Feb 6, 1986 - Miller, T. B., S. D. Van Camp, and D. A. Barnum. 1983. The effects of intra-amniotic inoculation of Haemophilus somnus on the bovine fetus and ...
INFECTION
AND
Vol. 54, No. 2
IMMUNITY, Nov. 1986, p. 555-560
0019-9567/86/110555-06$02.00/0 Copyright X) 1986, American Society for Microbiology
Experimental Abortion and the Systemic Immune Response to "Haemophilus somnus" in Cattle PHILLIP R. WIDDERS,l* LARRY G. PAISLEY,2 RONALD P. GOGOLEWSKI,' JAMES F. EVERMANN,2'3 JOE W. SMITH,1 AND LYNETTE B. CORBEIL' Department of Veterinary Microbiology and Pathology,' Department of Veterinary Clinical Medicine and Surgery,2 and Washington Animal Disease Diagnostic Laboratory,3 Washington State University, Pullman, Washington 99164 Received 6 February 1986/Accepted 15 August 1986
"Haemophilus somnus" has been identified in the etiology of bovine abortion on the basis of the isolation of the organism from aborted fetal and placental tissues. To investigate the role of hematogenous dissemination of "H. somnus" in the pathogenesis of abortion and to monitor the humoral immune response to infection, 19 pregnant cows (gestation ages, 1.4 to 7 months) were challenged intravenously (11 cows) or intrabronchially (8 cows). Five cows challenged intravenously aborted, and one cow challenged intrabronchially resorbed her fetus. "H. somnus" was isolated in large numbers from aborted tissues, and placental lesions were similar to those reported in a field case of "H. somnus" abortion. Antibody titers in serum were measured by the microagglutination test (MAT) and by enzyme-linked immunosorbent assay (ELISA). A response to challenge was measured by MAT; it was also measured by ELISA within the immunoglobulin Gl (IgGl), IgG2, and IgM isotypes. On comparison of pre- and postchallenge antibody titers, the greatest and most persistent response was detected within the IgG2 isotype. Prechallenge antibody titers (measured by MAT and by IgG2 ELISA) were lower in animals that aborted than in normal calving animals, indicating that IgG2 antibody may have a role in limiting hematogenous dissemination of "H. somnus." "Haemophilus somnus" causes a variety of clinical syndromes in cattle, including thromboembolic meningoencephalitis, arthritis, pneumonia, and reproductive failure. Both pneumonia (R. P. Gogolewski, L. B. Corbeil, C. W. Leathers, and H. D. Liggitt, submitted for publication) and thromboembolic meningoencephalitis (6) have been well characterized and reproduced experimentally, but reproductivedisease syndromes have been less well defined. "H. somnus" is incriminated as a cause of abortion on the basis of isolation in pure culture from aborted fetuses (15) and by reproduction of the disease by intra-amniotic challenge with "H. somnus" (8). However, challenge by this route cannot measure the capacity of the organism to cross the placenta and avoids the potentially protective effect of a systemic or local immune response. The ability of "H. somnus" to survive systemically and cause lesions in the central nervous system and joints (6) suggests that hematogenous dissemination may be responsible for the sporadic abortions caused by "H. somnus" in the field. Panciera and co-workers (10) observed that respiratory disease may precede outbreaks of thromboembolic meningoencephalitis, suggesting that the respiratory tract is a principal site for systemic invasion by "H. somnus." Conversely, the significance of an ascending route of infection in the pathogenesis of "H. somnus"induced abortion is equivocal, since most normal bulls (4, 5) and 10 to 28% of normal cows (12, 16) carry the organism on the genital mucosae. On this basis, animals in this experiment were challenged by intrabronchial (i.b.) or intravenous (i.v.) inoculation with "H. somnus." The bovine immune response after systemic or mucosal infections with "H. somnus" is not well defined. The aim of the present study was to investigate the role of systemic infection with "H. somnus" in the etiology of bovine abortion and to monitor the systemic isotypic immune responses before and after "H. somnus" infection in cattle. *
MATERIALS AND METHODS
Bacteria. The isolate of "H. somnus" (649-4) used in this study was recovered from the stomach contents of an aborted bovine fetus. The organism was stored in phosphatebuffered glycerol at -70°C, and from initial isolation, it was passaged a maximum of four times on Columbia blood agar plates. Challenge inocula were prepared by two passages in yeast extract (10 g/liter)-RPMI (11 g/liter) culture medium supplemented with thiamine monophosphate (10 ,ug/ml) and agitated in air at 37°C. The number of viable bacteria in the challenge inocula was estimated spectrophotometrically and confirmed by plate count immediately before and after challenge. Cattle. A total of 19 pregnant cows and heifers were used in the study. Fetal ages were estimated by rectal palpation and were confirmed by calving dates for most animals. Animals were assigned to a treatment group by gestation age in an attempt to obtain a uniform fetal age distribution for both treatment groups. For the course of the experiment, the cattle were confined on straw in partly covered runs. They were observed for signs of abortion daily for the first 3 weeks after challenge and at regular intervals thereafter. Animals were challenged by i.v. or i.b. inoculation. The jugular vein was used for i.v. challenge, and i.b.-challenged animals were intubated by a nasotracheal tube inserted into the right diaphragmatic lung lobe (Gogolewski et al., submitted). Experimental protocol. Cows P2 (6.5 months gestation) and P3 (2.5 months gestation) were initially challenged in a preliminary study by i.v. and i.b. inoculation, respectively, of 20 ml of broth containing 4 x 109 "H. somnus." No abortion resulted. These animals were then rechallenged 6 weeks later in the larger study when the 19 animals each received 4 x 1010 "H. somnus" in 10 ml of broth. Immediately before challenge and at 1, 2, 5, and 30 weeks after challenge, serum was collected as were nasal and
Corresponding author. 555
556
WIDDERS ET AL.
vaginal swabs for culture. Serum samples were frozen in aliquots at -70°C for subsequent antibody assay. Aborted fetuses and placentae were examined for gross lesions, and cultures were taken from the placenta, fetal stomach contents, lung, heart, brain, and eyelid. In addition, swabs from maternal uterine and vaginal secretions were cultured for "H. somnus." At abortion, blood from cows 4 and 5 was cultured in C02-enriched biphasic Trypticase soy medium (BBL Microbiology Systems, Cockeysville, Md.). Sections were cut from the fetal placenta (when available), lung, and liver, and stained with hematoxylin and eosin. To eliminate other infectious agents in the etiology of these abortions, fetal fluids were assessed for antibody to infectious bovine rhinotracheitis virus, bovine virus diarrhea, parainfluenza 3 virus, Brucella abortus, Leptospira spp., and Campylobacter spp. Virus isolation was attempted from all aborted fetal tissues. Cows 8 and 13, which were clinically unaffected by challenge, were aborted by intramuscular administration of prostaglandin (500 pug of cloprostenol) 17 days after challenge. Antibody assays. Agglutinating antibody was measured by the microagglutination test (MAT), and isotypic antibody was measured by enzyme-linked immunosorbent assay (ELISA) for the immunoglobulin G (IgG) subclasses IgGl and IgG2 and for IgM. Isolates 8025 and L1034 were used as the antigen in the MAT, since Washington Animal Disease Diagnostic Laboratory uses a mixture of these two isolates as the antigen in this diagnostic test. Triphenyltetrazolium chloride (80 mg) in H20 was added to 200 ml of an overnight culture of "H. somnus" to provide color. After incubation for an additional 20 min, the stained bacterial suspensions were washed and resuspended in 0.1% formaldehyde in sterile buffered saline. This antigen preparation was stable at 4°C for up to 6 months. Antigen was adjusted to 11% transmission at 440 nm, and 0.05 ml was added to serial dilutions of heatinactivated test sera (0.05 ml) in the wells of a round-bottom microtiter plate. The plates were incubated at 37°C for 38 h, and the titer was calculated as the last dilution showing agglutination of the antigen. Antigen for ELISA was prepared by washing in 1% formalin-saline a log-phase culture of the challenge isolate 649-4 grown in yeast extract-RPMI broth. Bacterial concentration was adjusted in phosphate-buffered saline (pH 7.2) with 0.02 M MgCl2 to 75% transmission at 610 nm in a
spectrophotometer (Perkin-Elmer Corp., Norwalk, Conn.). Wells of microtiter plates (Costar, Cambridge, Mass.) were coated by incubating the antigen preparation (0.10 ml per well) for 16 h at 4°C. Doubling dilutions of test sera in PBS containing 0.05% Tween 20, 0.2% ovalbumin, and 2% polyvinylpyrrolidone were incubated in the wells for 1 h at 37°C; this was followed by bovine isotype-specific (IgGl, IgG2, and IgM) monoclonal antibodies (IgGl and IgG2 were provided by A. Guidry, U.S. Department of Agriculture, Beltsville, Md.; IgM was from W. Davis, Washington State University, Pullman, Wash.) and rabbit anti-mouse immunoglobulin antiserum conjugated with peroxidase (Zymed Labs, San Francisco, Calif.). Reactions were developed with 0.005% hydrogen peroxide in 0.1% 5-aminosalicylic acid, and the plates were read in the dual-wavelength mode of an ELISA reader (Dynatech Laboratories, Inc., Alexandria,
Va.). Endpoints for each serum sample were derived by fitting regression lines of optical density against serum dilution and from these lines determining the reciprocal of the serum
INFECT. IMMUN.
dilution at an optical density of 0.35 (the approximate midpoint of the standard curve). To control for variation between plates and assays, a high-titer positive control serum (HT serum) was included on each plate. This positive serum was produced by immunizing a convalescent cow with three subcutaneous injections of formalinized "H. somnus" 649-4 in Freund incomplete adjuvant. The mean HT serum endpoint was determined for each isotype from all plates assayed by ELISA. Sample titers were then calculated by the following formula: sample titer = mean HT serum endpoint x [(sample serum endpoint, plate X)/(HT serum endpoint, plate X)]. Significant differences in antibody titers between cows that aborted and normal calving cows were assessed using Wilcoxon's rank test. Bacterial antigens for serology. To assess the validity of comparing antibody titers measured by MAT and by ELISA when different isolates of "H. somnus" were used as the antigen, the reactivity of the HT serum for the ELISA antigen (649-4) or MAT antigen (L1034-8025) was compared by immunoblotting (14). Bacterial samples (approximately 5 x 108 CFU/ml) were boiled for 3 min in sample buffer containing 5% ,-mercaptoethanol. Discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis in 1.4mm-thick slab gels used stacking and gradient gels of 5% and 7.5 to 15% polyacrylamide, respectively. Electrophoresis was performed at a constant current of 30 mA for 4 h. After electrophoresis, the gel was equilibrated in transfer buffer (25 mM Tris and 192 mM glycine [pH 8.3] with 20% methanol) for 10 min, and the separated bacterial antigens were transferred to nitrocellulose at 70 V for 3 h (14). The blots were then incubated overnight at 4°C with HT serum diluted 1/4,000 in 0.02 M Tris-buffered saline (pH 7.5) containing 0.05% Tween 20. After two 10-min washes in Tris-buffered saline-Tween 20, bound antibody was detected by incubation with goat antiserum to bovine IgG (heavy and light chain) conjugated with peroxidase (Kirkegaard and Perry Laboratories, Gaithersburg, Md.) followed by substrate (0.05% 4-chloro-1-naphthol, 5% methanol, 0.015% H202)RESULTS Clinical response. The outcomes of the 19 pregnancies are recorded in Tables 1 and 2. In all, 10 animals calved normally, 2 cows unaffected by challenge were aborted by prostaglandin administration 17 days after challenge, 5 cows challenged i.v. and 1 cow challenged i.b. aborted or resorbed fetuses. The gestational ages at challenge of the five aborted fetuses among the cows inoculated i.v. ranged from 3.5 to 6 months (Table 1). The cow inoculated i.b. that resorbed its fetus (cow 1) was 1.5 months pregnant at challenge (Table 2). Bacteriology. Culture of nasal and vaginal swabs prechallenge and at 1, 2, 5, and 30 weeks postchallenge resulted in "H. somnus" isolation in low numbers from cows P2 and P4 prechallenge and in sporadic isolation from cows 1, 5, 16, 21, and 17 postchallenge (Table 3). In addition, the mucosae of cows 1, 5, 17, and 21 were swabbed at 24 or 27 weeks postchallenge. At this time, endometritis was diagnosed by rectal palpation of cow 17, and "H. somnus" was isolated from uterine and vaginal swabs in heavy, pure culture at 27 and 30 weeks from cow 17. This was the only animal for which culture-positive "H. somnus" endometrial disease was detected, although persistence of the organism in the uterus of cow 21 was suggested by culture of "H. somnus" at the second abortion 5.5 months after challenge. After abortion, "H. somnus" was isolated from the genital tract or from fetal tissues in high numbers for i.v.-challenged
BOVINE ABORTION AND IMMUNE RESPONSE TO "H. SOMNUS"
VOL. 54, 1986
TABLE 3. Isolation of "H. somnus" from bovine mucosae
TABLE 1. Abortions after i.v. challenge with "H. somnus" "H. somnus" isolation after
Gestational Abortion (days Cw age (mo) after challenge)
16
2-2.5
13 8
3 3
5 R4 17 4 21 P2(1) P1 2 P2(2)
3.5-4 4 4 4-4.5 4-4.5 4.5 4.5 5.5-6 6
+ (21)b + (21 + (3)c
pregnancy"
-
-
-
-
-
-
_
4+
-
_ +
4+
+ (5)c + (3) +
e
OG 1+ ND
4+ ND
4+d 4+ 3+e
ND ND
+ (5)
-
-
ND
4+
cows 4, 5, and P2 (Table 1). In addition, cow 5 was septicemic with "H. somnus" at abortion. The culture of fetal and maternal tissues from cow 17 was complicated by a protracted abortion (>18 h from the onset of cervical dilation), so that heavy overgrowth of all tissues by Escherichia coli and Corynebacterium pyogenes compromised the identification of "H. somnus." However, 27 and 30 weeks after challenge (Table 3), "H. somnus" was isolated in heavy, pure culture from uterine and vaginal swabs of this cow. Since abortion was not observed in cow 21, no cultures were done at the time of the first abortion. However, "H. somnus" was recovered from the subsequent fetal resorption 5.5 months after challenge. "H. somnus" was not isolated from the reproductive tract of cow 1 after fetal resorption because the uterus of this cow could not be swabbed until the onset of cervical dilation some 8 to 15 days after fetal death (18 days after challenge). Cow 4 died 4 days postinoculation, shortly after abortion, because of a ruptured uterine wall after surgical replacement TABLE 2. Abortions after i.b. challenge with "H. somnus" "H. somnus" isolation after
age (mo)
Abortion after (days challenge)
abortiona Fetus Placenta Genital mucosa
1 14
Pe(1°) 15 12 3 Pe(2°) P4 10
1.5 2to2.5 2.4 3 3 3 to 3.5 4 4 6.5to7
+ (3 to lO)b -
ND
ND
d
Materal blood
ND
-CND -
-
-
ND, Not done. Fetus resorbed. c Weak calf died shortly after birth. Since the culture was negative and the birth occurred 6 months after challenge, this was not considered an "H. somnus"-induced abortion. d -, No growth. a
b
A C
5
A
10 16 17 21 P2 P4
C C
Isolation of "H. somnus" at wkb: 5 27 1 2
N
ND
V
N
A A A
N
C
V
V V
-
N -
30
ND ND V _CC ND ND
V
V
A, Fetus aborted or resorbed; C, calved normally. Only cows positive for "H. somnus" are shown. N, Positive on nasal swab; V, positive on vaginal swab; ND, not determined; -, no "H. somnus" isolation. c Cow 21 was positive for "H. somnus" when fetal resorption was detected at 24 weeks. b
Bacterial isolation rate: -, no growth; 1 +,
