Effect of Anticapsular Antibodies on Neutrophi1 Phagocytosis of Staphylococcus 8ureus1 A. J. GUIDRY, s. P. OLIVER,~K. E. SQUIGGINS, E. F. ERBE H. H. DOWLENP C. N. HAMBLETON, and L. Y. BERNINGd
Milk Secretion and Mastitis Laboratory Agrkullural Research Service Beltsvllle, MD 20705
munized with compact. However, anticompact antibodies were opsonic for diffuse large clearing. These data show that bovine antibodies to S. aureus capsule are opsonic for bovine neutrophils and that capsule plays a role in inhibition of cell-wall opsonization of S. aurew. (Key words: capsule, phagocytosis, antibodies)
ABSTRACT
One of the major virulence factors of Staphylococcus aureus is development of an exopolysaccharide capsule in vivo, which inhibits recognition of antibodies to highly antigenic cell wall by neutrophils. To circumvent this inhibition, an attempt was made to produce anticapsular antibodies. Three cows per group were immunized in midactation by injections in the area of the supramammary lymph node and intramuscularly and were boosted on d 14, 42, and 70 with three variants of Smith S. aureus: compact, unencapsulated; diffuse, rigid c a p sule; and diffise large clearing, excep tionally large flaccid capsule using dextran sulfate as adjuvant. Serum agglutination and ELISA titers of cows immunized with diffuse and diffuse large clearing increased after immunization and after each boost and remained elevated to the end of the experiment at 112 d. Phagocytosis of diffuse and diffuse large clearing, m e a d by flow cytometry, was enhanced by immunization with either organism. No antibody response to capsule or enhanced phagocytosis of diffuse developed in cows im-
Abbreviation key: Cp = Smith compact Staphylococcus aureus, M = Smith diffuse SraphyZ~occusaureus, Dflc = Smith diffuse large clearing SraphyZococcus aureus, FS = forward scatter, PBS = phosphatebuffered saline, PMN = polymorphonuclear neutrophil, SS = side scatter. INTRODUCTION
Staphylococcus aureus is a major cause of mastitis in cows (4, 21, 22). Polymorphonuclear neutrophil (PMN) phagocytosis has been shown to be an important defense against S. aurew infection of the mammary gland (11). However, because milk PMN are less phagocytic because of ingestion of milk fat globules and because of a deficiency in phagocytosis promoting antibodies (opsonins), high concentrations of milk PMN (9 x 1 6 P M N / ~ )are required to prevent infection (17). Because this concentration exceeds the number of PMN in the normal healthy gland, increasing phagocyReceived April 11, 1991. tic efficiency of PMN via opsonins is the more Accepted h3ay 29. 1991. ‘Mention of a trade name, proprietary product, or desirable means of enhancing mammary gland specitic equipment does not constitate a guarantee 01 resistance to S. aureus infection. warranty by the US Department of Agriculture and does A major obstacle to producing a protective not imply its approval to the wrclosion of otber products immune response to S. aureus is development that may be suitable. 2Department of ~nimalsciaces. university of en- of extracellular polysaccharide (exopolysacnessec. Knoxville 37901-1071. charide), in vivo, in the form of a capsule or 3Dairy Expcrimmt Station, Lewisburg, TN 37091. “slime” layer (3,9,23). The exopolysaccharide 4Presen! address: Department of Dairy Science, California polytechnic state university, san Luis obispa allows antibodies to cell wall and complement to penetrate, but masks recognition of antibody 93407. 1991 J Dairy Sci 74:3360-3369
3360
STAPIirLOCOCCUS AUREUS CAPSU'LE OPSONINS
by PMN (5, 12, 23) and can prevent activation of complement (C3) (2, 12). In addidon, polysaccharides are low in immunogenicity and are Tcell-independent; thus, they have no immunological memory (13). This results in a transient antibody response to immunization and in no synergistic response, which is typical for Tcelldependent antigens. Both of these factors make the production of opsonins to capsule difficult. However, when antibodies are produced against capsule, they can serve as effective opsonins for human PMN (6, 7). Instability of the capsule has been a major obstacle to studying effects of capsule on S. aureus virulence and on production of o p sonizing antibodies to encapsulated S. aureus. It has been showii that 94 to 100% of fresh isolates from cases of S. aureus mastitis are encapsulated but fail to express capsule when cultured in vitro (9, 14), thus complicating studies on interaction of the organism with what most mastitis researchers consider the primary intramammary defense mechanism, PMN. Also, the structural nature of the capsule (Le., size and density) has not been elucidated. Both capsular size and density could be important factors inhibiting opsonization. This study was conducted to determine whether cows could produce an immune response to S. aureus capsule that would enhance phagocytosis of S. aureus by PMN. MATERIALS AND METHODS
Siaphylococcus aureus Vaccine
To circumvent changes in expression of the exopolysaccharide of S. aureus in vitro, we selected three variants of S. aureus Smith that have stable capsule characteristics: compact (Cp) unencapsulated; diffuse (Df) rigid capsule (both obtained from M. A. Melly, Vanderbilt University School of Medicine, Nashville, TN); and diffuse !xge clearing (Dflc), a mutant of Df with an exceptionally large flaccid c a p sule (isolated in our laboratory). The three organisms were grown in vivo according to Watson and prideaux (20). Briefly, dialysis bags (molecular weight 12 to 14O00, Spectrum Medical Industries Inc., Los Angeles, CA) containing Oxoid nutrient broth (Oxoid, Columbia, MD) were inoculated and placed in the peritoneal cavities of adult ewes for 48 h.
3361
The S . aureus were killed by adding formalin, 1% final concentration, and incubating 48 h at 37'C. They were washed once in sterile phosphate-buffered saline (PBS) (.01 M phosphatebuffered 35% NaCl, pH 7.2) and stored at 4'C in sterile PBS with .l% NaN3. The vaccine consisted of 5 x 109 S. aureusld in 500 mg/ ml of dextran sulfate (molecular weight 500,000) (Sigma, St. Louis, MO) in PBS. Demonstratlon of Exopolysaccharlde vla India Ink
The S. aureus were dispersed evenly in a loop of India ink on a glass slide. A coverslip was placed over the mixture, and excess ink was blotted to obtain a nearly transparent brown film. Slides were examined with phase contrast microscopy to determine relative clearing around each organism. Demonstmtlon of Exopolysaccharide via Electron Mlcroscopy
The three variants of S. aureus were grown as surface colonies on trypticase soy agar (BBL Microbiology Systems, Cockeysville, MD). Organisms were killed by incubation in 3% formalin at room temperature for 24 h and washed in water four times by centrifugation at 10,OOO x g for 5 min. Organisms were resuspended, and a small volume (1 to 2 pl) was transferred to the specimen slot of a 24-karat gold complementary freeze-etch specimen holder (18). The holder was frozen immediately in Freon 22 at -155'C and mounted in a complementary specimen cap, which was transferred under liquid frwn to a modified Denton DFE-2 freeze-etch modul- @enton Vacuum, Inc., Cherry Hill, NJ). The chamber was evacuated to 2 x 10-7 torr. The sample was surrounded by a liquid nitrogen cooled metal shroud and fractured by opening the hinged holder. Temperature was raised quickly to -9SC for 2 min to sublime water from the sample and then cooled to -150°C for platinum deposition. platinum was deposited at a 45' angle by resistance evaporation. Deposition was controlled with a resistance monitor (19). Carbon was evaporated onto the platinum to support the thin platinum film. The sample holder was removed from the vacuum system, and the organisms were dissolved in chromicJournal of Dairy Science Vol. 74, No. 10, 1991
3362
GUIDRY ET AL.
sulfuric acid, rinsed four times in distilled water, and mounted on Fonnvar-coated grids. Grids were viewed and photographed using an H-500-H Hitachi transmission electron microscope (Hitachi Ltd., Tokyo, Jpn.).
fied with 1 ml of dextran sulfate in the superficial inguinal lymph node and in the area of the semitendinosus muscle. Cows were injected on d 0, 14, 42, and 70. Approximately 10 ml of blood were obtained biweekly for 4 mo.
Capsule Removal
Agglutlnation Assay
Diffuse and DflCwere mixed separately in a Waring blender (Dynamics Corporation of America, New Hartford, Cl") at 21,500 rpm for a total of 60 and 20 min. respectively, with cooling in ice every 3 min. Organisms were washed with PBS by centrifugation at 20,000 x g for 15 min at 4'C. Following treatment with the Waring blender, the Df suspension was treated with a Tekmar tissuemizer (Tekmar, Cincinnati, OH) at 22,000 rpm for 20 min. The Df suspension was further treated with a Vibrogen cell mill (Edmund Buhler, Tubingen 1, Germany) with .l-mm diameter glass beads. Bead to cell ratio was initially 3:l for 5 min, then 3:2 for 15 min, and then 1:l for 10 min. The Df suspension was decanted after glass beads were dlowed to settle. The S. uureus were then washed in PBS as described earlier.
Serum samples were diluted in .01 M PBS, pH 7.4, containing .05% Tween 20 (diluent). Diluted serum (50 pl) was placed in Linbro conical microtiter plate wells (Dynatech Laboratories Inc., Chantilly, VA). Serial dilutions of 1:l were made for each sample. The three formalin-killed S. aweus variants (25 pl, 1 x 109 organisms/ml) in diluent were added to microtiter wells and incubated overnight in a humidity chamber at room temperature. The highest dilution showing a positive reaction was recorded as positive.
Serum-Soft Agar
The three S.uureus variants were grown in trypticase soy broth (BBL Microbiology Systems) to a Nephelos reading of 100% and diluted l/500,000 in PBS to contain approximately 10 to 20 cfu/lOO pI. One hundred microliters of the suspension were added to 2 ml of Staphylococcus Medium 110, pH 6.3 (Difco Laboratories, Detroit, MI), containing .15% agar and 4% normal rabbit serum at 56°C. in 4-ml sterile polystyrene Falcon tubes (Becton Dickerson, Oxnard, CA) and vortexed. After cooling to room temperature, 1 ml of 2% agar cap in Staphylococcus Medium 110 (Difco), pH 6.3 (Difco) was layered on top. Tubes were incubated at 37'C for 12 to 16 h. Colony growth was scored as compact or diffuse. lmmunlzatlon and Sampllng of cows
Jersey cows, three each, in midlactation were injected with 1 ml of killed Cp, Df, and Dfic containing 5 x 109 organisms/& emulsiJournal of Dairy Science Vol. 74. No. 10, 1991
Absorptlon of Sera
Prior to absorption, sezum samples were diluted M2.5 or geater in .01 M PBS, pH 7.4, mixed with an equal volume of protein A (200 pg/ml in .01 M PBS, pH 7.4), and incubated at room temperature for 18 h to block Ig binding to S. aureus protein A. One hundred microliters of formalin-killed Cp (1 x 10'0 organisms/&) were added for every 3 ml of sera-protein A solution and rotated 2 h at 4'C. The S.aureus were removed by centrifugation, and absorption was repeated until ELISA readings to Cp were at background levels. ELlSA for Specific Antlbody
For quantifying total specific antibody, formdin-killed S. aureus (2 x 1oS/m1 in .01 M PBS, .74% NaCl, pH 6.8) were added to Immulon 2 (Dynatech Laboratories Inc.) microtiter wells and incubated overnight at room temperature in 1 0 % humidity. Wells were washed twice with 3 5 % NaCl containing .05% Tween 20 using a 12channel pipetter and gentle flicking. Serum samples previously incubated with protein A and diluted in diluent were added and incubated 2 h. M e r washing three times, goat anti-bovine IgG (€ + L) I alkaline phosphatase (Kirkegaard and Perry Inc., Gaithersburg, MD) was added and incubated 2 h. Substrate @-nitrophenyl phos-
STAPHyulCOCCUS AUREUS CAPSULE OPSONINS
3363
phate disodium; Sigma) in .05 M carbonatecarbonate buffer (.001 M MgC12, pH 9.8) was added. Plates were read at 30, 60, and 90 min on a Titertek Multiskan (Flow Labs, McClean, VA). Because alkaline phosphatase conjugation blocks protein A binding, alkaline phosphatase antisera were not incubated with purified protein A. Phagocytosis
Phagocytosis was determined using flow cytometry according to Saad and Hageltom (15). Fluorescein-labeled Cp, Df, and Dflc were opsonized by mixing with protein Ablocked serum samples and rocked for 1 h at mom temperature. Isolated PMN were mixed with the opsonized S. aureus 1:20 and rocked gently for 30 min at 37'C. phagocytosis was stopped by adding icecold .09% NaCl with .04% EDTA. Extracellular fluorescence was quenched with 1% methylene blue. Samples were analyzed on an EPICS Profile flow cytometer (Coulter Electronics, Hialeah, FL). The PMN and S. aweus were separated using forward angle light scatter (FS) and log side scatter (SS). The PMN-associated fluorescence was quantified after gating PMN using FS and log SS. Staphylococcus aureus and PMN were counted using FS versus SS. Fluorescent intensity per PMN after quenching with methylene blue was used as a measure of S. uureus phagocytosed per PMN. Percentage PMN phagocytosing was calculated by dividing the number of PMN fluorescing by the total PMN count. statistics
Because agglutination data were logarithmic, homogeneity of residual variance was tested on nontransformed or log-transformed data by the Levene test (8). If residual variance was homogeneous (P > .01) for treatment x sampling period, least significant difference test (P < .05) was used to test differences in mean titers between 1) sampling periods within a treatment p u p and 2) treatment groups at each sampling period. One pooled standard error was reported for each treatment Figure 1. Transmission electron micrographs of freeze group. etched preparations of Staphylococcus oweus Smith comIf residual variance was heterogeneous (P < pact (A), diffuse (B), and diffuse large clearing Q. Bar .01) for treatment x period, the Levene test repnsents 1.0 pm. Journal of Dairy Science Vol. 74, No. 10, 1991
3364
GUKDRY ET AL.
Figure 2. Transmission electron micrographs of fmeeetched prepmatiom of St~phyiocomcsoureus Smith diffuse after treatment with Waring blender, time homogenizer, and Vihuga~cell m?l, and S. aweus Smith dithLPe large clearing (B) aAa treatment with Waring blmda only. Bar rqmscnts 1.0 p. (A)
was used to test homogeneity of sampljng periods within treatment group. If data were homogeneous, mean titers for each sampling period were compared within treatment group using least significant difference (P < .05). Pooled standard e m r s we= reported over all sampling periods for each treatment group. If period residual variances for treatment groups were heterogeneous (P e .Ol), Student’s f test (P < .05) was used to compare mean titer of sampling periods within treatment group. In this case, standard error was reported at each period within treatment p u p . RESULTS
Forty-one cows were screened for this study using ELISA blood titers to Cp and Df to obtain nine cows with low titers to Cp, Df, and Dflc. Samples with ELISA titers of twice diluent background were considered positive. Figure 1 shows electron micrographs of S. J o d of Dairy Scicnce Vol. 74, No. 10. 1991
aureus Smith Cp, Df, and Dflc. Compact (Cp) (A) has a smooth surface due to a total absence of exopolysaccharide (capsule) coating, whereas Df (B) and Dfk (C)show very distinct exopoysaccharide fibers on their surfaces. Use of a Waring blender alone was effective in removing a major portion of the exopolysaccharide capsule from Dflc (Figure 2B), whereas the Waring blender plus cell mill plus tissue homogenizer had little effect on the exopolysaccharide coating of JX (Figure 2A). Figun: 3 shows serum-soft agar preparations of the three variants. Compact (A) formed dense colcmies typical of nonencapsulated S. aureus. Diffuse (C)was more dispersed and elongated, which is typical of encapsulated S. aurem. Diffuse large clearing (B) was intermediate between Cp and Df, exhibiting extensive dispersion with limited elongation. India ink preparations in Figure 4 are typical for each variant. The slight half-moon clearing to one side of Cp (A) is artifact. The
STAPHYWCOCCUS AUREUS CAPSULE OPSONINS
figure 3. Srcrphylococclts r m r t u ~Smith compact (A). W
3365
e large clearing @), and diffuse (C)in senun-soft agar at
PH 7.0, 12-h h ~ ~ ~ b d ~ n
distinct clearing around Df (El) and the excep Titers decreased to near prebleed by d 42 but tionally large clearing around DflC (C) are rela- increased following the second boost. Cows tive in size to the exopolysaccharide strands immunized with either Df or Dfic showed a shown in Figure 1. Diffuse large clearing c a p similar increase in agglutination titers for Df sule had a mucoid, flaccid appearance. 'Ihe and Dfk to d 28 with a decrease at d 42. large clearing around Dfk was first observed in Second and third boosts resulted in an increase an India ink preparation of a mixed c u l m of in titers, which was sustained to the end of the m d y at 112 d. Df and ML, and led to its isolation. Serum ELISA titers of Df-immunized cows Immunization with S. aureus Smith Cp,Df, or D~L, significantly (P< .05) increased agglu- to Cp, Df, and Dflc after absorption with Cp to tination titers to Cp by d 14 (Figure 5; standard remove cell-wall antibodies are shown in Figerrors are in Table 1). The standard error for ure 6. After removal of antibodies to cell wall, Df antisera to Cp was not homogeneous; there- antibodies to capsule were clearly evident from fore, standard errors for each period are listed. d 14 to the end of the experiment. Also, as in J o d of Dairy Science Vol. 74, No. 10, 1991
3366
GUIDRY ET AL.
Figure 4. Light micrographsof M i inlc pnperations of Stophyfococw aweus Smith compact (A), diffuse (B), and diffuse huge clearing (C).Bar represents 50 pm.
the agglutination data, antibodies to Df and Dflc were shown to be crossreactive. Phagocytosis of Cp, Df, and DflC in the presence of prebleed and immune sera is shown in Table 2. Percentage PMN phagocytosing indicates percentage PMN containing one or more s. aureus. Intensity per PMN is related directly to the number of S. aureus per PMN. Neonatal bovine sera were used as the negative opsonin-void control. When incubated with prebleed sera, percentage PMN phagocytosing was 49% (Cp), 37% (JX& and 14% @f). Intensity per PMN was 17% (Cp). 9% (Dfi,), and 3% 0. Immunization with Cp showed no increase over prebleed in percentage PMN phagocytosing of either Cp, Df, or DflC and only a slight increase in intensity per PMN for DflC. Absorption with Cp removed essentially all the opsonizing effect of prebleed and Cp immune sera. Similarly, Df and Dflc immune sera showed no increase in opsonizing effect on Cp compared Joumal of Dairy Science Vol. 74, No. 10, 1991
with prebleed. Percentage PMN phagocytosing and intensity per PMN of Df were increased significantly over prebleed and Cp immune sera by either Df or DflC immune sera both before and after absorption with Cp. Diffuse and DflC immune sera also resulted in an increase in percentage PMN phagocytosing and intensity per PMN of Dfi,, but to a lesser degree, and was evident only after absorption with Cp. As in the agglutination and ELJSA data, crossreactivity was observed between Df and Dfic immune sera. DISCUSSION
These data show that the bovine can effect
an immune response to S. ufireus capsule that is opsonic for PMN. The sustained antibody response to capsule and the capsular response to boosts suggest a T-cell-dependent response (13). Earlier in vitro studies in our laboratory demonstrated a T-celldependent response to
3367
STAPHYLOCOCCUS AUREUS CAPSULE OPSONINS 600
I
I
j
1 .o
:. . ........
DlWUS0
250
-
1.0
0
.a
3
6-
v)
D
g
e 150
91
100
,:;:
..:.;~...,,,
,,:.
a
. . ., . . . . . . . . . .......................... ................ .-._,; ....
. . . .
50
0
-""
I Diffuselarge clearing I
, , ,,, ... I
.....
.,
cow Y2
. '. . -.:;' ...........,.;, ~, ...... ..................... ..........,,_.,: I
............. . . ,. ... ,. . .. . .., ..
,
I
I
I
.
:;.
.
'"i
3
I
1
.
.
:;4
............. .
.
$ 100
2
50
0
I
B
cow Y3
,/
..,.......... .:.
::,.__ . . . * . e . .
_-.-.-.'
0 ' . 0
I
,
,
0'.
,. .. I
150
:
"
-
4-
-
I
200
I
. .
I
300
E
1
I
I
./
.2
':'0
cow x1
.._ .. ..... _ -..._ ..,:........~..........::. ............,.:. .............-..-::. ,;.'" ... ..,
. 14 E
........_i._
/
.
28
.
42
.
56
.
70
.
84
I
98
112
B
B
Day
Figure 6. The ELISA titers of sera from three cows Figure 5 . Agglutination titers of sera from cows immunized with Staphylococcus uweus Smith compact (Cp) immunized with Staphylococcus oureus Smith diffuse to compact (Cp)-, Df ---, and diffuse large -, diffuse @f)----, and diffuse large C1g (Dfk) .... clearing (Dflc) .... after removal of cell-wall antibodies to Cp, Df. and Dfic. I = inrmunizsd; B = boosted. via absorption with Cp.
m,
isolated Dflc capsule using bovine T lymphocytes (10). Other workers have shown that immune response to exopolysaccharides can be TABLE 1. Slandard errors for data in Figure 5. increased dramatically by conjugating the exo- EUSA Antisera Pooled Period2 SE (x 10-2) SE' (d) polysaccharide to protein carrier molecules (1). Antigen to CP 1.32 0 6 purified capsule similar to that used in the in Cp 122 Df 14 vitro immunization study was later shown to 0 DflC 1.75 28 have trace amounts of nucleic and teichoic 23 CP 1.27 42 acids, which may have acted as a conjugate Df Df 1.31 56 46 vaccine. Immunizing with whole organism in 0 DflC 1.39 70 the current study could have effected a similar Dflc 244 CP 1.56 84 response, with the cell wall acting as the pro122 Df 1.31 98 Dflo 1.34 112 69 tein conjugate (13). The relatively high percentage PMN 'Pooled standard errors were used where SE were phagocytosing and intensity per PMN of Cp homogeneous. Means were converted to nontransformed incubated in prebleed sera compared with neo- data in tbe graphs, and SE must be multiplied and divided. h SE for diffase antisera to compact Q). natal calf serum suggest that, under normal which was not homogeneous, are given for each period conditions, cows form opsonizing antibodies to and must be added and subtracted in the conventional cell-wall components. This is not surprising, mamer. Dflc = Diffuse large clearing. Journal of Dairy Science Vol. 74, No. 10, 1991
3368
GUIDRY ET AL. Smith compact. diffwe, and diffuse large clearing Won and after Paccntage PMN
Iotcaritypn~
Phasocytoslng Antisera
s.
to
Phagocytosed
Reblced (n = 9)
mvew
a
SE
49 14 37
2.4 .8 2.4
CP Df
46 13 38 41 32 38 40 31
Dfk
40
6.8 2.5 7.2 5.1 5.7 5.0 9.5 1.8 52
4 3 3 5 3 4 5 21 8 5 19 8
2 .2 2 .2 .3 .5 .4 5.5 3.4 .5 3.0 2.0
CP Df CP Df CP Df Dfk
Dfk
with c p
SE
Dfk
Df
Unabsorbed
x
Dfk
CP
Absorbed
UMbOftd
X
17 3 9 18 4 13 16
9 13 15 9 13
SE 1.0 2 1.0 1.4 .5 1.8 1.7 .3 .9 2.2 .2 .6
Absorbed with cp
x .6 .6 .5
.6 .7 5
.5 9.8 4.6 .6 8.7 5.3
SE .06
.04 .03 .03 .07 .09 .03 1.25 2.32 .12 1.05 1.45
'N = 3 except for prebleed, where each value is the average of all nine cows. PMN = Polymorphonuclear nentrupbik cp = Compact s. aurew, Df = diffuse s. uurnrr; Dfk = diffuse S. aureus. Neonatal calf ~enrmwm used as tbe negative controt P?dN p h a g and ~ in- ~ ~ P m cp, 7 and .6; Df, 4 and .5; and Dfk. 5 and .4.
considering the highly antigenic nature of bacterid cell-wall proteins and their ubiquitous occurrence among a wide variety of bacterial species (16). Failure of the increased antibody titer to Cp to increase phagocytosis of Cp indicates that percentage PMN phagocytosing and the amount ingested were at a maximum for the PMN. Similarity in response between Cp and Dflc reflects the inability of the large flaccid DfIc capsule to block PMN recognition of opsonizing antibody. Conversely. the smaller but more rigid capsule of Df did inhibit phagocytosis. These data suggest that ability of capsule to inhibit phagocytosis is not solely dependent on size of capsule, as suggested by Lee d al. (7) and Peterson et at (12), but is also dependent on the rigidity of the capsule structure. Results of this study show clearly that opsonins, elicited in the cow by either Df or Dflc, increase phagocytosis of either organism, indicating common capsular antigens. The inability of immune sera to enhance phagocytosis of Dflc at the same level as Df, evidenced by lower percentage PMN phagocytosing and lower intensity per PMN after absorptim of antibodies to cell wall, could be attributed to the flaccid nature of DfiC capsule. However, Journal of Dahy Science Vol. 74, No. 10, 1991
the level of both percentage PMN phagocytosing and intensity per PMN of Dfk prior to absorption suggests that antibodies to cell wall can be more effective opsonins than anticapsular antibodies for organisms with large but flaccid capsules. The mctural nature of c a p sule on S. aureus from cases of bovine mastitis has not been studied. These data indicate that this could be a critical factor in determining the antigen to which opsonins should be produced. ACKNOWLEDGMENTS
Authors express appreciation to Chris Pooley for developing and printing the photographs and to L. W.Douglass for the statistical analysis. REFERENCES 1 Bixler, G. S., and P. Snbramonia. 1989. Thc cellular basis of the irnmune response to conjugate vaccines. Page 18 in Conjugate vaccines. J. M.Cruse and R. E. Lewis, ed. Conbib. Mimbiol. Immunol. Vol. 10. S. Kargcr AG. Basel, Switzerland. 2Blobcl. H.,J. Brucklcr, D. Kitmow, and W.Schaeg. 1980. Antiphagacytic factors of Staphylmoccus aur a . a m p . Immrmol. Mimbiol. Infect. DE.3447. 3Cqmty. G.G.. and J. W.Costcrton. 1984. Immuaological examination of the glycocalyces of Sfaphylo-
5 T X P W C O C C U S AUREUS CAPSULE OPSONINS coccus uureus strains Wiley and Smith. Curr. Mibiol. 11:297. 4 Foster,T. J. 1986. A new genetic approach to defining the virulence determinants of Stcrphylococcus aweus strains that Cause bovine mastitis. Ir. Vet. J. 40:llO. 5 W, A. W. 1981. Factors influencing the outcome of Escherichh coli mastitis in the dairy cow. Res. Vet. Sci. 31:107. 6 Karalrawa, W. W., A. SuttOq R. Schneerson. A. Kmpa^, and W. F. V-. 1988. Capantibodits induce type-specific phagocytosis of capsulated Staphylococcus mrreus :,y human polymorphormclear leukocytes. Infect. Immun. 561090. 7Lee, I. C.,M.J. Betley.C. A. Hopkins, N. E. Perez, and G. B. Pier. 1987. V i r u l m studies, in mice, of eansposon-inducedmutants of Stcrphylmoccus aureus differing in capsule size. J. Infect. Dis. 156741. 8 Milliken, G. A., and D. E. Johnson. 1984. ollbway treatment structure in a completely randomized design smcturc with heterogenous errors. Page 17 in Designing experiments. Vol. 1. Analysis of messy data. Van Nostrand Rhinehold Co., New York, NY. 9 Nonxoss, N. L., and J. P.Opdebeeck 1983. Encapsnlation of Staphylmoccrrs a w a s isolated from bovine milk. Vet. Microbiol. 8:397. 1OOjo-Amaizc. E. A., and A. J. Guidry. 1987. In vitro sensitization and stimulation of bovine lymphocytes with encapsulation or nonencapsulated Smith strain of Staphylococcus aureus. Am. J. Vet. Res. 48:1456. llPaape, M. J., W. P. Wergin,A. J. Guidry,and W.D. Schultzc. 1981. phagocytic defense of the ruminant mammary gland. Adv. Exp. Med. Biol. 137555. 12 Petason, P.K.,B. J. WilLinson, Y.Kim, D. Schmeling, and P.G. Quie. 1978. influence of encapsulation on staphylococcal opsonization and phagocytosis by haman polymorphonuclear kukocytes. Infect-Immun. 19943. 13 Poolman. I. T. 1990. Polysaccharides and membraoe vBccines. Page 57 in Bacterid vmines. A. Miziahi.
3369
ed. Contrib. Adv. Biol. Process. Wiley-Lis, Inc., New
Yo& NY. 14 Rather,P. N., A. P. Davis, and B. J. Wilkinson. 1986. S l i m production by bovine milk staphylococcal is+ lam. J. Clin. Mmbiol. 23:858. 15 Saad,A. M., and M.Hageltom.1985. mow cytometry charactahtion of bovine blood neutrophil phagocytosis of fluorescent bacteria and zymosan particles. Acta Vet. Scand. 26289. 16 Schleifer, K.H.1983. The cell envelope. Page 385 in Staphylococci and staphylococcal infections. Vol. 2. C.S.F. Easmon and C. Adlam, ed. Academic Press Inc.. London, Eogl. 1 7 S c h u l ~ W. , D., and M. J. Paape. 1984. Effect of outcom of mr-yi challenge exposure with Staphybcocclrr aureus of somatic cell concentration andpnsaceofan’ . ry device. Am. J. Vet. Res. 45:420. 18Steac, R L. 1981. Reparation of freezbfracture. freezactch, fraeze dry and from surface replica specimens for electron microscopy in the Denton DFE2 and DFE3 freeze-etch units. Page 131 in Current trends in morphological techuiques. Vol. II. J. E. Johmoo. Jr., ed. CRC Press, Boca Raton, FL. 19 Steac, R L.,E. F. Erbc. and J. M. Moseley. 1977. A resistance monitor with power cut-off for automatic regalation of shadow and support f b thickness in frazt-etching and related techniques. J. h4icrosc. (Oxf.)3:313. 20 Watson,D. L., and J. A. Prideam. 1979. Comparisons of Staphylococcus aureus grown in vitro and in vivo. Miaobiol. Immanol. 23543. 21 Watts,I. L. 1988. Etiological agents of bovine mastitis. Vet. Micmbiol. 1641. 22Watts, J. L,J. W. F’ankey, W. M. Oliver, S. C. Nickcrrron. and A. W. Lazams. 1986. Prevalence and effects of mastitis in beef cows. J. Anim. Sci. 6216. 23 W n . B. J. 1983. Staphylococcal capsules and slime. Page 481 in Staphyloccai aud staphylococcal iufections. Vol. 2. C.S.F. Easmon and C. Adlam, ed. Academic Ress Inc.. London. Engl.
J o d of Dairy Science Vol. 74. No. 10, 1991
Milk Secretion and Mastitis Laboratory Agrkullural Research Service Beltsvllle, MD 20705
munized with compact. However, anticompact antibodies were opsonic for diffuse large clearing. These data show that bovine antibodies to S. aureus capsule are opsonic for bovine neutrophils and that capsule plays a role in inhibition of cell-wall opsonization of S. aurew. (Key words: capsule, phagocytosis, antibodies)
ABSTRACT
One of the major virulence factors of Staphylococcus aureus is development of an exopolysaccharide capsule in vivo, which inhibits recognition of antibodies to highly antigenic cell wall by neutrophils. To circumvent this inhibition, an attempt was made to produce anticapsular antibodies. Three cows per group were immunized in midactation by injections in the area of the supramammary lymph node and intramuscularly and were boosted on d 14, 42, and 70 with three variants of Smith S. aureus: compact, unencapsulated; diffuse, rigid c a p sule; and diffise large clearing, excep tionally large flaccid capsule using dextran sulfate as adjuvant. Serum agglutination and ELISA titers of cows immunized with diffuse and diffuse large clearing increased after immunization and after each boost and remained elevated to the end of the experiment at 112 d. Phagocytosis of diffuse and diffuse large clearing, m e a d by flow cytometry, was enhanced by immunization with either organism. No antibody response to capsule or enhanced phagocytosis of diffuse developed in cows im-
Abbreviation key: Cp = Smith compact Staphylococcus aureus, M = Smith diffuse SraphyZ~occusaureus, Dflc = Smith diffuse large clearing SraphyZococcus aureus, FS = forward scatter, PBS = phosphatebuffered saline, PMN = polymorphonuclear neutrophil, SS = side scatter. INTRODUCTION
Staphylococcus aureus is a major cause of mastitis in cows (4, 21, 22). Polymorphonuclear neutrophil (PMN) phagocytosis has been shown to be an important defense against S. aurew infection of the mammary gland (11). However, because milk PMN are less phagocytic because of ingestion of milk fat globules and because of a deficiency in phagocytosis promoting antibodies (opsonins), high concentrations of milk PMN (9 x 1 6 P M N / ~ )are required to prevent infection (17). Because this concentration exceeds the number of PMN in the normal healthy gland, increasing phagocyReceived April 11, 1991. tic efficiency of PMN via opsonins is the more Accepted h3ay 29. 1991. ‘Mention of a trade name, proprietary product, or desirable means of enhancing mammary gland specitic equipment does not constitate a guarantee 01 resistance to S. aureus infection. warranty by the US Department of Agriculture and does A major obstacle to producing a protective not imply its approval to the wrclosion of otber products immune response to S. aureus is development that may be suitable. 2Department of ~nimalsciaces. university of en- of extracellular polysaccharide (exopolysacnessec. Knoxville 37901-1071. charide), in vivo, in the form of a capsule or 3Dairy Expcrimmt Station, Lewisburg, TN 37091. “slime” layer (3,9,23). The exopolysaccharide 4Presen! address: Department of Dairy Science, California polytechnic state university, san Luis obispa allows antibodies to cell wall and complement to penetrate, but masks recognition of antibody 93407. 1991 J Dairy Sci 74:3360-3369
3360
STAPIirLOCOCCUS AUREUS CAPSU'LE OPSONINS
by PMN (5, 12, 23) and can prevent activation of complement (C3) (2, 12). In addidon, polysaccharides are low in immunogenicity and are Tcell-independent; thus, they have no immunological memory (13). This results in a transient antibody response to immunization and in no synergistic response, which is typical for Tcelldependent antigens. Both of these factors make the production of opsonins to capsule difficult. However, when antibodies are produced against capsule, they can serve as effective opsonins for human PMN (6, 7). Instability of the capsule has been a major obstacle to studying effects of capsule on S. aureus virulence and on production of o p sonizing antibodies to encapsulated S. aureus. It has been showii that 94 to 100% of fresh isolates from cases of S. aureus mastitis are encapsulated but fail to express capsule when cultured in vitro (9, 14), thus complicating studies on interaction of the organism with what most mastitis researchers consider the primary intramammary defense mechanism, PMN. Also, the structural nature of the capsule (Le., size and density) has not been elucidated. Both capsular size and density could be important factors inhibiting opsonization. This study was conducted to determine whether cows could produce an immune response to S. aureus capsule that would enhance phagocytosis of S. aureus by PMN. MATERIALS AND METHODS
Siaphylococcus aureus Vaccine
To circumvent changes in expression of the exopolysaccharide of S. aureus in vitro, we selected three variants of S. aureus Smith that have stable capsule characteristics: compact (Cp) unencapsulated; diffuse (Df) rigid capsule (both obtained from M. A. Melly, Vanderbilt University School of Medicine, Nashville, TN); and diffuse !xge clearing (Dflc), a mutant of Df with an exceptionally large flaccid c a p sule (isolated in our laboratory). The three organisms were grown in vivo according to Watson and prideaux (20). Briefly, dialysis bags (molecular weight 12 to 14O00, Spectrum Medical Industries Inc., Los Angeles, CA) containing Oxoid nutrient broth (Oxoid, Columbia, MD) were inoculated and placed in the peritoneal cavities of adult ewes for 48 h.
3361
The S . aureus were killed by adding formalin, 1% final concentration, and incubating 48 h at 37'C. They were washed once in sterile phosphate-buffered saline (PBS) (.01 M phosphatebuffered 35% NaCl, pH 7.2) and stored at 4'C in sterile PBS with .l% NaN3. The vaccine consisted of 5 x 109 S. aureusld in 500 mg/ ml of dextran sulfate (molecular weight 500,000) (Sigma, St. Louis, MO) in PBS. Demonstratlon of Exopolysaccharlde vla India Ink
The S. aureus were dispersed evenly in a loop of India ink on a glass slide. A coverslip was placed over the mixture, and excess ink was blotted to obtain a nearly transparent brown film. Slides were examined with phase contrast microscopy to determine relative clearing around each organism. Demonstmtlon of Exopolysaccharide via Electron Mlcroscopy
The three variants of S. aureus were grown as surface colonies on trypticase soy agar (BBL Microbiology Systems, Cockeysville, MD). Organisms were killed by incubation in 3% formalin at room temperature for 24 h and washed in water four times by centrifugation at 10,OOO x g for 5 min. Organisms were resuspended, and a small volume (1 to 2 pl) was transferred to the specimen slot of a 24-karat gold complementary freeze-etch specimen holder (18). The holder was frozen immediately in Freon 22 at -155'C and mounted in a complementary specimen cap, which was transferred under liquid frwn to a modified Denton DFE-2 freeze-etch modul- @enton Vacuum, Inc., Cherry Hill, NJ). The chamber was evacuated to 2 x 10-7 torr. The sample was surrounded by a liquid nitrogen cooled metal shroud and fractured by opening the hinged holder. Temperature was raised quickly to -9SC for 2 min to sublime water from the sample and then cooled to -150°C for platinum deposition. platinum was deposited at a 45' angle by resistance evaporation. Deposition was controlled with a resistance monitor (19). Carbon was evaporated onto the platinum to support the thin platinum film. The sample holder was removed from the vacuum system, and the organisms were dissolved in chromicJournal of Dairy Science Vol. 74, No. 10, 1991
3362
GUIDRY ET AL.
sulfuric acid, rinsed four times in distilled water, and mounted on Fonnvar-coated grids. Grids were viewed and photographed using an H-500-H Hitachi transmission electron microscope (Hitachi Ltd., Tokyo, Jpn.).
fied with 1 ml of dextran sulfate in the superficial inguinal lymph node and in the area of the semitendinosus muscle. Cows were injected on d 0, 14, 42, and 70. Approximately 10 ml of blood were obtained biweekly for 4 mo.
Capsule Removal
Agglutlnation Assay
Diffuse and DflCwere mixed separately in a Waring blender (Dynamics Corporation of America, New Hartford, Cl") at 21,500 rpm for a total of 60 and 20 min. respectively, with cooling in ice every 3 min. Organisms were washed with PBS by centrifugation at 20,000 x g for 15 min at 4'C. Following treatment with the Waring blender, the Df suspension was treated with a Tekmar tissuemizer (Tekmar, Cincinnati, OH) at 22,000 rpm for 20 min. The Df suspension was further treated with a Vibrogen cell mill (Edmund Buhler, Tubingen 1, Germany) with .l-mm diameter glass beads. Bead to cell ratio was initially 3:l for 5 min, then 3:2 for 15 min, and then 1:l for 10 min. The Df suspension was decanted after glass beads were dlowed to settle. The S. uureus were then washed in PBS as described earlier.
Serum samples were diluted in .01 M PBS, pH 7.4, containing .05% Tween 20 (diluent). Diluted serum (50 pl) was placed in Linbro conical microtiter plate wells (Dynatech Laboratories Inc., Chantilly, VA). Serial dilutions of 1:l were made for each sample. The three formalin-killed S. aweus variants (25 pl, 1 x 109 organisms/ml) in diluent were added to microtiter wells and incubated overnight in a humidity chamber at room temperature. The highest dilution showing a positive reaction was recorded as positive.
Serum-Soft Agar
The three S.uureus variants were grown in trypticase soy broth (BBL Microbiology Systems) to a Nephelos reading of 100% and diluted l/500,000 in PBS to contain approximately 10 to 20 cfu/lOO pI. One hundred microliters of the suspension were added to 2 ml of Staphylococcus Medium 110, pH 6.3 (Difco Laboratories, Detroit, MI), containing .15% agar and 4% normal rabbit serum at 56°C. in 4-ml sterile polystyrene Falcon tubes (Becton Dickerson, Oxnard, CA) and vortexed. After cooling to room temperature, 1 ml of 2% agar cap in Staphylococcus Medium 110 (Difco), pH 6.3 (Difco) was layered on top. Tubes were incubated at 37'C for 12 to 16 h. Colony growth was scored as compact or diffuse. lmmunlzatlon and Sampllng of cows
Jersey cows, three each, in midlactation were injected with 1 ml of killed Cp, Df, and Dfic containing 5 x 109 organisms/& emulsiJournal of Dairy Science Vol. 74. No. 10, 1991
Absorptlon of Sera
Prior to absorption, sezum samples were diluted M2.5 or geater in .01 M PBS, pH 7.4, mixed with an equal volume of protein A (200 pg/ml in .01 M PBS, pH 7.4), and incubated at room temperature for 18 h to block Ig binding to S. aureus protein A. One hundred microliters of formalin-killed Cp (1 x 10'0 organisms/&) were added for every 3 ml of sera-protein A solution and rotated 2 h at 4'C. The S.aureus were removed by centrifugation, and absorption was repeated until ELISA readings to Cp were at background levels. ELlSA for Specific Antlbody
For quantifying total specific antibody, formdin-killed S. aureus (2 x 1oS/m1 in .01 M PBS, .74% NaCl, pH 6.8) were added to Immulon 2 (Dynatech Laboratories Inc.) microtiter wells and incubated overnight at room temperature in 1 0 % humidity. Wells were washed twice with 3 5 % NaCl containing .05% Tween 20 using a 12channel pipetter and gentle flicking. Serum samples previously incubated with protein A and diluted in diluent were added and incubated 2 h. M e r washing three times, goat anti-bovine IgG (€ + L) I alkaline phosphatase (Kirkegaard and Perry Inc., Gaithersburg, MD) was added and incubated 2 h. Substrate @-nitrophenyl phos-
STAPHyulCOCCUS AUREUS CAPSULE OPSONINS
3363
phate disodium; Sigma) in .05 M carbonatecarbonate buffer (.001 M MgC12, pH 9.8) was added. Plates were read at 30, 60, and 90 min on a Titertek Multiskan (Flow Labs, McClean, VA). Because alkaline phosphatase conjugation blocks protein A binding, alkaline phosphatase antisera were not incubated with purified protein A. Phagocytosis
Phagocytosis was determined using flow cytometry according to Saad and Hageltom (15). Fluorescein-labeled Cp, Df, and Dflc were opsonized by mixing with protein Ablocked serum samples and rocked for 1 h at mom temperature. Isolated PMN were mixed with the opsonized S. aureus 1:20 and rocked gently for 30 min at 37'C. phagocytosis was stopped by adding icecold .09% NaCl with .04% EDTA. Extracellular fluorescence was quenched with 1% methylene blue. Samples were analyzed on an EPICS Profile flow cytometer (Coulter Electronics, Hialeah, FL). The PMN and S. aweus were separated using forward angle light scatter (FS) and log side scatter (SS). The PMN-associated fluorescence was quantified after gating PMN using FS and log SS. Staphylococcus aureus and PMN were counted using FS versus SS. Fluorescent intensity per PMN after quenching with methylene blue was used as a measure of S. uureus phagocytosed per PMN. Percentage PMN phagocytosing was calculated by dividing the number of PMN fluorescing by the total PMN count. statistics
Because agglutination data were logarithmic, homogeneity of residual variance was tested on nontransformed or log-transformed data by the Levene test (8). If residual variance was homogeneous (P > .01) for treatment x sampling period, least significant difference test (P < .05) was used to test differences in mean titers between 1) sampling periods within a treatment p u p and 2) treatment groups at each sampling period. One pooled standard error was reported for each treatment Figure 1. Transmission electron micrographs of freeze group. etched preparations of Staphylococcus oweus Smith comIf residual variance was heterogeneous (P < pact (A), diffuse (B), and diffuse large clearing Q. Bar .01) for treatment x period, the Levene test repnsents 1.0 pm. Journal of Dairy Science Vol. 74, No. 10, 1991
3364
GUKDRY ET AL.
Figure 2. Transmission electron micrographs of fmeeetched prepmatiom of St~phyiocomcsoureus Smith diffuse after treatment with Waring blender, time homogenizer, and Vihuga~cell m?l, and S. aweus Smith dithLPe large clearing (B) aAa treatment with Waring blmda only. Bar rqmscnts 1.0 p. (A)
was used to test homogeneity of sampljng periods within treatment group. If data were homogeneous, mean titers for each sampling period were compared within treatment group using least significant difference (P < .05). Pooled standard e m r s we= reported over all sampling periods for each treatment group. If period residual variances for treatment groups were heterogeneous (P e .Ol), Student’s f test (P < .05) was used to compare mean titer of sampling periods within treatment group. In this case, standard error was reported at each period within treatment p u p . RESULTS
Forty-one cows were screened for this study using ELISA blood titers to Cp and Df to obtain nine cows with low titers to Cp, Df, and Dflc. Samples with ELISA titers of twice diluent background were considered positive. Figure 1 shows electron micrographs of S. J o d of Dairy Scicnce Vol. 74, No. 10. 1991
aureus Smith Cp, Df, and Dflc. Compact (Cp) (A) has a smooth surface due to a total absence of exopolysaccharide (capsule) coating, whereas Df (B) and Dfk (C)show very distinct exopoysaccharide fibers on their surfaces. Use of a Waring blender alone was effective in removing a major portion of the exopolysaccharide capsule from Dflc (Figure 2B), whereas the Waring blender plus cell mill plus tissue homogenizer had little effect on the exopolysaccharide coating of JX (Figure 2A). Figun: 3 shows serum-soft agar preparations of the three variants. Compact (A) formed dense colcmies typical of nonencapsulated S. aureus. Diffuse (C)was more dispersed and elongated, which is typical of encapsulated S. aurem. Diffuse large clearing (B) was intermediate between Cp and Df, exhibiting extensive dispersion with limited elongation. India ink preparations in Figure 4 are typical for each variant. The slight half-moon clearing to one side of Cp (A) is artifact. The
STAPHYWCOCCUS AUREUS CAPSULE OPSONINS
figure 3. Srcrphylococclts r m r t u ~Smith compact (A). W
3365
e large clearing @), and diffuse (C)in senun-soft agar at
PH 7.0, 12-h h ~ ~ ~ b d ~ n
distinct clearing around Df (El) and the excep Titers decreased to near prebleed by d 42 but tionally large clearing around DflC (C) are rela- increased following the second boost. Cows tive in size to the exopolysaccharide strands immunized with either Df or Dfic showed a shown in Figure 1. Diffuse large clearing c a p similar increase in agglutination titers for Df sule had a mucoid, flaccid appearance. 'Ihe and Dfk to d 28 with a decrease at d 42. large clearing around Dfk was first observed in Second and third boosts resulted in an increase an India ink preparation of a mixed c u l m of in titers, which was sustained to the end of the m d y at 112 d. Df and ML, and led to its isolation. Serum ELISA titers of Df-immunized cows Immunization with S. aureus Smith Cp,Df, or D~L, significantly (P< .05) increased agglu- to Cp, Df, and Dflc after absorption with Cp to tination titers to Cp by d 14 (Figure 5; standard remove cell-wall antibodies are shown in Figerrors are in Table 1). The standard error for ure 6. After removal of antibodies to cell wall, Df antisera to Cp was not homogeneous; there- antibodies to capsule were clearly evident from fore, standard errors for each period are listed. d 14 to the end of the experiment. Also, as in J o d of Dairy Science Vol. 74, No. 10, 1991
3366
GUIDRY ET AL.
Figure 4. Light micrographsof M i inlc pnperations of Stophyfococw aweus Smith compact (A), diffuse (B), and diffuse huge clearing (C).Bar represents 50 pm.
the agglutination data, antibodies to Df and Dflc were shown to be crossreactive. Phagocytosis of Cp, Df, and DflC in the presence of prebleed and immune sera is shown in Table 2. Percentage PMN phagocytosing indicates percentage PMN containing one or more s. aureus. Intensity per PMN is related directly to the number of S. aureus per PMN. Neonatal bovine sera were used as the negative opsonin-void control. When incubated with prebleed sera, percentage PMN phagocytosing was 49% (Cp), 37% (JX& and 14% @f). Intensity per PMN was 17% (Cp). 9% (Dfi,), and 3% 0. Immunization with Cp showed no increase over prebleed in percentage PMN phagocytosing of either Cp, Df, or DflC and only a slight increase in intensity per PMN for DflC. Absorption with Cp removed essentially all the opsonizing effect of prebleed and Cp immune sera. Similarly, Df and Dflc immune sera showed no increase in opsonizing effect on Cp compared Joumal of Dairy Science Vol. 74, No. 10, 1991
with prebleed. Percentage PMN phagocytosing and intensity per PMN of Df were increased significantly over prebleed and Cp immune sera by either Df or DflC immune sera both before and after absorption with Cp. Diffuse and DflC immune sera also resulted in an increase in percentage PMN phagocytosing and intensity per PMN of Dfi,, but to a lesser degree, and was evident only after absorption with Cp. As in the agglutination and ELJSA data, crossreactivity was observed between Df and Dfic immune sera. DISCUSSION
These data show that the bovine can effect
an immune response to S. ufireus capsule that is opsonic for PMN. The sustained antibody response to capsule and the capsular response to boosts suggest a T-cell-dependent response (13). Earlier in vitro studies in our laboratory demonstrated a T-celldependent response to
3367
STAPHYLOCOCCUS AUREUS CAPSULE OPSONINS 600
I
I
j
1 .o
:. . ........
DlWUS0
250
-
1.0
0
.a
3
6-
v)
D
g
e 150
91
100
,:;:
..:.;~...,,,
,,:.
a
. . ., . . . . . . . . . .......................... ................ .-._,; ....
. . . .
50
0
-""
I Diffuselarge clearing I
, , ,,, ... I
.....
.,
cow Y2
. '. . -.:;' ...........,.;, ~, ...... ..................... ..........,,_.,: I
............. . . ,. ... ,. . .. . .., ..
,
I
I
I
.
:;.
.
'"i
3
I
1
.
.
:;4
............. .
.
$ 100
2
50
0
I
B
cow Y3
,/
..,.......... .:.
::,.__ . . . * . e . .
_-.-.-.'
0 ' . 0
I
,
,
0'.
,. .. I
150
:
"
-
4-
-
I
200
I
. .
I
300
E
1
I
I
./
.2
':'0
cow x1
.._ .. ..... _ -..._ ..,:........~..........::. ............,.:. .............-..-::. ,;.'" ... ..,
. 14 E
........_i._
/
.
28
.
42
.
56
.
70
.
84
I
98
112
B
B
Day
Figure 6. The ELISA titers of sera from three cows Figure 5 . Agglutination titers of sera from cows immunized with Staphylococcus uweus Smith compact (Cp) immunized with Staphylococcus oureus Smith diffuse to compact (Cp)-, Df ---, and diffuse large -, diffuse @f)----, and diffuse large C1g (Dfk) .... clearing (Dflc) .... after removal of cell-wall antibodies to Cp, Df. and Dfic. I = inrmunizsd; B = boosted. via absorption with Cp.
m,
isolated Dflc capsule using bovine T lymphocytes (10). Other workers have shown that immune response to exopolysaccharides can be TABLE 1. Slandard errors for data in Figure 5. increased dramatically by conjugating the exo- EUSA Antisera Pooled Period2 SE (x 10-2) SE' (d) polysaccharide to protein carrier molecules (1). Antigen to CP 1.32 0 6 purified capsule similar to that used in the in Cp 122 Df 14 vitro immunization study was later shown to 0 DflC 1.75 28 have trace amounts of nucleic and teichoic 23 CP 1.27 42 acids, which may have acted as a conjugate Df Df 1.31 56 46 vaccine. Immunizing with whole organism in 0 DflC 1.39 70 the current study could have effected a similar Dflc 244 CP 1.56 84 response, with the cell wall acting as the pro122 Df 1.31 98 Dflo 1.34 112 69 tein conjugate (13). The relatively high percentage PMN 'Pooled standard errors were used where SE were phagocytosing and intensity per PMN of Cp homogeneous. Means were converted to nontransformed incubated in prebleed sera compared with neo- data in tbe graphs, and SE must be multiplied and divided. h SE for diffase antisera to compact Q). natal calf serum suggest that, under normal which was not homogeneous, are given for each period conditions, cows form opsonizing antibodies to and must be added and subtracted in the conventional cell-wall components. This is not surprising, mamer. Dflc = Diffuse large clearing. Journal of Dairy Science Vol. 74, No. 10, 1991
3368
GUIDRY ET AL. Smith compact. diffwe, and diffuse large clearing Won and after Paccntage PMN
Iotcaritypn~
Phasocytoslng Antisera
s.
to
Phagocytosed
Reblced (n = 9)
mvew
a
SE
49 14 37
2.4 .8 2.4
CP Df
46 13 38 41 32 38 40 31
Dfk
40
6.8 2.5 7.2 5.1 5.7 5.0 9.5 1.8 52
4 3 3 5 3 4 5 21 8 5 19 8
2 .2 2 .2 .3 .5 .4 5.5 3.4 .5 3.0 2.0
CP Df CP Df CP Df Dfk
Dfk
with c p
SE
Dfk
Df
Unabsorbed
x
Dfk
CP
Absorbed
UMbOftd
X
17 3 9 18 4 13 16
9 13 15 9 13
SE 1.0 2 1.0 1.4 .5 1.8 1.7 .3 .9 2.2 .2 .6
Absorbed with cp
x .6 .6 .5
.6 .7 5
.5 9.8 4.6 .6 8.7 5.3
SE .06
.04 .03 .03 .07 .09 .03 1.25 2.32 .12 1.05 1.45
'N = 3 except for prebleed, where each value is the average of all nine cows. PMN = Polymorphonuclear nentrupbik cp = Compact s. aurew, Df = diffuse s. uurnrr; Dfk = diffuse S. aureus. Neonatal calf ~enrmwm used as tbe negative controt P?dN p h a g and ~ in- ~ ~ P m cp, 7 and .6; Df, 4 and .5; and Dfk. 5 and .4.
considering the highly antigenic nature of bacterid cell-wall proteins and their ubiquitous occurrence among a wide variety of bacterial species (16). Failure of the increased antibody titer to Cp to increase phagocytosis of Cp indicates that percentage PMN phagocytosing and the amount ingested were at a maximum for the PMN. Similarity in response between Cp and Dflc reflects the inability of the large flaccid DfIc capsule to block PMN recognition of opsonizing antibody. Conversely. the smaller but more rigid capsule of Df did inhibit phagocytosis. These data suggest that ability of capsule to inhibit phagocytosis is not solely dependent on size of capsule, as suggested by Lee d al. (7) and Peterson et at (12), but is also dependent on the rigidity of the capsule structure. Results of this study show clearly that opsonins, elicited in the cow by either Df or Dflc, increase phagocytosis of either organism, indicating common capsular antigens. The inability of immune sera to enhance phagocytosis of Dflc at the same level as Df, evidenced by lower percentage PMN phagocytosing and lower intensity per PMN after absorptim of antibodies to cell wall, could be attributed to the flaccid nature of DfiC capsule. However, Journal of Dahy Science Vol. 74, No. 10, 1991
the level of both percentage PMN phagocytosing and intensity per PMN of Dfk prior to absorption suggests that antibodies to cell wall can be more effective opsonins than anticapsular antibodies for organisms with large but flaccid capsules. The mctural nature of c a p sule on S. aureus from cases of bovine mastitis has not been studied. These data indicate that this could be a critical factor in determining the antigen to which opsonins should be produced. ACKNOWLEDGMENTS
Authors express appreciation to Chris Pooley for developing and printing the photographs and to L. W.Douglass for the statistical analysis. REFERENCES 1 Bixler, G. S., and P. Snbramonia. 1989. Thc cellular basis of the irnmune response to conjugate vaccines. Page 18 in Conjugate vaccines. J. M.Cruse and R. E. Lewis, ed. Conbib. Mimbiol. Immunol. Vol. 10. S. Kargcr AG. Basel, Switzerland. 2Blobcl. H.,J. Brucklcr, D. Kitmow, and W.Schaeg. 1980. Antiphagacytic factors of Staphylmoccus aur a . a m p . Immrmol. Mimbiol. Infect. DE.3447. 3Cqmty. G.G.. and J. W.Costcrton. 1984. Immuaological examination of the glycocalyces of Sfaphylo-
5 T X P W C O C C U S AUREUS CAPSULE OPSONINS coccus uureus strains Wiley and Smith. Curr. Mibiol. 11:297. 4 Foster,T. J. 1986. A new genetic approach to defining the virulence determinants of Stcrphylococcus aweus strains that Cause bovine mastitis. Ir. Vet. J. 40:llO. 5 W, A. W. 1981. Factors influencing the outcome of Escherichh coli mastitis in the dairy cow. Res. Vet. Sci. 31:107. 6 Karalrawa, W. W., A. SuttOq R. Schneerson. A. Kmpa^, and W. F. V-. 1988. Capantibodits induce type-specific phagocytosis of capsulated Staphylococcus mrreus :,y human polymorphormclear leukocytes. Infect. Immun. 561090. 7Lee, I. C.,M.J. Betley.C. A. Hopkins, N. E. Perez, and G. B. Pier. 1987. V i r u l m studies, in mice, of eansposon-inducedmutants of Stcrphylmoccus aureus differing in capsule size. J. Infect. Dis. 156741. 8 Milliken, G. A., and D. E. Johnson. 1984. ollbway treatment structure in a completely randomized design smcturc with heterogenous errors. Page 17 in Designing experiments. Vol. 1. Analysis of messy data. Van Nostrand Rhinehold Co., New York, NY. 9 Nonxoss, N. L., and J. P.Opdebeeck 1983. Encapsnlation of Staphylmoccrrs a w a s isolated from bovine milk. Vet. Microbiol. 8:397. 1OOjo-Amaizc. E. A., and A. J. Guidry. 1987. In vitro sensitization and stimulation of bovine lymphocytes with encapsulation or nonencapsulated Smith strain of Staphylococcus aureus. Am. J. Vet. Res. 48:1456. llPaape, M. J., W. P. Wergin,A. J. Guidry,and W.D. Schultzc. 1981. phagocytic defense of the ruminant mammary gland. Adv. Exp. Med. Biol. 137555. 12 Petason, P.K.,B. J. WilLinson, Y.Kim, D. Schmeling, and P.G. Quie. 1978. influence of encapsulation on staphylococcal opsonization and phagocytosis by haman polymorphonuclear kukocytes. Infect-Immun. 19943. 13 Poolman. I. T. 1990. Polysaccharides and membraoe vBccines. Page 57 in Bacterid vmines. A. Miziahi.
3369
ed. Contrib. Adv. Biol. Process. Wiley-Lis, Inc., New
Yo& NY. 14 Rather,P. N., A. P. Davis, and B. J. Wilkinson. 1986. S l i m production by bovine milk staphylococcal is+ lam. J. Clin. Mmbiol. 23:858. 15 Saad,A. M., and M.Hageltom.1985. mow cytometry charactahtion of bovine blood neutrophil phagocytosis of fluorescent bacteria and zymosan particles. Acta Vet. Scand. 26289. 16 Schleifer, K.H.1983. The cell envelope. Page 385 in Staphylococci and staphylococcal infections. Vol. 2. C.S.F. Easmon and C. Adlam, ed. Academic Press Inc.. London, Eogl. 1 7 S c h u l ~ W. , D., and M. J. Paape. 1984. Effect of outcom of mr-yi challenge exposure with Staphybcocclrr aureus of somatic cell concentration andpnsaceofan’ . ry device. Am. J. Vet. Res. 45:420. 18Steac, R L. 1981. Reparation of freezbfracture. freezactch, fraeze dry and from surface replica specimens for electron microscopy in the Denton DFE2 and DFE3 freeze-etch units. Page 131 in Current trends in morphological techuiques. Vol. II. J. E. Johmoo. Jr., ed. CRC Press, Boca Raton, FL. 19 Steac, R L.,E. F. Erbc. and J. M. Moseley. 1977. A resistance monitor with power cut-off for automatic regalation of shadow and support f b thickness in frazt-etching and related techniques. J. h4icrosc. (Oxf.)3:313. 20 Watson,D. L., and J. A. Prideam. 1979. Comparisons of Staphylococcus aureus grown in vitro and in vivo. Miaobiol. Immanol. 23543. 21 Watts,I. L. 1988. Etiological agents of bovine mastitis. Vet. Micmbiol. 1641. 22Watts, J. L,J. W. F’ankey, W. M. Oliver, S. C. Nickcrrron. and A. W. Lazams. 1986. Prevalence and effects of mastitis in beef cows. J. Anim. Sci. 6216. 23 W n . B. J. 1983. Staphylococcal capsules and slime. Page 481 in Staphyloccai aud staphylococcal iufections. Vol. 2. C.S.F. Easmon and C. Adlam, ed. Academic Ress Inc.. London. Engl.
J o d of Dairy Science Vol. 74. No. 10, 1991
