INFECTION AND IMMUNITY, Feb. 1992, p. 590-595 Vol. 60, No.

2
0019-9567/92/020590-06$02.00/0
Copyright C) 1992, American Society for Microbiology

Antigenic Diversity in Haemophilus ducreyi as Shown by

Western Blot (Immunoblot) Analysis
ERWIN L. ROGGEN,* SABINE DE BREUCKER, EDDY VAN DYCK, AND PETER PIOT
Department of Microbiology, Institute of Tropical Medicine, Nationalestraat 155, B-2000 Antwerp, Belgium
Received 21 June 1991/Accepted 11 November 1991

The antigenic diversity within a panel of 63 Haemophilus ducreyi isolates was examined by Western blot
(immunoblot) analysis with a pool of 238 well-characterized human antisera. When a serum pool adsorbed on
a mixture of Haemophilus influenzae, H. parainfluenzae, and H. parahaemolyticus was used, the immunopro-
tiles suggested that prominent antigenic proteins involved in the human immune response have apparent
molecular masses of 63, 42, 34 to 30, and 28.5 to 28 kDa. Preliminary subcellular localization revealed that
these antigens are associated with the cellular membrane. Two subsets of antigens were discriminated by
detergent extraction. There was no evidence that the antigen composition is altered by changing the growth
conditions. With a serum pool adsorbed on the Haemophilus spp. mixture supplemented with Actinobacillus
actinomycetemcomitans, Pasteurella ureae, Neisseria gonorrhoeae, and Escherichia coli, antigenic determinants
more specific for H. ducreyi were identified. An immunodominant 28.5- to 28-kDa protein was expressed by all
H. ducreyi isolates. In the range from 34 to 30 kDa, 56 isolates revealed a dominant protein with variable
molecular mass. By using both proteins (28.5 to 28 kDa and 34 to 30 kDa) as immunotypic markers, seven
different immunopatterns were identified. Antigenic diversity among isolates from different geographical
origins as well as from a single area was observed.

Haemophilus ducreyi is a poorly known and fastidious the study of the human immune response to infection with
microorganism causing chancroid, a sexually transmitted this microorganism.
disease (STD) that is characterized by genital ulcers and by
abcedation of the inguinal lymph nodes (6). Recently, chan-
croid has received more interest because it is emerging as the MATERIALS AND METHODS
major cause of genital ulcer disease in several parts of the Bacterial isolates. H. ducreyi isolates were grown on both
developing world, and it is reappearing in poor minority MHVI agar (Mueller-Hinton agar [BBL] with 5% [vol/vol]
populations in Europe and the United States (10, 14, 16). In fetal calf serum [GIBCO], 5% [vol/vol] chocolatized horse
addition, several reports suggest that genital ulcer disease, in blood [GIBCO], 1% [vol/vol] IsoVitaleX [BBL], and 3 ,ug of
particular chancroid, is a risk factor for the sexual transmis- vancomycin [BBL] per ml) and on GCF agar (gonococcal
sion of the human immunodeficiency virus (9). Little is agar [GIBCO] with 1% [wt/vol] bovine hemoglobin [Be-
known about the antigenic composition of H. ducreyi, and thesda Research Laboratories], 5% [vol/vol] fetal calf serum
the nature of the immune response upon infection has been [GIBCO], 1% [vol/vol] IsoVitaleX [BBL], and 3 jig of
poorly studied. Serum immunoglobulin G (IgG) and IgM vancomycin [BBL] per ml). Incubation was for 3 days at
antibodies have been detected in patients with a clinical

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.

diagnosis of chancroid both by dot immunobinding and by an 35.5°C in a humidified CO2 atmosphere (Anaerobic GasPak
enzyme immunoassay (8, 13). Western blot analysis of sera System [BBL]) for MHVI medium and in a humidified
from normal and immunized rabbits has shown that rabbits aerobic milieu containing 5% [vol/vol] CO2 for GCF me-
immunized with H. ducreyi respond with a humoral immune dium. Haemophilus influenzae type b (ITG 3877), Haemoph-
ilus parainfluenzae (ITG 1094), and Haemophilus parahae-
response in which multiple antigenic polypeptides are de- molyticus (ITG 402) were grown on GCF agar (10 plates per
tected. The most prominent antigens are reported to have strain) for 48 h at 35.5°C and 5% [vol/vol] C02, as described
apparent molecular masses of 79, 62, 55, 49, and 26 kDa (7) for H. ducreyi. Neisseria gonorrhoeae (ITG 4446, ITG 4447,
and of 67, 42, 22.5, and 20 kDa (12). However, the specificity and ITG 4451), Actinobacillus actinomycetemcomitans (ITG
of the antigens towards H. ducreyi has not been demon- 1460), and Pasteurella ureae (ITG 3842) were grown on GC
strated. agar (Difco) supplemented with 1.7% [wt/vol] Bacto-Hemo-
In the present study, the antigenic diversity occurring globin (Difco) for 48 h at 35.5°C. Escherichia coli 0124 was
within a panel of H. ducreyi isolates was examined by suspended in liquid broth (0.5% [wt/vol] yeastolate [Difco],
Western blot (immunoblot) analysis. Prominent proteins 1.0% [wt/vol] sodium chloride [Merck], and 1.0% [wt/vol]
involved in the human immune response and H. ducreyi- Bacto-Trypton [Difco]). Incubation was in 50 ml of liquid
specific antigens were identified and localized. The depen- broth for 18 h at 37°C.
dency of the immunoprofiles on growth conditions was Preparation of antigen. Microorganisms were harvested in
assessed. On the basis of the immunotypic proteins, classi- 10 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic
fication of the H. ducreyi isolates resulted in seven immu- acid (HEPES), pH 7.0, and, like E. coli, were sedimented by
notypes. The information may be used for epidemiological centrifugation for 15 min at 1,200 x g. The pelleted cells
purposes, for the development of diagnostic tests, and for were washed twice with 50 ml of buffer and were stored
overnight at -20°C. The next day, they were thawed and
resuspended in 10 mM HEPES, pH 7.0. One-milliliter frac-
*
Corresponding author. tions of the cell suspensions were sonicated four times (20 s
590
VOL. 60, 1992 ANTIGENIC DIVERSITY IN HAEMOPHILUS DUCREYI 591

each) on ice. Sonication was performed with a Branson 0.05% [vol/vol] Tween 80 [Fluka]) and overnight with 75 ml
Sonifier 250 (duty cycle, 90; output, 1; diameter of the probe, of blocking buffer containing 1% [vol/vol] heat-inactivated
3 mm). The homogenates were centrifuged for 15 min at fetal calf serum. Antigenic determinants were localized by
1,200 x g to remove intact cells. The supernatants resulting sequential incubation for 2 h at room temperature with the
from the suspensions of H. influenzae, H. parainfluenzae, diluted serum pool and with goat anti-human IgG (Fc)
and H. parahaemolyticus as well as the supernatants result- alkaline phosphatase (Bethesda Research Laboratories), di-
ing from the suspensions of E. coli, N. gonorrhoeae, P. luted 1/3,000 in blocking buffer. Finally, the blot was incu-
ureae, and A. actinomycetemcomitans were pooled. To bated for 15 min at room temperature with 5-bromo-4-
these pools, hereafter referred to as the adsorbent and the chloro-3-indolyl phosphate and nitroblue tetrazolium from
supplemented adsorbent, respectively, sodium dodecyl sul- Bethesda Research Laboratories. In between the incuba-
fate (SDS) was added to a final concentration of 1% (wt/vol). tions, unbound antibody was removed by washing the nitro-
Like the individual SDS-free H. ducreyi supernatants, the cellulose sheet three times (10 min per wash) with blocking
adsorbents were stored at -20°C in 500-,ul aliquots with a buffer.
protein concentration of 10 mg/ml. The protein content of
antigen and adsorbent preparations was determined accord- RESULTS
ing to the method described by Bradford (2).
Subcellular fractionation. To assess the localization of the Western blot analysis of the H. ducreyi isolates by using a
H. ducreyi-specific antigens, a total membrane fraction was Haemophilus spp. adsorbed serum pool. Western blot analy-
obtained from the homogenate by ultracentrifugation for 2 h sis of 63 individual H. ducreyi lysates was performed with a
at 200,000 x g. Subsequently, this crude membrane fraction pool of human antisera after preadsorption of this pool on
was extracted with 2% (vol/vol) Triton X-100 in the presence the Haemophilus spp. adsorbent. H. ducreyi isolates re-
of MgCl2 (1 mM) for 1 h at room temperature. The insoluble vealed three separate clusters of antigenic determinants.
membrane fraction was recovered by ultracentrifugation as Cluster 1, which was bound by 63- and 42-kDa immuno-
described before. The protein content of the individual dominant proteins, and cluster 3, which was characterized
fractions was then determined (2). by several fuzzy immunoreactive bands with molecular
Serum pool. A serum pool was composed by using equal masses less than 20 kDa, were found to be very similar
volumes (100 ,ul) of 238 human serum samples. These sera among these isolates. Cluster 2 showed interisolate variation
were from patients with culture-proven chancroid as well as and was composed of a 28.5- or 28-kDa protein and a set of
from patients at STD clinics who were seropositive, as proteins with molecular sizes ranging from 34 to 30 kDa.
demonstrated by an experimental enzyme-linked immuno- This variation allowed the identification of 13 immunopro-
sorbent assay (ELISA) for the detection of IgG anti-H. files, of which one representative each is shown in Fig. 1.
ducreyi antibody (8). The specificity of this assay for the The specificity of the emerging immunoprofiles was sug-
diagnosis of acute chancroid in men and women is estimated gested by the absence of immunoreactivity of the serum with
at 98%. The selected sera originated from Kigali, Rwanda; adsorbent antigen and marker proteins on a Western blot.
The Gambia; Bangkok, Thailand; and Kinshasa, Zaire. The proteinaceous character of the antigens in clusters 1
Preadsorption of the serum pool. A 75-,ul portion of the and 2 was suggested by their sensitivity to proteinase K and
serum pool was incubated with 1.5 mg of adsorbent in a final by the lack of reaction with the periodic acid-Schiff reagent
volume of 50 ml of phosphate-buffered saline containing as well as with the biotin-labeled lectin from Phaseolus
0.05% (vol/vol) Tween 80. Incubation was for 1 h at room vulgaris. The determinants in cluster 3 were proteinase
temperature. Large aggregates were sedimented by centrif- insensitive, bound lectin, and were positive in the periodic
ugation for 5 min at 10,000 x g. The supernatant was diluted acid-Schiff reaction (data not shown).
to 150 ml, and another 1.5 mg of adsorbent was added. Because growth of H. ducreyi may be highly dependent on

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.

Incubation was allowed to proceed for 18 h at 4°C. The the culture conditions such that some isolates prefer MHVI
following day, residual aggregates were pelleted and the while others prefer GCF, the effect of these growth condi-
supernatant was immediately used for Western blot analysis. tions on the antigen composition was studied by using six
The effectiveness of preadsorption was demonstrated by the isolates growing on both media. Western blot analysis was
absence of immunoreactivity when preadsorbed pools of performed as described previously, and the results are
sera from Belgian STD patients and from African patients shown in Fig. 2. Only minor quantitative differences were
without STD were used for Western blot analysis on 20 observed.
randomly chosen H. ducreyi isolates. Western blot analysis of subcellular fractions of H. ducreyi.
Immunoblotting. The electrophoretic separation of pro- Total homogenates prepared from H. ducreyi isolates 3 to 7
teins was carried out with a Pharmacia Gel Electrophoresis and 11 (Fig. 1) were fractionated, and the subcellular frac-
Apparatus GE-2LS. After reduction, the proteins were alky- tions were analyzed by Western blot analysis with a serum
lated by 1.5% (wt/vol) iodoacetamide (Sigma). For protein pool adsorbed with the Haemophilus spp. adsorbent. The
separation, the buffer system of Laemmli (5) was used with immunoprofiles showed that all antigens believed to be
0.75-mm-thick 15% (wt/vol) polyacrylamide gels. Migration involved in the human immune response are associated with
of the proteins was allowed to proceed at 20°C for 1 h at 100 the membrane of the microorganism. Extraction of this
V and for 2 h at 220 V. Electrophoretic blotting was crude membrane fraction with Triton X-100 and 1 mM MgCl2
performed in an LKB Semy-dry blot cell with an initial revealed that two antigen subsets exist. The 63-kDa and
current of 0.6 mA/cm2. The electrotransfer was optimized by 28.5- to 28-kDa antigens were released under these condi-
slightly adapting the buffer conditions described by Dunn tions, while the 42-kDa and 34- to 30-kDa antigens remained
(3). Gels were incubated for 30 min at room temperature with membrane bound. This distribution of antigens was ob-
transfer buffer (10 mM NaHCO3, 3 mM Na2CO3, pH 10.4). served for all six H. ducreyi isolates but is shown only for
Subsequently, the proteins were transferred to a 0.22-pum isolate C90/81 (isolate 5 in Fig. 1) (Fig. 3).
nitrocellulose sheet for 2 h. After blotting, the nitrocellulose Western blot analysis of the H. ducreyi isolates by using a
was incubated for 2 h with 75 ml of blocking buffer (PBS plus serum pool adsorbed on the supplemented adsorbent. In an
592 ROGGEN ET AL. INFECT. IMMUN.

KDA KMA *.. KDA

94- _S4
Ci
I%W. %
..
-.4

-G8
j 43-
-43

C21 30

2!0. 1-
-20.1
C3
14.4 -14.4
3i

1 2 3 4 5 6 7 8 9 10 11 12 13 14
1 2 3 4 S 6 7
FIG. 1. Immunoprofiles representative of the H. ducreyi isolates
by using a Haemophilus spp. adsorbed serum pool. Protein (15 FLg) FIG. 2. Assessment of the effect of growth conditions on the
of the lysed H. ducreyi isolates was subjected to Western blot immunoprofiles of H. ducreyi. Lysed H. ducreyi isolates (15 ,ug)
analysis as described in Materials and Methods. The serum pool was grown on both culture media were subjected to Western blot
adsorbed on the mixture of Haemophilus spp. lysates. The emerging analysis as described in Materials and Methods. The serum pool was
immunoprofiles were grouped into 13 immunotypes. For each type, adsorbed on the mixture of Haemophilus spp. lysates. The emerging
a representative profile is shown. Analysis of the H. ducreyi CIP542 immunoprofiles were compared for both media. Lanes: 1, Hae-
antigen required the blot to be incubated for 30 min with alkaline mophilus spp. adsorbent; 2, HD47/86; 3, HD699/88; 4, BMC4/89; 5,
phosphatase substrate for unambiguous immunoprofiles to emerge. C90/81; 6, F91/81; 7, PU12/83. (In each set, the left lane is with
Lanes: 1, Haemophilus spp. adsorbent; 2, HD47/86; 3, HD699/88; 4, MHVI medium and the right lane is with GCF medium.) Marker
BMC4/89; 5, C90/81; 6, F91/81; 7, PU35/83; 8, PU12/83; 9, 247B1/86; proteins are phosphorylase b (94 kDa), bovine serum albumin (68
10, H2/86; 11, HD65/88; 12, C105/81; 13, HD569/87; 14, CIP542/54. kDa), ovalbumin (43 kDa), carbonic anhydrase (30 kDa), soybean
Origin and year of isolation of the individual isolates are given in trypsin inhibitor (20.1 kDa), and a-lactalbumin (14.4 kDa).
Fig. 5 (isolates indicated by asterisks). Marker proteins are phos-
phorylase b (94 kDa), bovine serum albumin (68 kDa), ovalbumin
(43 kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor
(20.1 kDa), and a-lactalbumin (14.4 kDa). Cl to C3, clusters 1 to 3 Analysis of a serum pool adsorbed on a Haemophilus spp.
(see Results). adsorbent suggested that proteins with molecular masses of
63, 42, 34 to 30, and 28.5 to 28 kDa may play an important
role in the human immune response and in the experimental
effort to improve the specificity of the immunoprofiles, ELISA (8). The abundancy and molecular mass of the larger
Western blot analysis was repeated with a serum pool protein led us to assume the presence of heat-shock proteins
adsorbed with the supplemented adsorbent. The number of in H. ducreyi. Further study is in progress to assess this
immunoprofiles was reduced to the seven profiles shown in possibility.
Fig. 4. The immunoreactivity with the 63-kDa immunodom- Of proteins identified in studies with rabbits, only a
inant protein was completely abolished, while the 42-kDa 62-kDa (7) and a 42-kDa (12) protein are similar to the
protein was hardly detected. However, the 28.5- and 28-kDa antigens identified in this study. This observation stresses

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.

proteins were still recognized. In the supplemented adsor- the need for careful evaluation of results from studies on
bent, an equivalent protein was detected only weakly. Very animal models before extrapolating to a human model.
pronounced was the effect of extended serum adsorption on A comparative study of the antigen composition of isolates
the immunodetection of the 34- to 30-kDa proteins. The grown on both MHVI and GCF culture media suggested that
majority of the immunoprofiles (56 of 63) showed one the immunopatterns were unaffected by changes in growth
immunodominant protein in this region. The different molec- conditions. This was supported by the absence of any
ular sizes observed included 30, 30.5, 31, 32, and 33 kDa. correlation between growth condition and antigenic diver-
sity. A similar observation was reported by Abeck et al. (1).
These investigators noticed that lipopolysaccharide pat-
DISCUSSION terns, but not protein patterns, were dependent on medium
composition, atmospheric conditions, and temperature. We
The antigenic determinants of 63 H. ducreyi isolates were were not able to assess the effect of temperature on the
characterized by Western blot analysis. The specificities of immunoprofiles because none of the isolates grew at temper-
the resulting immunoprofiles were improved by adsorbing atures outside the range of 30 to 35°C.
the serum pool on a Haemophilus sp. adsorbent supple- Preliminary localization of the antigens identified in H.
mented with N. gonorrhoeae, A. actinomycetemcomitans, ducreyi suggested that all important antigens are associated
E. coli, and P. ureae. Indeed, previous studies have reported with the membrane fractions but that the 63- and 29-kDa
cross-reactions of rabbit anti-H. ducreyi antisera not only antigens are associated in a way different from that of the
with Haemophilus spp. but also with Pasteurella sp. and other antigens. As far as the larger protein is concerned,
Actinobacillus sp. (15). Furthermore, Western blot analysis molecular weight, localization, and extractability by Triton
of N. gonorrhoeae and E. coli isolates with the human serum X-100 suggested the presence of a Chlamydia trachomatis-
pool revealed important immunodeterminants comparable to like, membrane-bound, GroEL-class antigen in H. ducreyi
determinants of H. ducreyi as far as molecular weight is (17). This issue now is under investigation.
concerned (data not shown). Most of the H. ducreyi isolates (57 of 63) expressed a
VOL. 60, 1992 ANTIGENIC DIVERSITY IN HAEMOPHILUS DUCREYI 593

KO
KDA
._4 -.4
-j4 -43 -43

43- -4 -30
,30
ebo. 4IM- - W-

420_ 30.1 -20J

-.20.1
20.1 _
1 X 3 4 5 0 ? a
144 _-
FIG. 4. Immunoprofiles representative of the H. ducreyi isolates
by using the serum pool adsorbed on the supplemented adsorbent.
T T S T Protein (15 ,g) of the lysed H. ducreyi isolates was subjected to
m Western blot analysis as described in Materials and Methods. The
serum pool was adsorbed on the mixture of Haemophilus spp.
F F lysates supplemented with a mixture of N. gonorrhoeae, A. actino-
mycetemcomitans, P. ureae, and E. coli lysates. The emerging
FIG. 3. Preliminary localization of H. ducreyi antigens. Individ- immunoprofiles were grouped into seven immunotypes. For each
ual subcellular fractions (15 ,ug) were analyzed by Western blotting type, a representative profile is shown. For analysis of the H.
with a serum pool adsorbed on the Haemophilus spp. mixture. The ducreyi CIP542 antigen, the blot was incubated for 30 min with
immunoprofiles shown were obtained with isolate C90/81 (Fig. 1). alkaline phosphatase substrate. Lanes: 1, HD699/88; 2, BMC4/89; 3,
TH, total homogenate; S, soluble fraction; TMF, total membrane C90/81; 4, F91/81; 5, PU35/83; 6, HD65/88; 7, supplemented adsor-
fraction; TX, Triton X-100 extract; OMF, outer membrane fraction bent; 8, ClP542/54. Origin and year of isolation of the individual
(Triton X-100-insoluble fraction). Marker proteins are phosphory- isolates are given in Fig. 5 (isolates indicated by open circles).
lase b (94 kDa), bovine serum albumin (68 kDa), ovalbumin (43 Marker proteins are phosphorylase b (94 kDa), bovine serum
kDa), carbonic anhydrase (30 kDa), soybean trypsin inhibitor (20.1 albumin (68 kDa), ovalbumin (43 kDa), carbonic anhydrase (30
kDa), and a-lactalbumin (14.4 kDa). kDa), soybean trypsin inhibitor (20.1 kDa), and ot-lactalbumin (14.4
kDa).

28-kDa protein, while a few (5 of 63) produced a 28.5-kDa

protein instead. Since the immunoreactivity of the serum
pool with these determinants was not negatively affected by cently, Sarafian and coworkers (11) reported that differences
either adsorbent and since in neither adsorbent an equivalent in genomic fingerprints could be used for typing H. ducreyi
protein showed similar immunodominance, these proteins isolates. How this typing correlates with the immunotyping

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.

were considered to be species specific. Recently, Karim and proposed in this report needs to be investigated.
coworkers (4) have reported the production of a monoclonal A low immunoreactivity between reference strain CIP542
antibody directed against a 29-kDa protein and reactive with (immunotype G) and adsorbed serum was observed. This
all strains of H. ducreyi. Although the relation between the observation may indicate that, since its isolation in 1954, the
two reported proteins remains to be demonstrated, our antigenic characteristics of this isolate have changed signif-
observation supports their usefulness as species-specific icantly because of cultivation in the laboratory, that this
markers. isolate is atypical, or that H. ducreyi as a species has
Great interisolate variation was observed among the anti- antigenically shifted over the years.
gens of the 34- to 30-kDa cluster. By using a serum pool
adsorbed on the supplemented adsorbent, the complexity of A low immunoreactivity between reference strain CIP542
this cluster was reduced to a single immunodominant pro- (immunotype G) and adsorbed serum was observed. This
tein, with variable molecular mass, present in 56 isolates. observation may indicate that, since its isolation in 1954, the
Whether the variation is due to expression of different antigenic characeristics of this isolate have changed signifi-
antigens or to size variation of the same determinant is not cantly because of cultivation in the laboratory, that this
yet clear. Whatever the reason, the appearance of this isolate is atypical, or that H. ducreyi as a species has
dominant protein was highly reproducible; consequently, the antigenically shifted over the years.
protein may be used as an immunotypic marker (Fig. 5). In conclusion, preadsorption of a pool of human sera
In this study, 63 H. ducreyi isolates were classified ac- allowed the identification by Western blot analysis of mem-
cording to the pattern of the 28.5- to 28-kDa and 34- to brane-bound proteins that may play a role in the humoral
30-kDa proteins. The classification is correlated with the immune response after infection by H. ducreyi and of
origin and year of collection. Though further analysis is species-specific as well as immunotypic markers. Serological
required, the classification suggests that there is as much classification of 63 H. ducreyi isolates yielded seven immu-
antigenic variation among isolates from one geographical notypes. This classification may be useful for studies on the
origin as there is among isolates from different areas. Re- diagnosis, epidemiology, and pathogenesis of chancroid.
594 ROGGEN ET AL.

~
E
C.Z-
o
r

0
00

ct-*

oCO
0 o
s_
U

CZ
0
U
0

U
0

0
o~
0

*c
.0

00
INFECT. IMMUN.

U
CZ0C
0

.^ ONON COON
CZ:
0) CZ
0.
0 .0~
ON 000 0000

0- u u r-
00 0° h DCEO 00
CO*
-o
Hs0 o oo
oo
0000t

ct
m ON ON OC) - ^O0
OINOC ON
u CO EC) 00 --ON0c0 N
ON

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.

C*

CO CO
.0 s

CO~~~~~~~~~~C 0.)
Ot _ O.0O-O
03 "C
000 000
0
00 0)

CO0
0 - 0 C C CO - 00 3 Q
cbo
N

0) 00 XX
r- A H _r 0 00 00
$.
. .
\C
~c
0 4 0000CA
O .0000 .00 en 0 (O 0nC

c0u C
,
CZO
'CO CO
U
0)U C,74

.0 Nt 0
CO0000
.000000000
00 0000
00 00
U
N ON) 0
0)
CZ 0) 0)~
r 00 . 0 g
4
-4N-4
CO \0 £0 00
._na ~00O)00
r- 00 00a'
00 O,
00000 C 0 0
00-
= 'IC "mON --- n0= : 0) cr-
CZ m 0)U
'I*IC
;iO)
C's
0- *At 0 CZ*
*-r
u~
1.0
CO

6
Ct
C-i
encn
000 SolWIOST JoO POj 0
H-
VOL. 60, 1992 ANTIGENIC DIVERSITY IN HAEMOPHILUS DUCREYI 595

ACKNOWLEDGMENTS 8. Museyi, K., E. Van Dyck, E. Vervoort, D. Taylor, C. Hoge, and

P. Piot. 1988. Use of an enzyme immunoassay to detect serum
This study was supported by a grant from the Swedish Agency for IgG antibodies to Haemophilus ducreyi. J. Infect. Dis. 157:
Research Cooperation with Developing Countries (SAREC). 1039-1048.
We thank S. De Breucker, D. Hendriksen, and K. Janssens for 9. Piot, P., and M. Laga. 1989. Genital ulcers and other sexually
technical assistance. Biological material was kindly provided by J. transmitted diseases as cofactors for the sexual transmission of
Bogaerts, E. Echeverria, D. Dance, M. Laga, W. Albritton, T. HIV. Br. Med. J. 298:623-624.
Lagergard, A. Sturm, and R. Ballard. 10. Piot, P., and A. Z. Meheus. 1985. Genital ulcerations in the
tropics, p. 207-236. In D. Taylor-Robinson (ed.), Clinical prob-
REFERENCES lems in sexually transmitted diseases. Martinus Nijhoff Publish-
1. Abeck, D., A. P. Johnson, R. C. Ballard, Y. Dangor, E. A. ers, Dordrecht, The Netherlands.
Fontaine, and D. Taylor-Robinson. 1987. Effect of cultural 11. Sarafian, S. K., T. C. Woods, J. S. Knapp, B. Swaminathan, and
conditions on the protein and lipopolysaccharide profiles of S. A. Morse. 1991. Molecular characterization of Haemophilus
Haemophilus ducreyi analysed by SDS-PAGE. FEMS Micro- ducreyi by ribosomal DNA fingerprinting. J. Clin. Microbiol.
biol. Lett. 48:397-399. 29:1949-1954.
2. Bradford, M. 1976. A rapid and sensitive method for the 12. Saunders, J. M., and J. D. Folds. 1986. Immunoblot analysis of
quantitation of microgram quantities of protein utilizing the antigens associated with Haemophilus ducreyi using serum from
principle of protein-dye binding. Anal. Biochem. 72:248-254. immunized rabbits. Genitourin. Med. 62:321-328.
3. Dunn, S. D. 1986. Effects of the modification of transfer buffer 13. Schalla, W. O., L. L. Sanders, G. P. Schmid, M. R. Tam, and S.
composition and the renaturation of proteins in gels on the Morse. 1986. Use of dot-immunobinding and immunofluores-
recognition of proteins on Western blot analysis by monoclonal cence assays to investigate clinically suspected cases of chan-
antibodies. Anal. Biochem. 57:144-153. croid. J. Infect. Dis. 153:879-887.
4. Karim, Q. N., G. Y. Finn, C. S. F. Easmon, Y. Dangor, D. A. B. 14. Schmid, G. P., L. L. Sanders, J. H. Blount, and E. R. Alexander.
Dance, Y. F. Ngeow, and R. C. Ballard. 1989. Rapid detection of 1987. Chancroid in the United States. Reestablishment of an old
Haemophilus ducreyi in clinical and experimental infections disease. JAMA 255:3265-3268.
using monoclonal antibody: a preliminary evaluation. Geni- 15. Slootmans, L. D., D. A. Vanden Berghe, and P. Piot. 1985.
tourin. Med. 65:361-365. Typing Haemophilus ducreyi by indirect immunofluorescence
5. Laemmli, U. K. 1970. Cleavage of structural proteins during the assay. Genitourin. Med. 255:123-126.
assembly of the head of bacteriophage T4. Nature (London) 16. Taylor, D. N., C. Duangmani, C. Suvongse, R. O'Connor, C.
227:680-685. Pitarangsi, K. Panikabutra, and P. Escheverria. 1984. The role
6. Morse, S. A. 1989. Chancroid and Haemophilus ducreyi. Clin. of Haemophilus ducreyi in penile ulcers in Bangkok, Thailand.
Microbiol. Rev. 2:137-157. Sex. Transm. Dis. 11:148-151.
7. Muller, F., and K.-H. Muller. 1988. Immunoglobulin M and G 17. Wagar, E. A., J. Schachter, P. Bavoil, and R. S. Stephens. 1991.
antibody response in rabbits after experimental Haemophilus Differential human serologic response to two 60,000 molecular
ducreyi infection. Zentralbl. Bakteriol. Microbiol. Hyg. Ser. A weight Chlamydia trachomatis antigens. J. Infect. Dis. 162:922-
268:238-244. 927.

Downloaded from https://journals.asm.org/journal/iai on 09 July 2024 by 196.249.96.9.