Microbial Pathogenesis 1991 ; 11 : 211-220
A periodate-sensitive anti-phagocytic surface structure, induced by growth in milk whey, on Staphylococcus aureus isolated from bovine mastitis C . Hallen Sandgren,' W . Mamo, 2 I . Larsson,' M . Lindahl 2 and I . Bjork' 'Department of Veterinary Medical Chemistry and 2Department of Veterinary Microbiology, Swedish University of Agricultural Sciences, Uppsala Biomedical Center, S-751 23 Uppsala, Sweden (Received March 18, 1991 ; accepted in revised form May 10, 1991)
Hallen Sandgren, C . (Dept of Veterinary Medical Chemistry, Swedish University of Agricultural Sciences, Uppsala Biomedical Center, S-751 23 Uppsala, Sweden), W . Mamo, I . Larsson, M . Lindahl and I . Bjork . A periodate-sensitive anti-phagocytic surface structure, induced by growth in milk whey, on Staphylococcus aureus isolated from bovine mastitis . Microbial Pathogenesis 1991 ; 11 : 211-220 . The phagocytic and chemiluminescent activity of purified bovine neutrophils in response to two Staphylococcus aureus strains isolated from mastitic bovine milk and grown in milk whey was studied . The activity was significantly reduced compared with the response elicited by the same strains grown in tryptic soy broth (TSB) . A mild periodate treatment of the milk wheygrown strains resulted in a significant increase of both chemiluminescence and phagocytosis, whereas trypsin, subtilisin or papain treatment had no effect . The decreased binding of complement factor C3 to milk-whey-grown bacteria was restored to the level of TSB-grown homologous organisms by periodate treatment . Moreover, this treatment, but not treatment with trypsin, increased the surface hydrophobicity of milk-whey-grown bacteria . The chemiluminescent activity was as high towards heat-killed as towards live bacteria . Also, incubation of heat-killed TSB-grown bacteria in milk whey did not alter the chemiluminescent response, indicating that the reduced neutrophil activity towards milk-whey-grown bacteria was not due to binding of milk components to the microorganisms . These results strongly suggest that bovine mastitis S . aureus strains grown in milk whey produce an anti-phagocytic surface structure . This structure is heat- and protease-resistant and renders the bacterial surface hydrophilic . The anti-phagocytic material is altered or, more likely, released from the bacterial surface on periodate treatment and is probably of carbohydrate nature . Key words : Staphylococcus aureus ; bovine ; mastitis ; neutrophils ; periodate ; milk whey .
Introduction It is well documented that the virulence of Staphylococcus aureus increases when the bacteria are grown in vivo ." 2 However, it has been difficult to relate this increase in virulence to a particular structural component of the bacteria .` Karakawa and Vann reported that most S. aureus strains isolated from human infections and opsonized with non-immune serum resisted phagocytosis . 6 The phagocytic resistance has been suggested to be due to a material surrounding the bacteria, referred to as a 'microcapsule' . 6 This material has been identified as a carbohydrate structure consisting of repeating trisaccharide units .' The extracellular capsule on human S . aureus strains 0882-4010/91 /090211 +10 $03 .00/0
© 1991 Academic Press Limited
C . Hallen Sandgren et al_
21 2
renders the microorganism unrecognizable to the complement system and thereby also to polymorphonuclear phagocytes .' ,' Some evidence indicates that the production of capsular material is preferentially induced when S. aureus is grown in certain media .""' An analogous capsular material has recently been suggested to be present also on S . aureus isolated from bovine mastitis . 12-15 We have previously found that when S . aureus, isolated from mastitic milk, is grown in milk or milk-supplemented media, the surface of the bacteria is altered from hydrophobic to hydrophilic in nature ." Recently, we were able to show that five out of six S . aureus strains grown in milk whey were not only more virulent to mice than the same strains grown in tryptic soy broth (TSB) but also expressed a diffuse-type colony morphology in serum soft agar, in contrast with the compact-type colony morphology of the TSB-grown homologous organisms ." The milk-whey-grown S. aureus strains showed a significant increase in resistance to both phagocytosis by bovine neutrophil granulocytes and in vivo clearance from the peritoneal cavity in mice ." Further examination of two of these strains demonstrated that the whey-grown bacteria bound complement factor C3 to a considerably lower extent than the bacteria grown in TSB ." In this work we show that the anti-phagocytic properties of these two S. aureus strains are due to a surface structure that is altered or, more likely, released by a mild periodate treatment and probably is of carbohydrate nature . Results Bacterial cell-surface hydrophobicity
Bacteria grown in TSB or milk whey and treated with periodate or trypsin were analysed by the salt aggregation test (Table 1) . The whey-grown staphylococci lost their hydrophilic surface properties after periodate treatment and instead acquired hydrophobic properties similar to that of TSB-grown bacteria . In contrast, trypsin treatment had no effect on the surface hydrophilicity of whey-grown bacteria, whereas it decreased the hydrophobicity of the bacteria grown in TSB . Phagocytosis
A considerable proportion of the bacteria grown in TSB were ingested by bovine neutrophils, whereas the uptake of the bacteria grown in milk whey was minimal (Fig . 1) . The difference between whey-grown and TSB-grown bacteria was completely Table 1 The effect of periodate and trypsin treatments on surface hydrophobicity of three S . aureus strains isolated from bovine mastitis and grown in milk whey or TSB SAT value (M)a
S . aureus strains
Bacteria grown in
Buffer
Periodate
Trypsin
F1440
whey TSB
1 .6 0 .05
0 .1 0 .05
1 .8 1 .6
Mil 0064
whey TSB
1 .6 0 .1
0 .05 0 .1
1 .8 2 .0
'Surface hydrophobicity was determined by the salt aggregation test (SAT) . Values are expressed as the lowest molar concentration of NH,(S0 4 ) 2 at which bacteria aggregated . Low SAT values indicate high hydrophobicity, whereas high SAT values indicate low hydrophobicity .29
A periodate-sensitive anti-phagocytic structure on S . aureus
213
100 80 ° 60 U O T
t 40 a 0 20
ao N H
m
2M2 HH CL
N
3
ro HH
UJ
H
3
3
a
M
07
3
T
mH
L3 a
C
H
GI
3
H
Fig . 1 . Percentage of added TSB-grown or milk-whey-grown bacteria, untreated or treated with periodate (P) or trypsin jr), phagocytized by bovine neutrophils after 30 min incubation at 37'C . (a) Unopsonized bacteria ; (b) serum-opsonized bacteria . Each bar represents the mean+SD of four experiments with neutrophils from different cows . Each experiment was run in triplicate . ∎ : strain F1440 ; D : strain mj10064 . "`P < 0 .001 obtained from analysis of variance . Comparisons were made with the non-treated, TSB-grown bacteria opsonized in the same manner .
abolished after treatment of the whey-grown bacteria with periodate (Fig . 1) . In contrast, such treatment had no effect on the ingestion by neutrophils of TSB-grown staphylococci . Moreover, no effect on phagocytosis was observed when bacteria grown under either of the two conditions were treated with trypsin (Fig . 1) . The uptake was similar for unopsonized bacteria and bacteria opsonized with an IgG-free serum . Chemiluminescence The chemiluminescent response of bovine neutrophils towards milk-whey-grown staphylococci was significantly lower than that towards homologous TSB-grown bacteria (Fig . 2) . These responses were essentially the same regardless of whether unopsonized or serum-opsonized bacteria were used (Fig . 2) . However, the results
300
200
100
0
M N H
N t
CO N H a
Ol
3 a
m HH H
T N
3 H
U) HH
3
m H a
T
L
3
a
M H H
C1 3 H
Fig . 2 . Total chemiluminescence (in arbitrary units) produced during 30 min at 37°C by bovine neutrophils in response to bacteria treated as in Fig . 1 . (a) Unopsonized bacteria ; (b) serum-opsonized bacteria . Each bar represents the mean+SD of four experiments with neutrophils isolated from different cows . Each experiment was run in duplicate . Symbols and abbreviations are as in Fig . 1 .
C. Hallen Sandgren et al.
21 4
obtained with unopsonized bacteria was more variable, presumably owing to a greater sensitivity to small changes in bacterial concentration observed with unopsonized bacteria . The difference between whey- and TSB-grown bacteria was abolished after treatment of the whey-grown bacteria with periodate, a treatment that did not affect the chemiluminescent response towards the TSB-grown bacteria (Fig . 2) . In contrast, no change in chemiluminescent activity could be detected between trypsin-treated and untreated bacteria grown under either of the two conditions, nor did treatment with papain or subtilisin alter the chemiluminescent reaction (data not shown) . A control experiment demonstrated that treatment of milk whey-grown bacteria with NaBH, alone had no effect on the chemiluminescent activity towards these bacteria (data not shown) . In a further control experiment TSB-grown bacteria were heat-killed and incubated overnight in milk whey . The chemiluminescent response evoked by these bacteria were found not to differ from that elicited by untreated homologous TSB-grown bacteria (Fig . 3) . This result indicates that the increased resistance to phagocytosis was not due to the binding of milk components to the bacterial surface . It was also established (Fig . 4) that the differences in activity of neutrophils towards the bacteria grown under the two conditions was not a result of heat-treatment, since the chemiluminescence elicited by unopsonized heat-killed bacteria was essentially the same as that evoked by homologous live bacteria grown in the same medium .
Binding of complement factor C3 We have previously shown that the binding of complement factor C3 to serumopsonized, milk-whey-grown bacteria is markedly reduced compared with the binding to homologous TSB-grown bacteria ." Periodate treatment of milk-whey-grown bacteria was found to restore the level of binding of complement factor C3 to these bacteria during serum-opsonization essentially back to the level of TSB-grown bacteria (Fig . 5) .
(a)
(b)
300 U C UU C
E
200
Ev
L U
100
*
0
Fig . 3 . The effect of incubation of heat-killed, TSB-grown bacteria in milk whey on total chemiluminescence (in arbitrary units) produced by bovine neutrophils . (a) Unopsonized bacteria ; (b) serumopsonized bacteria . The assay was performed as for Fig . 2 . Each bar represents the mean+SD of experiments run in duplicate with neutrophils isolated from three different cows. TSB-whey : TSB-grown, heat-killed bacteria incubated in milk whey; other abbreviations and symbols are as in Fig . 1 . ** _< 0 .001 ; "'P S 0.01 and `P S 0 .05 in a two-tailed Student's t-test of comparisons with the TSB-grown homologous bacteria opsonized in the same manner .
215
A periodate-sensitive anti-phagocytic structure on S . aureus
O c CO
O
v
c
a
0) Y
Y
T O L
CO IF
N
T a) L
3
3 Fig . 4 . Total chemiluminescence (in arbitrary units) produced by bovine neutrophils towards living and heat-killed unopsonized bacteria grown in TSB or milk whey . Each bar represents the mean+SD of experiments run in duplicate with neutrophils from three different cows . Incubation conditions, abbreviations and symbols are as in Figs 1 and 2 . "'P < 0 .01 in a two-tailed Student's t-test of comparisons with the TSB-grown homologous bacteria.
0 .5
0 .4
e 0 .3
0 Q
d 0 .2
0 .1
0 .0
m IF
VO
L
3
O G
3
a
Fig . 5 . Binding of bovine complement factor C3 to serum-opsonized bacteria, analysed by ELISA . The values of the ordinate were calculated by subtraction of the absorbance given by unopsonized bacteria from the absorbance measured for serum-opsonized bacteria . Each bar represents the mean+SD of an experiment run in triplicate . The titer of the primary F(ab') 2 fragments was 1 :8 for strain F1440 and 1 :7 for strain mj10064 . Abbreviations and symbols are as in Figs 1 and 2 . 'P < 0 .05 in a two-tailed Student's t-test of comparisons with the untreated, whey-grown bacteria .
21 6
C . Hallen Sandgren et al.
Discussion We have previously demonstrated that two S . aureus strains (F1440 and mj10064), isolated from bovine mastitic milk, elicit a significantly lower chemiluminescent response and are phagocytized to a lower degree by bovine neutrophil granulocytes when grown in milk whey than when conventionally grown in TSB . The reduced uptake of S. aureus after growth in milk whey was considered to be due to the appreciably lower amount of complement bound to these cells compared with that bound to S. aureus grown in TSB ." In the present study, however, we show that the difference in neutrophilic phagocytic activities elicited by bacteria grown in the two media is not dependent on prior opsonization with IgG-free serum . This observation may indicate the existence of lectin-like recognition mechanisms between TSB-grown bacteria and neutrophils, which are not present on whey-grown bacteria . This phenomenon, known as 'lectinophagocytosis' refers to the interaction of proteins (lectins) on either the phagocyte or the bacterium with specific sugar residues on the surface of the counterpart .' 9,2' The results demonstrate that both complement factor C3- and the putative lectin-receptor recognition mechanisms are totally abolished when S . aureus strains isolated from milk of cows with clinical mastitis are grown in milk whey . The resistance to phagocytosis induced by growth of the bacteria in milk whey is probably due to the expression of a specific anti-phagocytic surface structure on the bacteria that is highly periodate sensitive . Thus, both phagocytosis and chemiluminescence by neutrophils was fully restored to the level of the control (TSB-grown bacteria) by a mild treatment of bacteria grown in milk whey with periodate . Moreover, periodate treatment of milk-whey-grown bacteria increased the surface hydrophobicity and the binding of complement factor C3 to the level of TSB-grown bacteria . On the other hand, heating or treatment of the milk-whey-grown bacteria with proteases affected neither the bacterial cell surface hydrophobicity nor the phagocytic activities of neutrophils towards the bacteria . When the TSB-grown bacteria were treated with periodate or proteases, no effect on phagocytosis or chemiluminescence was detected . However, a total abolishment of surface hydrophobicity was observed when the TSBgrown bacteria were treated with trypsin . These results thus indicate that surface hydrophobicity is only of minor importance for the effective ingestion of these bacteria, in agreement with recently reported studies on the hydrophobicity of uropathogenic Escherichia coil in relation to phagocytosis . 21 The structural feature most sensitive to periodate under the conditions used is considered to be vicinal hydroxyl groups ." The dramatic changes of the hydrophobic surface properties of the bacteria grown in milk whey following periodate treatment indicate that the anti-phagocytic material on the surface of the milk-whey-grown organisms is released on treatment with periodate . It is therefore likely that this material is attached to the bacterial cell wall via a periodate-sensitive structure . This structure could be the teichoic acid of the cell wall, since this molecule contains vicinal diols and its release from S. aureus on periodate treatment has been reported . 23 Alternatively, periodate may oxidize essential sugar structures in the anti-phagocytic material without releasing this material . However, in such a case a polysaccharide of the type demonstrated in the 'microcapsule' of human strains of S . aureus is a poor candidate for the anti-phagocytic surface structure of the bovine strains studied in this work, as this carbohydrate appears to lack periodate-sensitive structures .' Moreover, the antigenic properties of the previously reported surface material of S . aureus strains isolated from bovine mastitis were shown to be similar to those of human strains, 13 - 14 indicating that also this material may not be sensitive to periodate .
A period ate-sensitive anti-phagocytic structure on S.
aureus
217
The importance of neutrophil phagocytosis and killing in host defence mechanisms towards S. aureus infections is well established, 24 and the pathophysiological implications of our results therefore are intriguing . The ability of bacteria to escape from neutrophil phagocytosis by altering their surface structures has been reported for S . aureus strains isolated from human bacteremia . The hydrophilic polysaccharide capsular material discussed above, which appears to be preferentially induced under certain conditions, thus renders the organism unrecognizable by the complement system, leading to a decreased phagocytosis of the bacteria .' - " It is apparent that the antiphagocytic surface material on bovine mastitis S . aureus strains grown in milk whey functions in an analogous manner . Moreover, the anti-phagocytic properties induced in this manner may well resemble those of bacteria colonizing the lactating udder . The development of the surface structure therefore could be of crucial importance for the ability of the bacteria to escape uptake and killing by phagocytes in the udder . The acquired surface structure may thus facilitate colonization of S. aureus in the udder tissue . Moreover, one of the reasons for the failure of many vaccination trials against bovine S. aureus mast itis25 may have been that the bacteria used for immunization lacked this surface structure . This possibility is indicated by the demonstration that an effective neutrophilic uptake of phagocytosis-resistant human S . aureus strains can be induced by specific antibodies against the polysaccharide surface material .' In summary, grown in milk whey of S . aureus strains isolated from clinical bovine mastitis induces the production of an anti-phagocytic surface structure on the bacteria . This structure is hydrophilic and resistant to heat and protease treatment . It is altered or, more likely, removed by a mild periodate treatment and is probably of carbohydrate nature . Studies on the structure of this material are in progress .
Materials and methods Chemicals. [ 3 H]-methyl thymidine was purchased from Amersham, U .K . Bovine serum albumin, trypsin, papain (type III), soy bean trypsin inhibitor, lyostaphin, luminol (5-amino2,3-dihydro-1,4-phtalazinedione), goat anti-rabbit IgG (Fab') 2 fragment and p-nitrophenylphosphate were from Sigma Chemical Company, St Louis, Missouri . Subtilisin was from Boehringer Mannheim, Penzberg, Germany . Rabbit anti-bovine C3 serum was purchased from Organontecniqa, Capel Industries, West Chester, Pennsylvania . Metrizamide was from Nyegaard, Oslo, Norway . TSB was purchased from Difco Chemical, Detroit, Michigan . Phosphate buffered saline (0.01 M sodium phosphate, 0 .14 M NaCl, 1 mm Ca 21, 1 MM Mg 2 ', pH 7 .4) was used as buffer . Modified Gey's solution had the following composition : 0.13 M NaCl, 20 mm Hepes, 5.0 mm KCI, 3 .0 mm NaHC0 3 , 2 .5 mm CaC1 2 , 1 .0 MM MgC1 2 and 11 .0 mm glucose ; the pH of this solution was adjusted to 7 .4 with concentrated NaOH . All other chemicals were of analytical grade . Bacteria . Staphylococcus aureus strains F1440 and mjl 0064 were isolated from the milk of cows with clinical mastitis. Isolation and identification were done according to recommendations for examination of milk samples at the National Veterinary Institute, Uppsala, Sweden .26 Preparation of milk whey and bacteria . Milk whey was prepared as described earlier ." A bacterial culture was prepared by inoculating an overnight blood agar culture into 1 .0 ml milk whey, followed by incubation at 37°C for 10 h . The culture was subsequently transferred into 100 ml milk whey and incubated at 37°C for another 10 h . The bacteria were washed twice in phosphate buffered saline, pH 7 .4, and killed by heating at 80°C for 10 min . For studies of phagocytosis, bacteria were cultured in 10 ml growth media containing 0 .02 mCi [ 3 H]-methyl thymidine (specific activity 5 Ci/mmol) . Following incubation at 37°C for 18 h, the bacterial cells were washed in phosphate buffered saline until the radioactivity in the supernatant was reduced to the background value . In all assays, homologous strains treated in the same way but grown in TSB were used as controls .
218
C . Hallen Sandgren et ai,
Treatment of bacterial cells . The bacterial concentration was 5x10 9/ml in all treatments Bacterial suspensions were treated with trypsin (1 mg/ml) at 37°C for 60 min in 0 .11 m sodium phosphate buffer, pH 7 .4 . The reaction was stopped by addition of soy bean trypsin inhibitor to 1 mg/ml, followed by 40 min incubation at 20°C . Treatments with papain (5 /I/ml) in 0 .11 m sodium phosphate buffer, pH 7 .0, and subtilisin (0 .1 mg/ml) in 0 .1 m Tris-HCI, pH 7 .6, were done at 37°C for 60 min . Periodate treatment was done by incubating bacterial suspensions with 20 mm periodic acid in 0 .11 m sodium phosphate buffer, pH 6 .5, at 4°C for 20 min in the dark. The cells were then washed twice and treated with 10 mm NaBH 4 in 0 .11 m sodium phosphate buffer, pH 6 .5, for 60 min at room temperature . In control experiments the bacteria were treated only with NaBH 4 . The influence of binding of milk whey components to the bacteria was analysed by incubating heat-killed TSB-grown bacteria in milk whey at 37°C for 10 h . After all treatments, the bacteria were washed twice, counted and suspended in phosphate buffered saline to an appropriate concentration . Opsonization procedures . IgG-poor (
A periodate-sensitive anti-phagocytic surface structure, induced by growth in milk whey, on Staphylococcus aureus isolated from bovine mastitis C . Hallen Sandgren,' W . Mamo, 2 I . Larsson,' M . Lindahl 2 and I . Bjork' 'Department of Veterinary Medical Chemistry and 2Department of Veterinary Microbiology, Swedish University of Agricultural Sciences, Uppsala Biomedical Center, S-751 23 Uppsala, Sweden (Received March 18, 1991 ; accepted in revised form May 10, 1991)
Hallen Sandgren, C . (Dept of Veterinary Medical Chemistry, Swedish University of Agricultural Sciences, Uppsala Biomedical Center, S-751 23 Uppsala, Sweden), W . Mamo, I . Larsson, M . Lindahl and I . Bjork . A periodate-sensitive anti-phagocytic surface structure, induced by growth in milk whey, on Staphylococcus aureus isolated from bovine mastitis . Microbial Pathogenesis 1991 ; 11 : 211-220 . The phagocytic and chemiluminescent activity of purified bovine neutrophils in response to two Staphylococcus aureus strains isolated from mastitic bovine milk and grown in milk whey was studied . The activity was significantly reduced compared with the response elicited by the same strains grown in tryptic soy broth (TSB) . A mild periodate treatment of the milk wheygrown strains resulted in a significant increase of both chemiluminescence and phagocytosis, whereas trypsin, subtilisin or papain treatment had no effect . The decreased binding of complement factor C3 to milk-whey-grown bacteria was restored to the level of TSB-grown homologous organisms by periodate treatment . Moreover, this treatment, but not treatment with trypsin, increased the surface hydrophobicity of milk-whey-grown bacteria . The chemiluminescent activity was as high towards heat-killed as towards live bacteria . Also, incubation of heat-killed TSB-grown bacteria in milk whey did not alter the chemiluminescent response, indicating that the reduced neutrophil activity towards milk-whey-grown bacteria was not due to binding of milk components to the microorganisms . These results strongly suggest that bovine mastitis S . aureus strains grown in milk whey produce an anti-phagocytic surface structure . This structure is heat- and protease-resistant and renders the bacterial surface hydrophilic . The anti-phagocytic material is altered or, more likely, released from the bacterial surface on periodate treatment and is probably of carbohydrate nature . Key words : Staphylococcus aureus ; bovine ; mastitis ; neutrophils ; periodate ; milk whey .
Introduction It is well documented that the virulence of Staphylococcus aureus increases when the bacteria are grown in vivo ." 2 However, it has been difficult to relate this increase in virulence to a particular structural component of the bacteria .` Karakawa and Vann reported that most S. aureus strains isolated from human infections and opsonized with non-immune serum resisted phagocytosis . 6 The phagocytic resistance has been suggested to be due to a material surrounding the bacteria, referred to as a 'microcapsule' . 6 This material has been identified as a carbohydrate structure consisting of repeating trisaccharide units .' The extracellular capsule on human S . aureus strains 0882-4010/91 /090211 +10 $03 .00/0
© 1991 Academic Press Limited
C . Hallen Sandgren et al_
21 2
renders the microorganism unrecognizable to the complement system and thereby also to polymorphonuclear phagocytes .' ,' Some evidence indicates that the production of capsular material is preferentially induced when S. aureus is grown in certain media .""' An analogous capsular material has recently been suggested to be present also on S . aureus isolated from bovine mastitis . 12-15 We have previously found that when S . aureus, isolated from mastitic milk, is grown in milk or milk-supplemented media, the surface of the bacteria is altered from hydrophobic to hydrophilic in nature ." Recently, we were able to show that five out of six S . aureus strains grown in milk whey were not only more virulent to mice than the same strains grown in tryptic soy broth (TSB) but also expressed a diffuse-type colony morphology in serum soft agar, in contrast with the compact-type colony morphology of the TSB-grown homologous organisms ." The milk-whey-grown S. aureus strains showed a significant increase in resistance to both phagocytosis by bovine neutrophil granulocytes and in vivo clearance from the peritoneal cavity in mice ." Further examination of two of these strains demonstrated that the whey-grown bacteria bound complement factor C3 to a considerably lower extent than the bacteria grown in TSB ." In this work we show that the anti-phagocytic properties of these two S. aureus strains are due to a surface structure that is altered or, more likely, released by a mild periodate treatment and probably is of carbohydrate nature . Results Bacterial cell-surface hydrophobicity
Bacteria grown in TSB or milk whey and treated with periodate or trypsin were analysed by the salt aggregation test (Table 1) . The whey-grown staphylococci lost their hydrophilic surface properties after periodate treatment and instead acquired hydrophobic properties similar to that of TSB-grown bacteria . In contrast, trypsin treatment had no effect on the surface hydrophilicity of whey-grown bacteria, whereas it decreased the hydrophobicity of the bacteria grown in TSB . Phagocytosis
A considerable proportion of the bacteria grown in TSB were ingested by bovine neutrophils, whereas the uptake of the bacteria grown in milk whey was minimal (Fig . 1) . The difference between whey-grown and TSB-grown bacteria was completely Table 1 The effect of periodate and trypsin treatments on surface hydrophobicity of three S . aureus strains isolated from bovine mastitis and grown in milk whey or TSB SAT value (M)a
S . aureus strains
Bacteria grown in
Buffer
Periodate
Trypsin
F1440
whey TSB
1 .6 0 .05
0 .1 0 .05
1 .8 1 .6
Mil 0064
whey TSB
1 .6 0 .1
0 .05 0 .1
1 .8 2 .0
'Surface hydrophobicity was determined by the salt aggregation test (SAT) . Values are expressed as the lowest molar concentration of NH,(S0 4 ) 2 at which bacteria aggregated . Low SAT values indicate high hydrophobicity, whereas high SAT values indicate low hydrophobicity .29
A periodate-sensitive anti-phagocytic structure on S . aureus
213
100 80 ° 60 U O T
t 40 a 0 20
ao N H
m
2M2 HH CL
N
3
ro HH
UJ
H
3
3
a
M
07
3
T
mH
L3 a
C
H
GI
3
H
Fig . 1 . Percentage of added TSB-grown or milk-whey-grown bacteria, untreated or treated with periodate (P) or trypsin jr), phagocytized by bovine neutrophils after 30 min incubation at 37'C . (a) Unopsonized bacteria ; (b) serum-opsonized bacteria . Each bar represents the mean+SD of four experiments with neutrophils from different cows . Each experiment was run in triplicate . ∎ : strain F1440 ; D : strain mj10064 . "`P < 0 .001 obtained from analysis of variance . Comparisons were made with the non-treated, TSB-grown bacteria opsonized in the same manner .
abolished after treatment of the whey-grown bacteria with periodate (Fig . 1) . In contrast, such treatment had no effect on the ingestion by neutrophils of TSB-grown staphylococci . Moreover, no effect on phagocytosis was observed when bacteria grown under either of the two conditions were treated with trypsin (Fig . 1) . The uptake was similar for unopsonized bacteria and bacteria opsonized with an IgG-free serum . Chemiluminescence The chemiluminescent response of bovine neutrophils towards milk-whey-grown staphylococci was significantly lower than that towards homologous TSB-grown bacteria (Fig . 2) . These responses were essentially the same regardless of whether unopsonized or serum-opsonized bacteria were used (Fig . 2) . However, the results
300
200
100
0
M N H
N t
CO N H a
Ol
3 a
m HH H
T N
3 H
U) HH
3
m H a
T
L
3
a
M H H
C1 3 H
Fig . 2 . Total chemiluminescence (in arbitrary units) produced during 30 min at 37°C by bovine neutrophils in response to bacteria treated as in Fig . 1 . (a) Unopsonized bacteria ; (b) serum-opsonized bacteria . Each bar represents the mean+SD of four experiments with neutrophils isolated from different cows . Each experiment was run in duplicate . Symbols and abbreviations are as in Fig . 1 .
C. Hallen Sandgren et al.
21 4
obtained with unopsonized bacteria was more variable, presumably owing to a greater sensitivity to small changes in bacterial concentration observed with unopsonized bacteria . The difference between whey- and TSB-grown bacteria was abolished after treatment of the whey-grown bacteria with periodate, a treatment that did not affect the chemiluminescent response towards the TSB-grown bacteria (Fig . 2) . In contrast, no change in chemiluminescent activity could be detected between trypsin-treated and untreated bacteria grown under either of the two conditions, nor did treatment with papain or subtilisin alter the chemiluminescent reaction (data not shown) . A control experiment demonstrated that treatment of milk whey-grown bacteria with NaBH, alone had no effect on the chemiluminescent activity towards these bacteria (data not shown) . In a further control experiment TSB-grown bacteria were heat-killed and incubated overnight in milk whey . The chemiluminescent response evoked by these bacteria were found not to differ from that elicited by untreated homologous TSB-grown bacteria (Fig . 3) . This result indicates that the increased resistance to phagocytosis was not due to the binding of milk components to the bacterial surface . It was also established (Fig . 4) that the differences in activity of neutrophils towards the bacteria grown under the two conditions was not a result of heat-treatment, since the chemiluminescence elicited by unopsonized heat-killed bacteria was essentially the same as that evoked by homologous live bacteria grown in the same medium .
Binding of complement factor C3 We have previously shown that the binding of complement factor C3 to serumopsonized, milk-whey-grown bacteria is markedly reduced compared with the binding to homologous TSB-grown bacteria ." Periodate treatment of milk-whey-grown bacteria was found to restore the level of binding of complement factor C3 to these bacteria during serum-opsonization essentially back to the level of TSB-grown bacteria (Fig . 5) .
(a)
(b)
300 U C UU C
E
200
Ev
L U
100
*
0
Fig . 3 . The effect of incubation of heat-killed, TSB-grown bacteria in milk whey on total chemiluminescence (in arbitrary units) produced by bovine neutrophils . (a) Unopsonized bacteria ; (b) serumopsonized bacteria . The assay was performed as for Fig . 2 . Each bar represents the mean+SD of experiments run in duplicate with neutrophils isolated from three different cows. TSB-whey : TSB-grown, heat-killed bacteria incubated in milk whey; other abbreviations and symbols are as in Fig . 1 . ** _< 0 .001 ; "'P S 0.01 and `P S 0 .05 in a two-tailed Student's t-test of comparisons with the TSB-grown homologous bacteria opsonized in the same manner .
215
A periodate-sensitive anti-phagocytic structure on S . aureus
O c CO
O
v
c
a
0) Y
Y
T O L
CO IF
N
T a) L
3
3 Fig . 4 . Total chemiluminescence (in arbitrary units) produced by bovine neutrophils towards living and heat-killed unopsonized bacteria grown in TSB or milk whey . Each bar represents the mean+SD of experiments run in duplicate with neutrophils from three different cows . Incubation conditions, abbreviations and symbols are as in Figs 1 and 2 . "'P < 0 .01 in a two-tailed Student's t-test of comparisons with the TSB-grown homologous bacteria.
0 .5
0 .4
e 0 .3
0 Q
d 0 .2
0 .1
0 .0
m IF
VO
L
3
O G
3
a
Fig . 5 . Binding of bovine complement factor C3 to serum-opsonized bacteria, analysed by ELISA . The values of the ordinate were calculated by subtraction of the absorbance given by unopsonized bacteria from the absorbance measured for serum-opsonized bacteria . Each bar represents the mean+SD of an experiment run in triplicate . The titer of the primary F(ab') 2 fragments was 1 :8 for strain F1440 and 1 :7 for strain mj10064 . Abbreviations and symbols are as in Figs 1 and 2 . 'P < 0 .05 in a two-tailed Student's t-test of comparisons with the untreated, whey-grown bacteria .
21 6
C . Hallen Sandgren et al.
Discussion We have previously demonstrated that two S . aureus strains (F1440 and mj10064), isolated from bovine mastitic milk, elicit a significantly lower chemiluminescent response and are phagocytized to a lower degree by bovine neutrophil granulocytes when grown in milk whey than when conventionally grown in TSB . The reduced uptake of S. aureus after growth in milk whey was considered to be due to the appreciably lower amount of complement bound to these cells compared with that bound to S. aureus grown in TSB ." In the present study, however, we show that the difference in neutrophilic phagocytic activities elicited by bacteria grown in the two media is not dependent on prior opsonization with IgG-free serum . This observation may indicate the existence of lectin-like recognition mechanisms between TSB-grown bacteria and neutrophils, which are not present on whey-grown bacteria . This phenomenon, known as 'lectinophagocytosis' refers to the interaction of proteins (lectins) on either the phagocyte or the bacterium with specific sugar residues on the surface of the counterpart .' 9,2' The results demonstrate that both complement factor C3- and the putative lectin-receptor recognition mechanisms are totally abolished when S . aureus strains isolated from milk of cows with clinical mastitis are grown in milk whey . The resistance to phagocytosis induced by growth of the bacteria in milk whey is probably due to the expression of a specific anti-phagocytic surface structure on the bacteria that is highly periodate sensitive . Thus, both phagocytosis and chemiluminescence by neutrophils was fully restored to the level of the control (TSB-grown bacteria) by a mild treatment of bacteria grown in milk whey with periodate . Moreover, periodate treatment of milk-whey-grown bacteria increased the surface hydrophobicity and the binding of complement factor C3 to the level of TSB-grown bacteria . On the other hand, heating or treatment of the milk-whey-grown bacteria with proteases affected neither the bacterial cell surface hydrophobicity nor the phagocytic activities of neutrophils towards the bacteria . When the TSB-grown bacteria were treated with periodate or proteases, no effect on phagocytosis or chemiluminescence was detected . However, a total abolishment of surface hydrophobicity was observed when the TSBgrown bacteria were treated with trypsin . These results thus indicate that surface hydrophobicity is only of minor importance for the effective ingestion of these bacteria, in agreement with recently reported studies on the hydrophobicity of uropathogenic Escherichia coil in relation to phagocytosis . 21 The structural feature most sensitive to periodate under the conditions used is considered to be vicinal hydroxyl groups ." The dramatic changes of the hydrophobic surface properties of the bacteria grown in milk whey following periodate treatment indicate that the anti-phagocytic material on the surface of the milk-whey-grown organisms is released on treatment with periodate . It is therefore likely that this material is attached to the bacterial cell wall via a periodate-sensitive structure . This structure could be the teichoic acid of the cell wall, since this molecule contains vicinal diols and its release from S. aureus on periodate treatment has been reported . 23 Alternatively, periodate may oxidize essential sugar structures in the anti-phagocytic material without releasing this material . However, in such a case a polysaccharide of the type demonstrated in the 'microcapsule' of human strains of S . aureus is a poor candidate for the anti-phagocytic surface structure of the bovine strains studied in this work, as this carbohydrate appears to lack periodate-sensitive structures .' Moreover, the antigenic properties of the previously reported surface material of S . aureus strains isolated from bovine mastitis were shown to be similar to those of human strains, 13 - 14 indicating that also this material may not be sensitive to periodate .
A period ate-sensitive anti-phagocytic structure on S.
aureus
217
The importance of neutrophil phagocytosis and killing in host defence mechanisms towards S. aureus infections is well established, 24 and the pathophysiological implications of our results therefore are intriguing . The ability of bacteria to escape from neutrophil phagocytosis by altering their surface structures has been reported for S . aureus strains isolated from human bacteremia . The hydrophilic polysaccharide capsular material discussed above, which appears to be preferentially induced under certain conditions, thus renders the organism unrecognizable by the complement system, leading to a decreased phagocytosis of the bacteria .' - " It is apparent that the antiphagocytic surface material on bovine mastitis S . aureus strains grown in milk whey functions in an analogous manner . Moreover, the anti-phagocytic properties induced in this manner may well resemble those of bacteria colonizing the lactating udder . The development of the surface structure therefore could be of crucial importance for the ability of the bacteria to escape uptake and killing by phagocytes in the udder . The acquired surface structure may thus facilitate colonization of S. aureus in the udder tissue . Moreover, one of the reasons for the failure of many vaccination trials against bovine S. aureus mast itis25 may have been that the bacteria used for immunization lacked this surface structure . This possibility is indicated by the demonstration that an effective neutrophilic uptake of phagocytosis-resistant human S . aureus strains can be induced by specific antibodies against the polysaccharide surface material .' In summary, grown in milk whey of S . aureus strains isolated from clinical bovine mastitis induces the production of an anti-phagocytic surface structure on the bacteria . This structure is hydrophilic and resistant to heat and protease treatment . It is altered or, more likely, removed by a mild periodate treatment and is probably of carbohydrate nature . Studies on the structure of this material are in progress .
Materials and methods Chemicals. [ 3 H]-methyl thymidine was purchased from Amersham, U .K . Bovine serum albumin, trypsin, papain (type III), soy bean trypsin inhibitor, lyostaphin, luminol (5-amino2,3-dihydro-1,4-phtalazinedione), goat anti-rabbit IgG (Fab') 2 fragment and p-nitrophenylphosphate were from Sigma Chemical Company, St Louis, Missouri . Subtilisin was from Boehringer Mannheim, Penzberg, Germany . Rabbit anti-bovine C3 serum was purchased from Organontecniqa, Capel Industries, West Chester, Pennsylvania . Metrizamide was from Nyegaard, Oslo, Norway . TSB was purchased from Difco Chemical, Detroit, Michigan . Phosphate buffered saline (0.01 M sodium phosphate, 0 .14 M NaCl, 1 mm Ca 21, 1 MM Mg 2 ', pH 7 .4) was used as buffer . Modified Gey's solution had the following composition : 0.13 M NaCl, 20 mm Hepes, 5.0 mm KCI, 3 .0 mm NaHC0 3 , 2 .5 mm CaC1 2 , 1 .0 MM MgC1 2 and 11 .0 mm glucose ; the pH of this solution was adjusted to 7 .4 with concentrated NaOH . All other chemicals were of analytical grade . Bacteria . Staphylococcus aureus strains F1440 and mjl 0064 were isolated from the milk of cows with clinical mastitis. Isolation and identification were done according to recommendations for examination of milk samples at the National Veterinary Institute, Uppsala, Sweden .26 Preparation of milk whey and bacteria . Milk whey was prepared as described earlier ." A bacterial culture was prepared by inoculating an overnight blood agar culture into 1 .0 ml milk whey, followed by incubation at 37°C for 10 h . The culture was subsequently transferred into 100 ml milk whey and incubated at 37°C for another 10 h . The bacteria were washed twice in phosphate buffered saline, pH 7 .4, and killed by heating at 80°C for 10 min . For studies of phagocytosis, bacteria were cultured in 10 ml growth media containing 0 .02 mCi [ 3 H]-methyl thymidine (specific activity 5 Ci/mmol) . Following incubation at 37°C for 18 h, the bacterial cells were washed in phosphate buffered saline until the radioactivity in the supernatant was reduced to the background value . In all assays, homologous strains treated in the same way but grown in TSB were used as controls .
218
C . Hallen Sandgren et ai,
Treatment of bacterial cells . The bacterial concentration was 5x10 9/ml in all treatments Bacterial suspensions were treated with trypsin (1 mg/ml) at 37°C for 60 min in 0 .11 m sodium phosphate buffer, pH 7 .4 . The reaction was stopped by addition of soy bean trypsin inhibitor to 1 mg/ml, followed by 40 min incubation at 20°C . Treatments with papain (5 /I/ml) in 0 .11 m sodium phosphate buffer, pH 7 .0, and subtilisin (0 .1 mg/ml) in 0 .1 m Tris-HCI, pH 7 .6, were done at 37°C for 60 min . Periodate treatment was done by incubating bacterial suspensions with 20 mm periodic acid in 0 .11 m sodium phosphate buffer, pH 6 .5, at 4°C for 20 min in the dark. The cells were then washed twice and treated with 10 mm NaBH 4 in 0 .11 m sodium phosphate buffer, pH 6 .5, for 60 min at room temperature . In control experiments the bacteria were treated only with NaBH 4 . The influence of binding of milk whey components to the bacteria was analysed by incubating heat-killed TSB-grown bacteria in milk whey at 37°C for 10 h . After all treatments, the bacteria were washed twice, counted and suspended in phosphate buffered saline to an appropriate concentration . Opsonization procedures . IgG-poor (
