soluble plasma antigen in experimental salmonella typhimurium infection in mice

FEMS Immunology and Medical Microbiology 9 (1994) 325-332 © 1994 Federation of European Microbiological Societies 0928-8244/94/$07.00 Published by Elsevier

325

FEMSIM00441

Soluble plasma antigen in experimental typhimurium infection in mice

Salmonella

Thomas Butler a,*, Rial D. Rolfe b, Gina Marie James b and David J. Hentges b

a Department of Internal Medicine, Texas Tech University Health Sciences Center, Lubbock, Texas 79430, USA, b Department of Microbiology and Immunology, Texas Tech University Health Sciences Center, Lubbock, Texas 79430, USA

(Received 1 February 1994; revision received 14 July 1994; accepted 25 July 1994)

Abstract: To detect and characterize Salmonella antigen in blood, outbred CF-1 female mice were inoculated intraperitoneally with S. typhimurium LT-2 and blood was assayed by ELISA for Salmonella common structural antigen. Plasma antigen was detectable early in the course of infection and increased in quantity later in the course of illness when animals showed high grade bacteremia and high counts of splenic bacteria. Antigen was associated with a cell-free plasma fraction of blood, passed through filters with cut-offs of 0.2 g and molecular mass of 1000 kDa, and was enhanced in detectability after heating to 100°C for 15 min. Antigen was concentrated by diluting plasma 1:4 in 0.1 M EDTA, heating to 100°C, and concentrating the supernate with an ultrafiltration membrane with a molecular mass cut-off of 15 kDa. By gel filtration, antigen was associated with a peak at about molecular mass 300 kDa in heated plasma and a peak at about 380 kDa in unheated plasma. These results indicate that murine typhoid infection results in circulating soluble plasma antigen, which is heat-stable with a molecular mass of approximately 300 kDa.

Key words: Salmonella typhimurium; Antigenemia; Enzyme-linked immunoassay

Introduction

Detection of Salmonella antigens in blood and urine has been attempted for rapid diagnosis of human typhoid fever by investigators who used methods of counterimmunoelectrophoresis, co- agglutination, and ELISA [1-7], but successful use of these tests has been limited by their vari- able sensitivity and specificity. Some lack of specificity of tests may have resulted because antisera produced against Salmonella antigens

* Corresponding author. Tel: (806) 743 1215; Fax: (806) 743 3148.

cross-react with other bacterial species [8]. An- other source of poor specificity when plasma is used for antigen detection is cross-reaction between anti-Salmonella antibodies and transferrin [9].

Experimental infection with Salmonella typhimurium in mice [10,11] and rats [12] is an intracellular infection of mononuclear phagocytes in the spleen, liver, and bone marrow accompa- nied by bacteremia. This model of murine typhoid mimics human typhoid fever caused by S. typhi and paratyphoid fever caused by S. paraty- phi A by producing prolonged bacteremia and, in severe cases, death. Porin antigen of S. typhimurium has been detected in serum of experi-

SSDI 0928-8244(94)00049-2

326

mentally infected mice in quantities that corre- lated with counts of bacteria in the spleen [13]. To improve the performance of antigen detection tests, we investigated antigenemia in mice experimentally infected with S. typhimurium LT-2 using an ELISA with antibody against Salmonella common structural antigen that has been employed by other investigators in clinical studies [4,7].

Materials and Methods

Bacteria and lipopolysaccharide We used the LT-2 strain of S. typhimurium

which was cultivated in trypticase soy broth (TSB). To obtain heat-killed organisms for antigen standard, log-phase cultures that were turbid were centrifuged at 1500 x g for 15 min to sediment bacteria. After the supernatant was removed, the sediment was suspended in 0.9% NaC1 to the density of a 0.5 MacFarland standard and placed into a boiling water bath for 15 min. Lipopolysac- charide (LPS) W of S. typhimurium (Difco Labo- ratories, Detroit, MI) was dissolved in 0.9% NaCI for antigen detection.

Mouse infection We used outbred CF-1 female mice, which

were approximately eight weeks old and weighed about 20 g. An individual colony of S. typhimurium LT-2 was inoculated into TSB, which was incubated overnight at 35°C. To obtain log-phase growth, the turbid culture was diluted 1:100 in TSB and incubated at 35°C for 3 -6 h until turbid- ity equivalent to a 0.5 MacFarland standard was reached. Bacteria were centrifuged at 1500 × g for 10 min, the supernatant was poured off, and the pellet was resuspended in 0.9% NaC1 to the density of a 0.5 MacFarland standard (approx 1.5 X 108 cfu ml-1). Mice were inoculated intraperitoneally with approximately 1 x 105 cfu in 0.2 ml. Exact counts of bacteria were determined by performing viable counts on inoculum suspen- sions. This inoculum was about four times the LDs0 of this infection by the intraperitoneal route. The typical course of infection is for mice to appear ill after 2-3 days with piloerection and wasting. The spleens and livers increase in size

and are sites of intracellular multiplication of bacteria. The blood contains bacteria typically in counts per ml 103-104 times fewer than spleen and liver per g tissue. Deaths ensue 4-8 days after inoculation. Blood was obtained in Pasteur pipets containing approximately 50 units heparin sodium (Lyphomed, Deerfield, IL) per ml from the retro-orbital plexus. Blood was pooled from two or more mice to obtain required amounts. 20 /xl blood was streaked onto trypticase soy agar (TSA) plates for quantitative culture. Plasma was removed after blood was centrifuged at 1500 x g for 10 min. After bleeding, mice were killed and the spleens removed for homogenization in 1 ml cold 0.9% NaC1 in glass tubes fitted with teflon pestles. Serial ten-fold dilutions of the homogenates were prepared and 20 /xl volumes were spread in duplicate on TSA plates for colony counts.

Handling of mouse plasma Pools of plasma from infected and uninfected

mice were diluted 1:10 in phosphate buffered saline (PBS). In some experiments, diluted plasma was heated to 100 ° C in plastic microcentrifuge tubes placed in a boiling water bath for 15 rain. The tubes were centrifuged at 10000 x g for 10 min and clear supernates removed and frozen until tested for antigen. Diluted plasma was filtered through 0.2 micrometer syringe filters (Acrodisc R, Gelman Sciences, Ann Arbor, MI) and filtrates frozen until tested. Diluted plasma was filtered through type C cellulose ultrafilters with molecular mass cut-offs of 10 kDa or 100 kDa or type F polyvinylidene fluoride membrane with molecular mass cut-off of 1000 kDa (Spec- t r a / Pot R, Spectrum Medical Industries, Los An- geles, CA) and filtrates frozen until tested.

Concentration of plasma antigen was performed by the method described by McIllmurray and Moody [14]. Plasma after dilution to 1 :4 in 0.1 M ethylenediaminetetraacetic acid (EDTA) adjusted to pH 7.4 with sodium hydroxide was heated for 15 min in a boiling water bath and the precipitate sedimented by centrifugation at 2500 x g for 1 h. The supernatant was placed into a Minicon-B15 Concentrator (Amicon, Beverly, MA) with molecular mass cut-off of 15 kDa and

concentrated to one-fourth its original volume. Concentrates were frozen until tested.

Enzyme-finked immunoassay An ELISA kit (BacTrace 1~ Microwell kit,

Kirkegaard and Perry Laboratories, Gaithers- burg, MD) using affinity purified goat antibody directed against a common structural antigen (CSA-1) of Salmonellae was used. The antibody was made from serum of goats immunized with 17 serotypes of Salmonella that was adsorbed onto four additional serotypes of Salmonella. Cross-reacting antibodies were removed by ad- sorption onto Escherichia coli. The antibody re- acts with all 83 strains of Salmonella in the Amer- ican Type Culture Collection. The targets are both antigens released into solution upon heating at 100°C and antigens on the surface of bacterial cells. ELISA plates used for the assays were flat-bottomed microtiter 96-well plates (Immulon 2, Dynatech Industries, Alexandria, VA). The capture antibody was diluted 1 : 100 in the coating solution (PBS) and 0.1 ml was added to each well. Following incubation for 1 h, wells were emptied and blocked using 0.3 ml of a milk blocking solution and were incubated for an additional 10

327

min. Samples of 0.1 ml to be tested for antigen were placed in triplicate wells. The temperature of incubation of capture antibody and antigen was tested using 10 ° intervals between 20°C and 80°C. 60°C was selected because it gave an opti- mal difference between positive control and milk blocking solution. After an incubation of 1 h, the wells were emptied and washed twice with 0.3 ml wash solution (PBS-Tween 20). Into each well was placed 0.1 ml of the peroxidase labelled antibody, diluted 1 : 100 in milk blocking solution, for 1 h and washed three times. During the final wash, the wells were exposed to the wash solution for 5 min. The substrate provided in the kit was a two-component ABTS ~ (2,2'-azino-di-(3-ethylbe- nzthiazoline sulfonate)) and hydrogen peroxide system. The two reagents were mixed 1 : 1 (v/v) and 0.1 ml was added to each well. Plates were read in an ELISA plate-reader (Bio-Tek Instru- ments, Winooski, VT) at 405 nm after 2 h.

Gel filtration of plasma Pools of plasma from infected and uninfected

mice were diluted 1:1 in PBS. Aliquots of the diluted plasma were heated to 100°C in a boiling water bath for 15 min, centrifuged at 10000 × g

2-

o o

I I I

0 j l + ~ J i l

LPS Milk (ng/ml) D 50 25 12.5 6.3 3.2 1.6 0.8

LT-2 (cells/ml) O 2x10 5 l x 1 0 5 5x10 4 3x10 4 l x 1 0 4 6x10 3 3x10 3

Fig. 1. Detectability of whole heat-killed S. typhimurium LT-2 organisms and lipopolysaccharide (LPS) of S. typhimurium in solution by ELISA. Milk refers to milk blocking solution.

328

for 10 min and clear supernates were removed. 0.3 ml of heated and unheated plasma from infected and uninfected mice were applied to a Sepharose CL-6B column (25 cm × 1.5 cm; Phar- macia, Piscataway, N J). The samples were eluted with Tris (Tris(hydroxymethyl)aminomethane) buffer at pH 7.5 containing 0.05 M NaC1 (TBS) under constant pressure at a flow rate of 12 ml h - t . The effluent was monitored for optical ab- sorbance at 280 nm, with 1.0 ml fractions col- lected. To estimate molecular mass of antigen detected in column fractions, a linear regression was made from standards eluted from the column (Pharmacia, Piscataway, NJ). Standards included blue dextran for void volume, ferritin (molecular mass 440 kDa), aldolase (molecular mass 158 kDa), albumin (molecular mass 67 kDa), ovalbu- min (molecular mass 43 kDa), and ribonuclease (molecular mass 13 kDa).

Protein and carbohydrate determinations Protein concentrations were determined by the

procedure of Lowry et al. [15] with bovine serum albumin as the standard. A colorimetric assay based on the periodic ac id /Schif f stain was used to determine carbohydrate concentrations with glycogen as the standard [16].

Results

Sensitivity of ELISA for S. typhimurium and LPS Using a temperature of 60°C for reaction of

capture antibody with antigen, optical densities were determined for serial two-fold dilutions of heat-killed S. typhimurium LT-2 and LPS of S. typhimurium (Fig. 1). The limits of sensitivity for detection of LT-2 was about 104 organisms per ml and of LPS about 2 ng per ml.

Effects of heating and membrane filtration on antigen detection

When mouse plasma was diluted 1 : 10 in PBS, heating to 100°C for 15 min before testing in ELISA resulted in reduction in optical density for uninfected control plasma (Table 1). Heating diluted plasma of infected mice resulted in increases in optical density. Testing filtrates of diluted plasma showed that most antigen passed through filters with cut-offs of 0.2 /z and 1000 kDa. Filters with a cut-off of 100 kDa excluded most of the antigen in one experiment and a part of the antigen in a second experiment. Filters with a cut-off of 10 kDa excluded most of the antigen in both experiments.

Table 1

Effects of heating and membrane filtration on Salmonella antigen detection in plasma pooled from 2-3 uninfected and infected mice diluted 1 : 10 in phosphate buffered saline. Data given as optical density at 405 nm_+ S.D. of triplicate readings *

Unhea ted Hea ted 100°C 15 Min

Filtrates using membranes with cut-off size in/x or kDa:

0.2/x 1,000 100 10

Uninfected controls Experiment 1 0.178 + 0.018 0.146 + 0.003 Experiment 2 0.146 + 0.002 0.091 + 0.004

Infected mice Experiment 1 0.560 ___ 0.009 0.996 _+ 0.033

Blood culture LOgl0 c fu /ml : 3.8

Experiment 2 0.384 + 0.005 0.808 + 0.016 Blood culture Log10 c fu /ml : 2.7

0.521 _+ 0.009 0.500 + 0.011 0.085 + 0.003 0.107 _+ 0.005

0.398 +_ 0.012 0.366 _+ 0.014 0.325 + 0.059 0.092 + 0.005

* Mean optical densi ty+ S.D. for milk blocking solution background of these experiments was 0.079 + 0.008. Limits of detectability in each experiment were calculated to be the mean of uninfected control plasma plus 2 S.D. units; this was optical density 0.152 in experiment 1 and 0.099 in experiment 2.

329

Table 2

Relation of plasma Salmonella antigen to quantitative cultures of S. typhimurium LT-2 of whole blood, plasma, and spleen homogenates. Data given as optical density at 405 nm-+ S.D. of triplicate readings. * Blood from 2-3 mice was pooled to obtain sufficient volume

Days after Plasma optical density at 405 nm Geometric means of log10

inoculation Diluted 1 : 10 Diluted 1:4 in quantitative cultures

or in PBS, heated EDTA, heated to Whole-blood Plasma Spleen uninfected to IO0°C to IO0°C concentrated

Experiment 1 Uninfected 0.116_+0.007 0.202:1:0.018 < 1.7 < 1.7 ND Intraperitoneal 1 0.419 _+ 0.008 1.633 _+ 0.020 2.7 < 1.7 4.6 inoculum of 5 x 105 2 0.373+0.007 0.939-+0.017 2.0 <1.7 5.3 S. typhimurium LT-2 3 1.649 -+ 0.117 2.625 _+ 0.056 4.7 < 1.7 6.5

4 0.525 _+ 0.007 1.222 ± 0.098 3.1 < 1.7 6.2 5 0.120-+0.004 0.149-+0.004 < 1.7 < 1.7 3.3

Experiment 2 Uninfected 0.145 _+ 0.012 0.238 :t: 0.007 ND ND ND Intraperitoneal 2 0.276 + 0.015 0.674 ± 0.025 < 1.7 < 1.7 5.3 inoculum of 1 × 105 3 0.181 -+ 0.004 1.757 + 0.038 1.7 < 1.7 3.9 S. typhimurium LT-2 5 0.159+0.007 0.430-+0.009 < 1.7 < 1.7 3.3

* Mean optical density + S.D. for milk blocking solution background of these experiments was 0.078 + 0.007. Limits of detectability in each experiment were calculated to be the mean of uninfected control plasma plus 2 S.D. units.

ND: not done.

Relation of antigen to bacteremia and spleen cultures

O p t i c a l d e n s i t i e s o f d i l u t e d p l a s m a f r o m in-

f e c t e d m i c e v a r i e d f r o m levels i n d i s t i n g u i s h a b l e

f r o m u n i n f e c t e d c o n t r o l s to leve ls m o r e t h a n t e n

t i m e s g r e a t e r t h a n u n i n f e c t e d c o n t r o l s ( T a b l e 2).

T h e g r e a t e s t c o n c e n t r a t i o n s o f a n t i g e n w e r e de -

t e c t e d in m i c e t h a t s h o w e d t h e h i g h e s t g e o m e t r i c

2-

Infected Mice (heated)

s ~ Infected Mice , / \ ' . . , t~//(unheated)

.. ~ Control Mice VO s / (unheated) \ ~ (heated)

i I , . , •

. : y : 12: - - . _ -

0 . . . . I . . . . I . . . . I . . . . I . . . . I 10 20 30 40 50 60

Fraction Number

Fig. 2. Salmonella antigen in plasma of infected mice in fractions of gel filtration detected by ELISA. Plasma was either heated to 100°C followed by removal of precipitate by centrifugation or was applied unheated to column. Heparinized plasma was pooled from four mice bled one day after intraperitoneal inoculation with 2 × 106 viable S. typhimurium LT-2 when pooled blood culture

showed 7.2 × 10 4 organisms per ml. Heparinized plasma from uninfected control mice are also shown. V o indicates void volume.

330

mean numbers of bacteria per spleen. Most pools of infected mice had blood cultures positive for S. typhimurium, with the highest concentration of antigen from a pool with the greatest number of bacteria per ml. However, negative blood cultures occurred in two pools of blood, which showed presence of antigen. Cultures of plasma obtained before dilution in PBS were negative ( < 50 bacteria ml-1), indicating that centrifugation removed bacteria effectively from blood.

Our use of dilution of plasma in E D T A followed by heating and ultra-filtration of the supernatant resulted in concentration of antigen (Ta- ble 2). The increases in optical density following concentration varied from less than two-fold in uninfected control plasma to nearly ten-fold. Only one pool of plasma from infected mice, obtained five days after inoculation in experiment 1, failed to show antigen by optical densities of both diluted and concentrated plasma; this pool of blood showed a negative blood culture and the counts of bacteria in spleens were the lowest that were measured in all pools of infected mice.

Plasma antigen after gel filtration Antigen was detected after gel filtration in

plasma of infected mice that had not been heated

and plasma of infected mice that had been heated to 100°C for 15 min with removal of precipitate by centrifugation (Fig. 2). Heating resulted in a narrower, taller peak at a fraction number that corresponded to a molecular mass of about 300 kDa from a linear regression using molecular mass standards. The estimated molecular mass of the antigen peak of unheated plasma was 380 kDa. Protein concentrations in fractions were low in heated plasma compared to unheated, and no rise in protein concentrations was noted in infected plasmas in the fractions associated with antigen (Fig. 3). Carbohydrate assays showed peaks in unheated plasma at positions corresponding to molecular mass about 300 kDa and this material was completely removed by heating. No increase in carbohydrate concentrations occurred in fractions from heated plasma of infected mice.

Discussion

Our results of quantitative cultures of blood and splenic homogenates of mice experimentally infected with S. typhimurium confirmed previous studies [10,11] that murine typhoid is predomi-

1200-

1000-

800- E

.=_ 600-

2 O_

40o-

Control Mice Infected Mice :..'.... / (unheated) (unheated) ::

/ \ - / \ \

i / \ \ 1 \~ ::1 \\

.~1 \ Contro Mce "1 "~, . . . . . theated~ :1 ~ inlected MiCe

V ' / / ' / '....\ / (heated)

200- ,o / / - \ / J, ~/ z,. 4' .,,

o . . . . ~ ' ~ " r . ~ . ' ~ . . " 7 , ~-. ,'~,'-, . . .-~.--';-..:.........:- 10 20 30 40 50 60 70 80

Fraction Number

Fig. 3. Protein concentrations of fractions after gel filtration of heparinized plasma before and after heating from infected and control mice as performed in Fig 2. V o indicates void volume.

331

nantly an intracellular infection showing at least a thousand-fold more bacteria per gram of spleen than of blood. Salmonella antigen was detected in plasma with an ELISA employing polyclonal goat antiserum reactive against common structural antigen of Salmonella [4,7]. Most plasma pools from infected mice showed antigen, and greater concentrations of antigen were found in mice with high spleen counts of organisms and with bacteremia.

The Salmonella antigen detected in plasma of infected mice was shown to be stable after heating to 100°C. Furthermore, heating plasma increased detectability of antigen, suggesting that an inhibitor of the antigen was inactivated or that additional antigenic sites were exposed to the capture antibody following heating. The plasma antigen was not comprised of whole bacteria because viable bacteria were removed from plasma by centrifugation and the sensitivity of the ELISA for whole bacteria was about 104 ml-1, which was considerably more organisms than quantitative blood cultures showed the mice harbored. Be- sides, detectable plasma antigen passed through membrane filters with cut-offs of 0.2 Iz and 1000 kDa. Virtually no antigen passed through a filter with a cut-off of 10 kDa. This information on heat-stability and molecular size allowed us to increase sensitivity of the ELISA by concentrating the antigen by heating in a solution of E D T A and concentrating the supernatant by use of an ultrafiltration membrane with cut-off of 15 kDa.

The soluble plasma antigen detected in mice had a molecular mass of about 300 kDa. This molecular mass of 300 kDa was too large to be flagellin with a molecular mass around 55 kDa [17], the 55 kDa surface protein of S. typhimurium that conferred immunological protection described by Foulaki et al. [18], the 30 kDa antigens of the adhesive phenotype of S. typhimurium reported by Isaacson et al. [19], or the 17 kDa outer membrane protein that confers complement resistance to S. typhimurium [20,21].

The antibody used in our ELISA was polyclonal and affinity purified for common structural antigen, which is shared by several Salmonella serotypes. The ability of this antibody to detect whole organisms of different serotypes, LPS of S.

typhimurium in our study, and serum antigen in patients with S. typhi infection [4,7] suggests that the antigen detected in these mice is a component of the bacterial outer membrane [22]. It is possible that our antigen is LPS of S. typhimurium which was complexed with the 28 kDa binding protein in mouse serum described by Brade et al. [23]. Studies using monoclonal antibodies against defined Salmonella antigens might help to characterize further this plasma antigen. Alternatively, the use of recombinant or synthetic antigens as competitive inhibitors in the ELISA could assist in identifying which Salmonella antigen in mouse plasma was detected.

Previous efforts to make rapid diagnosis of typhoid fever by antigen detection have shown limited success because of poor sensitivity and specificity [2,6,7]. An attempt to detect endotoxin with the limulus test with a sensitivity of about 1 ng m1-1 in plasma gave negative results in typhoid fever [24], suggesting that low concentrations of bacterial antigen circulate during typhoid fever. Our finding of antigen in plasma of infected mice without bacteremia that could be concentrated and detected by ELISA suggest that further improvements in preparing plasma for antigen detection may enable development of more effective rapid diagnostic tests for typhoid fever [25].

Acknowledgements

This work was supported by Public Health Service grant R03-AI30994 from the Indo-U.S. Vaccine Action Program. The authors thank Woosun Song for technical assistance.

References

1 Appassakij, H., Bunchuin, N., Sarosombath, S., Rungpi- tarangsi, B., Manatsathit, S., Komolpit, P. and Sukosol, T. (1987) Enzyme-linked immunosorbent assay for detection of Salmonella typhi protein antigen. J. Clin. Microbiol. 25, 273-277.

2 Chaicumpa, W., Ruangkunaporn, Y., Burr, D., Chongsa- Nguan, M. and Echeverria, P. (1992) Diagnosis of typhoid

332

fever by detection of Salmonella typhi antigen in urine. J. Clin. Microbiol. 30, 2513-2515.

3 Chaicumpa, W., Thin-inta, W., Khusmith, S., Tapchaisri, P., Echeverria, P., Kalambaheti, T. and Chongsa-nguan, M. (1988) Detection with monoclonal antibody of Sal- monella typhi antigen 9 in specimens from patients. J. Clin. Microbiol. 26, 1824-1830.

4 Heiba, I., Girgis, N.I. and Farid, Z. (1989) Enzyme-linked immunosorbent assays (ELISA) for the diagnosis of en- teric fever. Trop. Geogr. Med. 41,213-217.

5 Sadallah, F., Brighouse, G., Guidice, G.D., Drager-Dayal, R., Hocine, M. and Lambert, P.H. (1990) Production of specific monoclonal antibodies to Salmonella typhi flagellin and possible application to immunodiagnosis of typhoid fever. J. Infect. Dis. 161, 59-64.

6 Sridharan, G. and John, T.J. (1986) Salmonella typhi. In: Antigen Detection to Diagnose Bacterial Infections, Vol- ume II, Applications (Kohler, R.B., Ed.), pp. 85-96. CRC Press, Boca Raton.

7 Araj, G.F. and Chugh, T.D. (1987) Detection of Salmonella spp. in clinical specimens by capture enzyme-linked immunosorbent assay. J. Clin. Microbiol. 25, 2150-2153.

8 Jesudason, M., Sridharan, G. and John, T.J. (1991) Anti- gen sharing of Salmonella typhi non-flagellar proteins with other salmonella and some shigellae and Escherichia coli. Ind. J. Exp. Biol. 29, 338-341.

9 Jesudason, M., Sridharan, G., Santosh, M. and John, T.J. (1992) Salmonella typhi cell wall protein showing antigenic similarity to transferrin. Ind. J. Exp. Biol. 30, 250-251.

10 Butler, T., Shuster, C.W. and Franco, A. (1979) Effect of a Salmonella group HI R factor on virulence and response of infections to antimicrobial therapy. Antimicrob. Agents Chemother. 15, 478-480.

11 MacKaness, G.B., Blanden, R.V. and Collins, F.M. (1966) Host-parasite relations in mouse typhoid. J. Exp. Med. 124, 573-583.

12 Nnalue, N.A., Shnyra, A., Hultenby, K. and Lindberg, A.A. (1992) Salmonella choleraesuis and Salmonella typhimurium associated with liver cells after intravenous inoculation of rats are localized mainly in Kupffer cells and multiply intracellularly. Infect. Immun. 60, 2758-2768.

13 Muthukkumar, S. and Muthukkaruppan, V.R. (1992) De- tection of porin antigen in serum for early diagnosis of mouse infections with Salmonella typhimurium. FEMS Mi- crobiol. Immunol. 89, 147-154.

14 Mclllmurray, M.B. and Moody, M.D. (1986) Latex agglutination. In: Antigen Detection to Diagnose Bacterial Infec- tions. Volume I. Methodology (Kohler, R.B., Ed.), pp. 9-26. CRC Press, Boca Raton.

15 Lowry, O.H., Rosebrough, N.F., Farr, A.L. and Randall, R.J. (1951) Protein measurement with Folin phenol reagent. J. Biol. Chem. 193, 265-275.

16 Mantle, M. and Allen, A. (1978) A colorimetric assay for glycoproteins based on the periodic/Schiff stain. Biochem. Soc. Trans. 6, 607-609.

17 Macnab, R.M. (1987) Flagella. In: Escherichia coli and Salmonella typhimurium. Cellular and molecular biology (Niedhardt, F.C., Ed.), pp. 70-83. American Society for Microbiology, Washington, DC.

18 Foulaki, K., Gruber, W. and Schlect, S. (1989) Isolation and immunological characterization of a 55-kilodalton surface protein from Salmonella typhimurium. Infect. Immun. 57, 1399-1404.

19 Isaacson, R.E. and Kinsel, M. (1992) Adhesion of Salmonella typhimurium to porcine intestinal epithelial surfaces: identification and characterization of two pheno- types. Infect. Immun. 60, 3193-3200.

20 Heffernan, E.J., Harwood, J., Fierer, J. and Guiney, D. (1992) The Salmonella typhimurium virulence plasmid complement resistance gene rck is homologous to a family of virulence-related outer membrane protein genes, in- cluding pagC and ail. J. Bacteriol. 174, 84-91.

21 Heffernan, E.J., Reed, S., Hackett, J., Fierer, J., Roudier, C. and Guiney, D. (1992) Mechanism of resistance to complement-mediated killing of bacteria encoded by the Salmonella typhimurium virulence plasmid gene. J. Clin. Invest. 90, 953-964.

22 Nikaido, H. and Vaara, M. (1987) Outer membrane. In: Escherichia coli and Salmonella typhimurium. Cellular and molecular biology (Niedhardt, F.C., Ed.), pp. 7-22. Ameri- can Society for Microbiology, Washington, DC.

23 Brade, L., Brade, H. and Fischer, W. (1990) A 28 kDa protein of normal mouse serum binds lipopolysaccharides of Gram-negative and lipoteichoic acids of Gram-positive bacteria. Microb. Path. 9, 355-362.

24 Butler, T., Bell, W.R., Levin, J., Linh, N.N. and Arnold, K. (1978) Typhoid fever. Studies of blood coagulation, bacteremia, and endotoxemia. Arch. Intern. Med. 138, 407- 410.

25 Rubin, F.A., McWhirter, P.D., Burr, D., Panjabi, N.H., Lane, E., Kumala, S., Sudarmono, P., Pulungsih, S.P., Lesmana, M., Tjaniadi, P., Sukri, N. and Hoffman, S.L (1990) Rapid diagnosis of typhoid fever through identification of Salmonella typhi within 18 h of specimen acquisi- tion by culture of the mononuclear cell-platelet fraction of blood. J. Clin. Microbiol. 28, 825-827.

soluble plasma antigen in experimental salmonella typhimurium infection in mice

Documents