into GCB (Difco) containing 2 and 1% Kelloggdefined supplements 1 and 3 (0.5 M NaHCO3), respectively.Growth was allowed to continue to a Klett reading of 80 to100 U (Klett-Summerson colorimeter, filter no. 540; KlettManufacturing, Inc., N.Y.) at 37°C in 5% CO2. Lag-phasegonococci were harvested by scraping 16- to 24-h-old GCBagar colonies into the desired buffer. Final gonococcalconcentration was determined by direct enumeration in aPetroff-Hausser counting chamber or by measurement ofturbidity in a Klett-Summerson colorimeter which had pre-viously been correlated with direct counting or CFUs (2).

738

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF HUMAN SERUM ON NEISSERIA GONORRHOEAE

Gonococcal oxygen consumption. 02 consumption wasmeasured with a Clark oxygen electrode (Yellow SpringsInstrument Co., Yellow Springs, Ohio) by using a 1-mlvolume containing 108 gonococci. Results were expressed asthe maximal rate of 02 consumption (nanomoles perminute). Our previous work (6) demonstrated that serumincreased gonococcal 02 consumption in a dose-dependentfashion with no further increase in stimulation above a serumconcentration of 10%. Therefore, unless otherwise specified,10% serum was used in all assay systems.

[14C]adenine uptake. We grew 108 gonococci to log phasein GCB and then washed and suspended them in 1 ml of thedesired buffer containing 0.4 ,uCi of [14C]adenine (NewEngland Nuclear Corp., Boston, Mass.). Incorporation of[14C]adenine into macromolecules after a 30-min incubationat 37°C was determined by precipitation with cold trichloro-acetic acid as previously described (28).

Glucose utilization. Pathways of gonococcal glucose utili-zation were determined by using a modification of a previ-ously described procedure (6). Briefly, 108 log-phasegonococci suspended in phosphate-buffered saline (PBS) orin 0.9% saline (NS) were added to 25-ml Erlenmeyer flaskswhich contained 0.5 ml of either [1-14Clglucose, [3,4-14C]glucose, or [6-14C]glucose (0.025 ,uCi/ml of PBS or NS; NewEngland Nuclear) or [1_14C]acetate (0.5 ,uCi/ml of PBS orNS; New England Nuclear). The final volume of the reactionmixture was 1.5 ml. Suspensions also contained either 10%serum (vol/vol) or 150 ,ug of unlabeled glucose, which is anamount equivalent to that contributed by 10% serum. Flaskswere stoppered, pierced by a plastic well containing 0.2 ml of10% KOH to capture 14CO2, and incubated in a shaker waterbath at 37°C for 30 min. Reactions were ended by theaddition of 0.5 ml of 1 N HCl to the reaction mixture. Flaskswere reincubated for an additional 15 min, after which eachwell was transferred to a scintillation vial and counted in aPackard Tri-Carb liquid scintillation spectrometer (PackardInstrument Co., Inc., Rockville, Md.). This technique pro-vided a simple, reproducible means of measuring glucoseconsumption and CO2 generation over a defined time period.Addition of filter paper to the plastic wells to increaseabsorptive surface area did not increase recovery of 14CO2.

Gonococcal growth. Growth in broth cultures was deter-mined by measurement of Klett units or by serial dilutionand inoculation of bacteria onto GCB agar plates withdetermination of CFUs after 24 to 48 h of growth at 37°C in5% C02.

Preparation of serum. Serum from normal donors lacking ahistory of gonococcal infection was pooled, filtered, andfrozen at -70°C. Some serum was thawed and dialyzed(dialyzed serum) in 1,000- or 3,500-molecular-weight mem-brane exclusion tubing overnight (24 h) against 2 liters ofHBSS. Chloroform extraction was performed by combiningchloroform and serum in a 2:1 ratio. The mixture wasallowed to sit for 2 h, and the aqueous layer was removedwith a glass pipette. In some experiments, serum pH waslowered to 2.6 by addition of 1 N HCl for 1 h and then wasreturned to pH 7.4 to 7.5 by the addition of 1 N NaOH(acid-treated serum). For other work, monobasic sodiumphosphate was added to normal serum to achieve a finalphosphate concentration of 9.2 mg/dl of serum (postdialysisconcentration). CaCl2 and NaHCO3 were added to dialyzedserum to effect final Ca2+ and HC03- concentrations of 9.7mg/dl of serum and 21 meq/liter of serum (predialysis con-centration), respectively. Serum electrolyte concentrationswere determined by using a Beckman autoanalyzer (Electro-lyte-2; Beckman Instruments, Inc., Palo Alto, Calif.) and a

Gilford autoanalyzer (SBA 300; Gilford Instrument Labora-tories, Inc., Oberlin, Ohio).

Erythrocyte lysate. Heparinized whole blood from normalhuman donors was obtained, and erythrocytes were sepa-rated from leukocytes and serum via Plasmagel (RogerBellon, Neuilly, France) separation as previously described(6). Erythrocytes were washed twice in NS and were lysedwith sterile water.

Preparation of other reagents. KCN, lactate, pyruvate,cysteine, ATP, AMP, NADPH, NADH, glutathione, malicacid (Sigma Chemical Co., St. Louis, Mo.),acetate, CaC12,NaHCO3, Na2HPO4 (Fisher Scientific Co., Pittsburgh, Pa.),BME essential amino acids 100x (Flow Laboratories, Inc.,McLean, Va.), and minimal essential medium nonessentialamino acids 100x (GIBCO Laboratories, Grand Island,N.Y.) were suspended in HBSS. Amino acids were dilutedto 2x concentration (approximate concentration present in10% serum), and pH was increased to 7.4 by addition of 1 NNaOH. All of the other reagents were prepared such thatfinal concentrations were equivalent to those found in 8%serum (8): 5.4 mg/dl, 1 mg/dl, 0.8 mg/dl, 45 mg/dl, 1.1 mg/dl,35.4 mg/dl, 0.01 mg/dl, 0.07 mg/dl, and 0.5 mg/dl of serum forlactate, pyruvate, cysteine, ATP, AMP, glutathione,NADPH, NADH, and malate, respectively.

Statistics. Paired or unpaired Student's t tests were usedfor all statistical analyses. Results were considered signifi-cant if P < 0.05. Although for the purpose of the presenta-tion many of the data are expressed as a percentage ofappropriate control, only original (raw) data were used forstatistical comparisons.

RESULTS

We have previously shown that N. gonorrhoeae sus-pended in HBSS undergo a marked increase in oxygen,adenine, and glucose utilization after exposure to pooledhuman serum, the rates increasing in a dose-dependentfashion to a serum concentration of 10% (6). These meta-bolic events can be demonstrated in opaque, transparent,piliated, and nonpiliated colony variants as well as in gono-coccal strains isolated from. local or disseminated sites ofinfection (6).Serum supplementation of standard gonococcal growth

broth. Serum appears to provide a factor important togonococcal metabolism. Accordingly, we evaluated the ef-fect of serum supplementation of a commonly used gono-coccal growth broth (GCB plus 2% Kellogg'defined sup-plements 1 [glucose, L-glutamine, cocarboxylase], 1% 2[Fe(NO3)3], and 1% 3 [NaHCO3]) on several parameters ofgonococcal growth and metabolism. Oxygen consumptionrates for log-phase gonococci suspended in GCB were 321 ±49% (mean ± standard error, n = 4; P < 0.05) of those inHBSS (Fig. 1). Addition of 2, 1, and 1% Kellogg supple-ments 1, 2, and 3 to GCB did not increase oxygen consump-tion rates (Fig. 1). Gonococci in GCB supplemented with10% serum exhibited oxygen consumption rates of 139 ± 7%(mean ± standard error, n = 4; P < 0.01) of gonoco%ci inGCB alone and 456 ± 90% (mean ± standard error, n = 4; P< 0.01) of those in HBSS (Fig. 1).

Similar effects were seen when growth was assessed (Fig.2). A twofold increase in growth (as measured by viableCFUs after 4 h of incubation or by Klett units) resulted when10% serum was added to GCB containing Kellogg definedsupplements. However, increased agglutination (clumping)of gonococci was noted in the presence of serum (Britiganand Cohen, unpublished data). Differences in gonococcal

VOL. 50, 1985 739

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

740 BRITIGAN ET AL.

**

HBSS GCB GCB+ GCB +Supplemerits 10% Serum

I-En

FIG. 1. Comparison of peak 02 consumption of gonococci sus-

pended in HBSS, GCB, GCB plls,.2 and 1% Kellogg definedsupplements 1ito 3, respectively, orGCB plus 10% serum. Resultsare the mean + standard error of four separate experiments, eachdone in duplicate. *, Significant increase relative to HBSS (P <0.05); **, significant increase relative to HBSS, GCB, and GCB plus2 and 1% Kellogg defined supplements 1 to 3, respectively. (P <0.01).

agglutination could affect both Klett unit readings andCFUs, confounding data interpretation.

["'C]adenine incorporation via macromolecular (nucleicacid) synthesis provides further evidence of bacterial metab-olism and indirect evidence of the initiation of bacterialreplication. However, DNA synthesis cannot be differenti-ated from that of RNA. We measured cold trichloroacetic

109

* 10% SERUM* 6% BSA* CONTROL

108

A

300 _

200 _

150 -

io

F-

TIME (hours)

18

16

14

120

0

0

E

10

8

6

4

2

HBSS

10% Serum

H BSS

GCBGCB +10% Serum

5 10 15 20 25 30TIME (min.)

FIG. 3. Assimilation of ["'Cladenine into cold trichloroaceticacid-precipitable material by gonococci over 30 min while sus-

pended in HBSS, HBSS with 10% serum, GCB, or GCB with 10%serum. Results are expressed as mean ± standard error of threeseparate experiments. Addition of 10% serum to HBSS markedlyinicreased ["'C]adenine incorporation. Calculated uptake with GCBand GCB with 10% serum was much lower than uptake with HBSS,probably reflecting competition by unlabeled adenine found in GCB.

acid-precipitable ["'C]adenine from gonococci incubated for30 min in the presence and absence of 10% serum. Additionof 10% serum to HBSS increased recovery of trichloroaceticacid-precipitable ["'C]adenine from gonococci 331% relativeto HBSS alone (Fig. 3).- Calculated ma,cromolecular["'C]adenine incorporation for gonococci in GCB or GCBplus 10% serum was less than that for gonococci in HBSS(Fig. 3). It seems likely that GCB contains a large adeninepool which competes with "'C-labeled adenine for macromo-lecular incorporation, precluding interpretation of these lat-ter results.Pathways of glucose utilization after serum exposure. N.

gonorrhoeae glucose metabolism occurs chiefly via theEntner-Doudoroff and pentose phosphate pathways (11, 23,25), although the organism is capable of Embden-Meyerhofand tricarboxylic acid pathway activity as well (9, 11-14, 18,23) (Fig. 4). Previous work has shown that serum induces a

threefold increase in gonococcal glucose utilization relativeto HBSS (6). Activity of each glycolytic pathway is influ-enced by several different regulatory compounds. Determi-nation of the glycolytic pathway(s) stimulated by serumcould provide important clues to the nature of stimulatoryfactors. Each pathway generates CO2 from specific carbonatoms of the glucose molecul,e (Fig. 4). Determination of thepattern of 14CO2 generation from glucose selectively radio-labeled at different carbon positions (radiospirometry) al-

FIG. 2. Comparison of gonococcal growth in GCB and GCB with10% serum as assessed by CFU (A) and Klett units (B). Serumsupplementation of GCB increased gonococcal growth rates. BSA,Bovine serum albumin.

50C

EI 40 .E

C 30.00.

E: 20.Co0 .c00).

0~

INFECT. IMMUN.

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF HUMAN SERUM ON NEISSERIA GONORRHOEAE

Pentose Pathway1- CO2 NR Ribulose-5-P

Entner- Doudoroff- CO2 | -Acetate

Glucose -_ Glucose-6- p 4C " TCA

\-' Embden-Meyerhof "

3-CO2~ cttGlyceraldehyde-3-P {4-C 2Acet e

(Cl=6, 2=5,3=4) L............

FIG. 4. Summary of glycolytic pathways of N. gonorrhoeae. Contained within each box are the glucose carbon atoms from which the CO,generated from that pathway are derived. Of note is that, because of the activity of an isomerase reaction, the carbon atoms from carbonpositions 1 and 6, 2 and 5, and 3 and 4 become indistinguishable at the point of glyceraldehyde-3-phosphate production. Subsequentcatabolism of this compound, therefore, leads to equal CO, production from C-3 and C-4, C-2 and C-5, or C-1 and C-6.

lows identification of the relative activity of the variousglycolytic pathways (11, 16, 25, 33). Accordingly, gonococciwere incubated for 30 min in the presence of [1-'4C]glucose,[3,4-14C]glucose, or [6-'4C]glucose; 14CO2 generation wasmeasured. Phosphate containing (PBS) and non-containing(NS) buffer systems were both studied because of a previousreport suggesting that may influence gonococcal glycolyticpathway activity (11).

Yield of gonococcus-derived 14CO2 from each carbonposition was C, > C3,4 > C6 regardless of buffer system(Table 1). Gonococci suspended in NS exhibited 10-foldlower glycolytic activity relative to those suspended in PBS(Table 1). Addition of 10% serum to PBS induced a three tofourfold increase in 14CO2 generation from all three[14C]glucose molecules (Table 1). The supplementation ofNS with 10% serum resulted in CO2 generation equivalent tothat with PBS plus 10% serum (Table 1). This represented amore substantial increase relative to NS than to PBS be-cause of the lower activity seen with unsupplemented NS(Table 1).The increased CO2 production from C1 and C3,4 iS consis-

TABLE 1. Effect of 10% serum on gonococcal 14CO2 generationfrom selective ['4C]glucose and [14C]acetate isotopes

'4CO2 cpm recovered from buffer'4C Source Buffer supplemented with:

No serum 10% seruma

[1-_4C]glucose PBS 6,638 ± 348 20,703 ± 1,564 (3.1)NS 738 ± 82 20,177 ± 580 (27.3)

[3,4-14C]glucose PBS 1,960 ± 115 6,003 ± 522 (3.1)NS 384 ± 41 8,236 ± 45 (21.5)

[6-14C]glucose PBS 1,442 ± 251 5,294 ± 506 (3.7)NS 73 ± 21 5,028 ± 454 (68.8)

[1-_4C]acetate PBS 39,503 ± 4,756 4,540 ± 1,044 (0.1)NS 3,188 ± 279 4,320 ± 973 (1.4)

a For all glucose isotopes, 10% serum induced a significant increase in4CO2 recovery with both PBS and NS (n = 3 to 10; P < 0.001). Addition of10%o serum to PBS reduced 41CO2 recovery from [1-_4C]acetate relative toPBS (n = 4; P < 0.001) but did not alter recovery when added to NS (n = 4;P > 0.05). Numbers in parentheses indicate the ratio of 1'CO2 production in10%0 serum compared with production in the absence of serum for eachisotope and buffer. Reaction mixtures contained 20-fold greater counts of[1-14C]acetate as compared with glucose isotopes, precluding comparison ofdata from these two compounds. The concentration of [1-_4C]acetate was>1,000-fold that of unlabeled acetate, likely contributed by 1O0o serum (8).

tent with increased activity of the Entner-Doudoroff andpentose phosphate pathways (Fig. 4) (11, 23, 25). Gonococ-cal C6 utilization could increase as a consequence of in-creased tricarboxylic acid cycle activity. Alternatively, C6may be inserted into the C, position of glucose-6-phosphatederived from glyceraldehyde-3-phosphate, yielding CO2from the original C6 atom with further pentose phosphate orEntner-Doudoroff pathway activity (Fig. 4). The two possi-bilities can be differentiated by the effect of serum on 14CO2production from [1-_4C]acetate, which occurs only as aconsequence of the tricarboxylic acid cycle (9, 14). Sinceacetate utilization varies with glucose availability (9, 14),reaction mixtures were adjusted to contain equivalent glu-cose concentrations. [1-14C]acetate concentrations usedwere far below those reported to inhibit gonococcal metab-olism (22) (Table 2). 14CO2 generation from [1-14C]acetate(i.e., tricarboxylic acid cycle activity) did not increase in thepresence of 10% serum (Table 1). Organisms exposed toserum demonstrated a marked decrease in tricarboxylic acidcycle activity relative to those suspended in PBS.Mechanism(s) of serum effect on gonococcal metabolism.

Several investigators have reported that fatty acid by-prod-ucts of gonococcal catabolism inhibit subsequent growth andmetabolism (22, 32). Acetate is the principal product ofgonococcal carbohydrate metabolism (11, 23). Consistentwith previous reports (22, 32), addition of exogenous acetateto HBSS decreased gonococcal 02 consumption in a dose-dependent fashion, independent of pH change, which is alsoknown to influence gonococcal metabolism (23) (Fig. 5).Gonococci suspended in HBSS containing 0.010 N acetate(pH 4.0) showed no oxygen consumption, whereas thosesuspended in HBSS whose pH had been decreased to 4.0 bythe addition of HCl had oxygen rates only mildly reducedrelative to control. Addition of 10% serum negated theinhibitory effect of low acetate concentrations, resulting inoxygen consumption rates near those induced by serum inthe absence of exogenous acetate (Fig. 5). As acetate con-centrations were increased, greater serum concentrationswere necessary to stimulate gonococcal oxygen consump-tion (Fig. 5). Serum which had been dialyzed to remove thelow-molecular-weight factor(s) responsible for stimulation ofgonococcal metabolism and addition of bovine albumin toHBSS at protein concentrations found in serum were able toreverse the acetate effect, returning 02 consumption rates tonearly those of HBSS controls (Fig. 5).Nature of serum factor(s) stimulating gonococcal metabo-

lism. The dialysis procedure previously used to remove thestimulatory factor (6) was repeated by using 1,000-

VOL. 50, 1985 741

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

742 BRITIGAN ET AL.

TABLE 2. Effect of electrolytes, low-molecular-weight factors,acid-treated serum, and erythrocyte lysate onr gonococcal oxygen

consumption

% of controlIncubation conditiona (HBSS) oxygen

consumptionb c

10% Serum 427 ± 40d10% Serum + P042- 410 ± 39e10% Dialyzed serumf 99 ± 9910% Dialyzed serum (1,000 Mw) 122 ± 20Y

10% Dialyzed serum + Ca2+ + 106 ± 7HC03-

10% Dialyzed serum + lactate + 116 ± 10pyruvate + cysteine

10% Dialyzed serum + amino 94 ± 9acids + thiamine

10% Dialyzed serum + ATP 95 ± 210% Dialyzed serum + AMP 89 ± 110% Dialyzed serum + NADPH 81 ± 110% Dialyzed serum + malate 78 ± 810% Dialyzed serum + 87 ± 7

glutathione10% Dialyzed serum + ATP, 86 ± 5AMP, NADPH, malate, andglutathione

10% Dialyzed serum + NADH 127 ± 510% Chloroform-extracted serum 302 ± 18HBSS (chloroform extracted) 22 ± 6hErythrocyte lysate 96 ± 610% Acid-treated serum 380 ± 25e

a All incubation media contained HBSS.b Results are expressed as mean ± standard error peak 02 consumption

(nanomoles per minute) of three experiments.c Control (HBSS) gonococci consumed 7.4 ± 3 nmol of 02 per min.d Significant increase relative to control; P < 0.001.e Significant increase relative to control; P < 0.05.f 3,500-Mw dialyzed serum was used except as noted.g Significant decrease relative to HBSS plus 10% serum; P < 0.05.h Significant decrease relative to control; P < 0.05.

molecular-weight exclusion tubing. This also resulted in"inactivation" of the retentate (Table 2).An "add back" strategy was then employed in an attempt

to identify the serum factor(s) responsible for bacterialstimulation. We first examined the effects of dialysis onconcentrations of specific anions and cations. Na+, K+, Cl-,and Mg2+ concentrations were essentially unchanged. Dial-ysis elevated the serum P042- concentration from 3.7 to 9.2mg/dl of serum. Addition of P042- to normal serum toachieve the postdialysis concentration did not reduce serumstimulatory capacity (Table 2). Ca2' and bicarbonate con-centrations were reduced from 9.7 mg/dl of serum and 21meq/liter of serum to 7.3 mg/dl of serum and 1 meq/liter ofserum, respectively. Correction of these ion concentrationsdid not reinstate stimulatory activity to dialyzed serum(Table 2).We next examined a variety of low-molecular-weight

factors known to influence gonococcal metabolism. Cysteineis critical for gonococcal growth (4), and pyruvate andlactate may serve as gonococcal energy sources in theabsence of glucose (9, 25). Simultaneous addition of concen-trations of these substances found in serum to dialyzedserum did not renew stimulatory activity (Table 2). A varietyof amino acids are required by different gonococcal strainsfor growth (4). Concomitant addition of all essential andnonessential amino acids as well as cocarboxylase (thiamine)to dialyzed serum in physiologic concentrations was ineffec-

tive (Table 2). Malate and NADH have been reported tostimulate gonococcal oxygen consumption via a KCN-sensitive cytochrome system (18). NADPH and glutathioneplay important roles in oxygen metabolism (7, 19), and ATPand AMP concentrations are known to influence rates ofglucose in other biological systems (19). Addition of thesecompounds to dialyzed serum at physiologic concentrationswas similarly unsuccessful in reestablishing stimulatory ac-tivity (Table 2).The role of lipids other than fatty acids as factors in

gonococcal metabolism has not been extensively investi-gated. Removal of serum lipids by chloroform extractionresulted in a slight but statistically significant (P < 0.05)decrease in serum-induced stimulation ofgonococcal oxygenconsumption (Table 2). However, this probably representstoxicity of residual chloroform, as HBSS treated with chlo-roform in a similar fashion also lowered gonococcal metab-olism (Table 2).

Smith et al. have reported that a heat- and acid-labilelow-molecular-weight serum factor is able to convertgonococci from serum sensitive to serum resistant (20, 21).This factor appears to be a glucopeptide derived from lysederythrocytes (26, 27). Hemolyzed erythrocytes added toHBSS did not increase gonococcal oxgen consumption (Ta-ble 2).

DISCUSSIONIn this work, we set out to categorize the impact of serum

on the growth and metabolism of N. gonorrhoeae. This goalseems important, since N. gonorrhoeae is an obligate humanpathogen with no other natural reservoir. Previous work hasdemonstrated that exposure to a low-molecular-weight heat-stable factor in human serum initiates an immediate increasein gonococcal metabolism (6). Vaginal mucosal secretionsand human ascitic fluid have similar effects (5).Gonococcal growth broth (GCB plus Kellogg defined (17)

supplements 1, 2, and 3) is used in many laboratories as anoptimal growth medium for N. gonorrhoeae. As expected,gonococci suspended in GCB demonstrated oxygen con-sumption rates three times those of gonococci suspended inHBSS. Addition of Kellogg defined supplements, includingiron, to GCB had little effect. However, serum appears toprovide a factor not present in this artificial medium which isimportant to gonococcal metabolism. Addition of 10% serumto GCB yielded gonococcal oxygen consumption rates 139%of gonococci in GCB alone and induced a twofold increase ingonococcal growth. Serum also increased macromolecular['4C]adenine incorporation for gonococci in HBSS, provid-ing direct evidence for increased gonococcal metabolism andindirect evidence for gonococcal replication.

Quantitation of 14CO2 generated from glucose selectivelyradiolabeled at different carbon positions has been routinelyutilized to determine glycolytic pathways for various micro-organisms (11, 16, 25, 33). Consistent with previous reports(11, 23, 25), gonococci suspended in PBS or NS appeared torely on a combination of the Entner-Doudoroff and pentosephosphate pathways as their principal mechanisms of glu-cose catabolism. Serum markedly stimulated the activity ofboth pathways, as reflected in 14CO2 production from [1-14C]glucose and [3,4-14C]glucose. An increase in C6 utiliza-tion was also noted; this could be attributed to the recyclingof glyceraldehyde-3-phosphate through the pentose phos-phate or Entner-Doudoroff pathways or both (11, 16, 25, 33).Serum also appeared to decrease tricarboxylic acid cycleactivity among gonococci suspended in PBS.

Fatty acid by-products of gonococcal catabolism, includ-

INFECT. IMMUN.

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF HUMAN SERUM ON NEISSERIA GONORRHOEAE 743

z0I-~~~~~~~~~~~ta. ~~~~~I ttt~400-z0

3000

W 200a.

00

pH 7.0 4.0 7.0 4.0 10% 10% 0.6% 10% 10% 80% 80%+10% +10% Serum Diol BSA Serum Serum Serum DialSerum Serum Serum Serum

HBSS 0.003 N ACETATE 0.006 N ACETATE 0.01 N ACETATE

BUFFER

FIG. 5. Effect of the interaction of acetate and serum components on gonococcal oxygen consumption. Results are expressed as thepercentage of peak 02 consumption relative to gonococci suspended in HBSS (pH 7.4) (control) from three to six separate experiments. Dataare grouped according to the different acetate concentrations examined with the addition of various serum components noted (Serum, normalserum; Dial Serum, dialyzed serum; BSA, bovine serum albumin). Acetic acid (0.010 N) induced a significantly greater decrease in 02metabolism than did HBSS at comparable pH (4.0); P < 0.02. t, Significant increase relative to control (P < 0.001); tt, significant increaserelative to control (P < 0.02); ttt, significant increase relative to control (P < 0.05); *, significant decrease relative to control (P < 0.05); *significant decrease relative to control (P < 0.01).

ing acetate (the major by-product of gonococcal glucosecatabolism), have been previously shown to reduce gono-coccal growth and oxygen consumption (22, 32). Our resultsshow that this inhibition can be reversed by serum throughthe binding of acetate to serum proteins. Therefore, serumincreases gonococcal metabolism by providing a factorwhich stimulates directly as well as a separate factor whichinactivates inhibitory influences.The identity of the serum factor(s) responsible for the

direct stimulation of gonococcal metabolism remains un-clear. This factor is heat and acid stable and has a molecularsize of <1,000 daltons. Anion-cation concentration changesinduced by the dialysis procedure do not account for theresulting loss of stimulatory capacity in serum. Amino acids,thiamine, pyruvate, lactate, malate, and NADH are knownto increase gonococcal growth or to stimulate metabolism (4,9, 18, 23) and have molecular sizes of <1,000 daltons.Neither these compounds nor NADPH, ATP, AMP, orglutathione (which are of known importance to oxygen andglucose metabolism in other biological systems) (7, 19)reconstituted stimulatory activity when added to dialyzedserum at concentrations found in normal human serum.These results suggest that an add-back strategy is unlikely toallow final determination of the serum factor(s) important togonococcal metabolism.

Smith et al. have previously reported that a heat- andacid-labile low-molecular-weight factor in human serum andgenital secretions is able to induce resistance to antibody-complement-mediated killing in formerly serum-sensitivegonococcal strains (20, 21). This factor appears to be aglycopeptide derived from erythrocyte membranes (26, 27).The inability of erythrocyte lysate to induce a gonococcalmetabolic burst and the heat and acid stability of ourmetabolic factor(s) would suggest that these are two separatesubstances.Anderson et al. have reported that a heat-stable low-

molecular-weight serum factor induces serum resistance invitro and increases virulence in an animal model amongstrains of H. influenzae type b (1, 30). Only further clarifi-

cation of the identity of this factor and of ours will determinetheir relationship to one another.

Historically, serum has been viewed as a major compo-nent of host defenses against bacterial pathogens. Serum-derived complement and immunoglobulins may cause lysisof certain bacterial strains in vitro, and may provideopsonins which enhance ingestion of bacteria by phagocyticcells (15). However, the role of serum in defense againstgonococcal infection is unclear (3). Most gonococcal strainsresponsible for both mucosal and disseminated infections areresistant to antibody-complement-mediated killing by nor-mal pooled serum (29, 34). In focusing on the mechanisms bywhich serum and mucosal secretions protect the host frominfection, the incredible adaptability of pathogens such as N.gonorrhoeae has to some extent been overlooked. On thebasis of our work (2, 5, 6) and that of other investigators (1,20, 21, 26, 27, 30), an alternative view of the interactionbetween serum and N. gonorrhoeae appears warranted.These organisms are capable of a variety of rapid anddramatic changes in response to components of theirmicroenvironment. Exposure of bacteria to serum andmucosal secretions may promote metabolic and structuralchanges which enhance their resistance to several hostdefenses. Supplementation of artificial growth media withserum should be considered to provide conditions whichallow gonococci to grow in a fashion more consistent withtheir natural milieu.

ACKNOWLEDGMENTS

This work was supported by Public Health Service grantsA115036-07 and 2T32AI07001-09 from the National Institutes ofHealth and by an R. J. Reynolds Award. Yaacov Chai is supportedby the Israel Institute for Biological Research, Ness-Ziona, Israel.We acknowledge the excellent technical assistance of Marjorie

Huck Cooney and Martha French, and we thank Stephen Morse andP. F. Sparling for their helpful suggestions. The serum electrolytemeasurements were performed in the Clinical Research Laboratoryof North Carolina Memorial Hospital. Ronda Terrell and BelindaWilson helped with the preparation of this manuscript.

VOL. 50, 1985

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

744 BRITIGAN ET AL.

LITERATURE CITED1. Anderson, P., A. Flesher, S. Shaw, A. L. Harding, and D. H.

Smith. 1980. Phenotypic and genetic variation in the suscepti-bility of Haemophilus influenzae type B. J. Clin. Invest.65:885-891.

2. Britigan, B. E., and M. S. Cohen. 1985. Growth in serum inhibitsuptake of Neisseria gonorrhoeae by human neutrophils. J.Infect. Dis. 152:330-338.

3. Britigan, B. E., M. S. Cohen, and P. F. Sparling. 1985. Thegonococcus: a model of molecular pathogenesis. N. Engl. J.Med. 312:1683-1694.

4. Catlin, B. W. 1973. Nutritional profiles of Neisseria gonor-rhoeae, Neisseria meningitidis, and Neisseria lactima in chem-ically defined media and the use of growth requirements forgonococcal typing. J. Infect. Dis. 128:178-194.

5. Cohen, M. S., J. R. Black, R. A. Proctor, and P. F. Sparling.1985. Host defenses and the vaginal mucosa. Scand. J. Urol.Nephrol. Suppl. 86:13-22.

6. Cohen, M. S., and M. H. Cooney. 1984. A bacterial respiratoryburst: stimulation of the metabolisfn of Neisseria gonorrhoeaeby human serum. J. Infect. Dis. 150,49-56.

7. Cohen, M. S., and R. K. Root. 194t. The microbicidal mecha-nisms of human neutrophils and eosinophils. Rev. Infect. Dis.3:565-598.

8. Diem, K. 1962. Documenta Giegy scientific tables. Giegy Phar-maceuticals, Ardsley, N.Y.

9. Hebeler, B. H., and S. A. Morse. 1976. Physiology and metab-olism of pathogenic Neisseria: tricarboxylic acid cycle activityin Neisseria gonorrhoeae. J. Bacteriol. 128:192-201.

10. Holten, E. 1974. 6-phosphogluconate dehydrogenase and en-zymes of the Entner-Doudoroff pathway in Neisseria. ActaPathol. Microbiol. Scand. Sect. B 82:207-213.

11. Holten, E. 1975. Radiospirometric studies in genus Neisseria. 1.The catabolism of glucose. Acta Pathol. Microbiol. Scand. Sect.B 83:353-366.

12. Holten, E. 1976. Pyridine nucleotide independent oxidation ofL-malate in genus Neisseria. Acta Pathol. Microbiol. Scand.Sect. B 84:17-21.

13. Holten, E. 1976. Radiospirometric studies in genus Neisseria. 2.The catabolism of glutamate and fumarate. Acta Pathol. Micro-biol. Scand. Sect. B 84:1-8.

14. Holten, E. 1976. Radiospirometric studies in genus Neisseria. 3.The catabolism of pyruvate and acetate. Acta Pathol. Microbiol.Scand. Sect. B 84:9-16.

15. Horwitz, M. A. 1982. Phagocytosis of microorganisms. Rev.Infect. Dis. 4:104-123.

16. Katz, J., and H. G. Wood. 1963. The use of "4CO2 yields fromglucose-i- and -6-C'4 for the evaluation of pathways of glucosemetabolism. J. Biol. Chem. 238:517-523.

17. Kellogg, D. S., Jr., W. L. Peacock, Jr., W. E. Deacon, L. Brown,and C. I. Perkle. 1963. Neisseria gonorrhoeae. I. Virulencegenetically linked to clonal variation. J. Bacteriol. 85:1274-1279.

18. Kenimer, E. A., and D. F. Lapp. 1978. Effects of selectedinhibitors on electron transport in Neisseria gonorrhoeae. J.

Bacteriol. 134:537-545.19. Lehninger, A. L. 1970. Biochemistry, p. 409-412. Worth Pub-

lishers, Inc., New York.20. Martin, P. M. V., P. V. Patel, N. J. Parsons, and H. Smith. 1981.

Induction of phenotypically determined resistance of Neisseriagonorrhoeae to human serum by factors in human serum. J.Gen. Microbiol. 127:213-217.

21. Martin, P. M. V., P. V. Patel, N. J. Parsons, and H. Smith. 1983.Induction of serum resistance in recent isolates of Neisseriagonorrhoeae by a low molecular weight fraction of guinea pigserum. J. Infect. Dis. 148:334.

22. Miller, R. D., K. E. Brown, and S. A. Morse. 1977. Inhibitoryaction of fatty acids on the growth of Neisseria gonorrhoeae.Infect. Immun. 17:303-312.

23. Morse, S. A. 1977. Physiology and metabolism of Neisseriagonorrhoeae, p. 213-253. In R. B. Roberts (ed.), Thegonococcus. John Wiley and Sons, Inc., Publishers, New York.

24. Morse, S. A., and B. H. Hebeler. 1978. Effect of pH on thegrowth and glucose metabolism of Neisseria gonorrhoeae.Infect. Immun. 21:87-95.

25. Morse, S. A., S. Stein, and J. Hines. 1974. Glucose metabolismin Neisseria gonorrhoeae. J. Bacteriol. 120:702-714.

26. Patel, P. V., P. M. V. Martin, M. Goldner, N. J. Parsons, and H.Smith. 1984. Fractionation of guinea pig serum for an inducer ofgonococcal resistance to killing by human serum: active frac-tions containing glucopeptides similar to those from human redblood cells. J. Gen. Microbiol. 130:2757-2766.

27. Patel, P. V., P. M. V. Martin, M. Goldner, N. J. Parsons, and H.Smith. 1984. Red blood cells, a source of factors which induceNeisseria gonorrhoeae to resistance to complement-mediatedkilling by human serum. J. Gen. Microbiol. 130:2767-2770.

28. Rosenthal, R. S., R. S. Fulbright, M. E. Eads, and W. D.Sawyer. 1977. Ethylenediaminetetraacetic acid-sensitiveantiphagocytic acitivity of Neisseria gonorrhoeae. Infect. Im-mun. 15:817-827.

29. Schoolnik, G. K., T. M. Buchanan, and K. K. Holmes. 1976.Gonococci causing disseminated infection are resistant to thebactericidal action of normal human serum. J. Clin. Invest.58:1163-1173.

30. Shaw, S., A. L. Smith, D. Anderson, and D. H. Smith. 1976. Theparadox of Haemophilis influenzae type B bacteremia. J. Clin.Invest. 58:1019-1029.

31. Swanson, J., S. J. Kraus, and E. C. Gotschlich. 1971. Studies ongonococcus infection. I.P.1. and zones of adhesion: their rela-tion to gonococcal growth patterns. J. Exp. Med. 134:886-906.

32. Walstad, D. L., R. C. Reitz, and P. F. Sparling. 1974. Growthinhibition among strains of Neisseria gonorrhoeae due to pro-duction of inhibitory free fatty acids and lysophosphatidyleth-anolamine: absence of bacteriocins. Infect. Immun. 10:481-488.

33. Wang, C. H., I. Stern, C. M. Gilmour, S. Klungsoyr, D. J. Reed,J. J. Bialy, B. E. Christensen, and V. H. Cheldelin. 1958. Com-parative study of glucose catabolism by the radiorespirometricmethod. J. Bacteriol. 76:207-216.

34. Ward, M. E., P. J. Watt, and A. A. Glynn. 1970. Gonococci inurethral exudates possess a virulence factor lost on subculture.Nature (London) 227:382-384.

INFECT. IMMUN.

on February 21, 2020 by guest

http://iai.asm.org/

Dow

nloaded from