Folia Microbiol.39 (5), 428-434 (1994) Antibacterial Activity of Human Mononuclear Leukocytes against Staphylococcus aureus L. PROKE~OV,~ a, DINH HUU DUNGa, I. TREBICHAVSKY b, E. FORMANKOVA a, V. ~TI~PANKOVA a, and C. JOHNa alst Medical Faculty, Charles University, 128 O0 Prague, Czech Republic bLaboratory of Gnotobiology, Institute of Microbiology, Academy of Sciences of Czech Republic, 549 22 Nov~ Hrddek, Czech Republic ReceivedApril 13, 1994 Revised version September30, 1994 ABSTRACT. Human mononuclearleukocyteskill Staphylococcus aureus cells in vitro. The killing of the bacteria takes place even in the absence of antibodies. The presence of antibodies (in an autologous inactivated serum) usually enhances the antibacterial activity of mononuclear ieukocytes.In some cases, however, this activity is markedly decreased by the serum, probably depending of the spectrum of antibodiescontained in the serum. The antibacterial activityof mononuclear leukocytes is mostly due to monocytes because their depletion causes substantial drop or the activity disappearance. We failed to demon- strate in the case of S. aureus the antibacterial cytotoxicity of T lymphocytes described by some authors dealing with Gram-negativebacteria. Large differencesin the structure of the bacterial cell wall underlie apparently the different sensitivity of G + and G- bacteria to some protective mechanisms of the host. In the antibacterial assay against S. aureus, electron microscopyrevealed a maximal activation of monocyteswhich phagocytizedthe bacteria although extracellularkilling is not excluded. Eiectronopticalfindings point also to a possible participation of NK cells in the antibacterial cytotoxicity against S. aureus. In spite of the considerable number of studies dealing with staphylococcal infections, the pathogenesis of these infections and the character of immune reactions involved in the protection of the macroorganism have not yet been fully elucidated. Important factors are undoubtedly phagocytosis and opsonization by antibodies and complement but a direct effector action of lymphocytes cannot be entirely excluded. Not much is as yet known about the cytotoxic action of lymphocytes on bacteria but the existence of these immunological mechanisms has been convincingly proved by several groups. Lowell et al. demonstrated the killing of meningococci (1979, 1980a,b) and Shigella flexneri (1980c) by human leucocytes. Kleinman and Hunt (1982) described the killing ofAcinetobacter calcoaceticus and Escherichia coli in a purified suspension of human lymphoeytes. Neneioni et al. (1983) and Tagliabue et al. (1983, 1984) described an antibacterial activity of mouse lymphocytes against salmonellas and shigellas while Tagliabue et al. (1985a,b) demonstrated antibacterial activity of purified human T lymphocytes against Salmonella typhi. Markham et al. (1985) described the killing of Pseudomonas aeruginosa by isolated mouse T lymphocytes without the participation of antibodies. According to these authors the antibacterial action of mononuclear leukocytes (probably mainly ADCC) involves the effector action of monocytes, T lymphocytes and non B-non T cells. Our study aimed at assessing the significance of a direct effector action of lymphocytes in the killing of S. aureus. We demonstrated antibacterial activity against S. aureus in mononuclear leukocytes. However, the main effector cells in our case appear to be monocytes. MATERIALS AND METHODS Isolation of mononuclear leukocytes. Mononuclear leukocytes were isolated from heparinized blood (10 U heparin per mL) of healthy adult donors by centrifugation on Ficoll-Verografin (B~yum 1984). Isolated cells were washed twice with MEM medium, incubated for 1 h at 37 ~ (release of Ig bound to cell surface) and again washed three times with MEM medium. Monocyte depletion. 107 mononuclear leukocytes in 1 mL RPMI-1640 medium with 10 % fetal calf serum (FCS) was incubated on a polystyrene Petri dish 60 mm in diameter (Koh-i-Noor Hardt- muth, Czech Republic) for 1 h at 37 ~ at a maximum humidity and in 5 % CO2. After incubation the cells were separated by suction, the dish carefully rinsed, the cells were centrifuged and the cell adher- ence performed once more. Nonadherent cells were washed three times with MEM medium.
Transcript
Folia Microbiol. 39 (5), 428-434 (1994)
Antibacterial Activity of Human Mononuclear Leukocytes against Staphylococcus aureus L. PROKE~OV,~ a, DINH HUU DUNG a, I. TREBICHAVSKY b, E. FORMANKOVA a, V. ~TI~PANKOVA a, and C. JOHN a
alst Medical Faculty, Charles University, 128 O0 Prague, Czech Republic bLaboratory of Gnotobiology, Institute of Microbiology, Academy of Sciences of Czech Republic,
549 22 Nov~ Hrddek, Czech Republic
Received April 13, 1994 Revised version September 30, 1994
ABSTRACT. Human mononuclear leukocytes kill Staphylococcus aureus cells in vitro. The killing of the bacteria takes place even in the absence of antibodies. The presence of antibodies (in an autologous inactivated serum) usually enhances the antibacterial activity of mononuclear ieukocytes. In some cases, however, this activity is markedly decreased by the serum, probably depending of the spectrum of antibodies contained in the serum. The antibacterial activity of mononuclear leukocytes is mostly due to monocytes because their depletion causes substantial drop or the activity disappearance. We failed to demon- strate in the case of S. aureus the antibacterial cytotoxicity of T lymphocytes described by some authors dealing with Gram-negative bacteria. Large differences in the structure of the bacterial cell wall underlie apparently the different sensitivity of G + and G- bacteria to some protective mechanisms of the host. In the antibacterial assay against S. aureus, electron microscopy revealed a maximal activation of monocytes which phagocytized the bacteria although extracellular killing is not excluded. Eiectronoptical findings point also to a possible participation of NK cells in the antibacterial cytotoxicity against S. aureus.
In spite of the considerable number of studies dealing with staphylococcal infections, the pathogenesis of these infections and the character of immune reactions involved in the protection of the macroorganism have not yet been fully elucidated. Important factors are undoubtedly phagocytosis and opsonization by antibodies and complement but a direct effector action of lymphocytes cannot be entirely excluded. Not much is as yet known about the cytotoxic action of lymphocytes on bacteria but the existence of these immunological mechanisms has been convincingly proved by several groups. Lowell et al. demonstrated the killing of meningococci (1979, 1980a,b) and Shigella flexneri (1980c) by human leucocytes. Kleinman and Hunt (1982) described the killing ofAcinetobacter calcoaceticus and Escherichia coli in a purified suspension of human lymphoeytes. Neneioni et al. (1983) and Tagliabue et al. (1983, 1984) described an antibacterial activity of mouse lymphocytes against salmonellas and shigellas while Tagliabue et al. (1985a,b) demonstrated antibacterial activity of purified human T lymphocytes against Salmonella typhi. Markham et al. (1985) described the killing of Pseudomonas aeruginosa by isolated mouse T lymphocytes without the participation of antibodies. According to these authors the antibacterial action of mononuclear leukocytes (probably mainly ADCC) involves the effector action of monocytes, T lymphocytes and non B-non T cells.
Our study aimed at assessing the significance of a direct effector action of lymphocytes in the killing of S. aureus. We demonstrated antibacterial activity against S. aureus in mononuclear leukocytes. However, the main effector cells in our case appear to be monocytes.
M A T E R I A L S A N D M E T H O D S
Isolation o f mononuclear leukocytes. Mononuclear leukocytes were isolated from heparinized blood (10 U heparin per mL) of healthy adult donors by centrifugation on Ficoll-Verografin (B~yum 1984). Isolated cells were washed twice with MEM medium, incubated for 1 h at 37 ~ (release of Ig bound to cell surface) and again washed three times with MEM medium.
Monocyte depletion. 107 mononuclear leukocytes in 1 mL RPMI-1640 medium with 10 % fetal calf serum (FCS) was incubated on a polystyrene Petri dish 60 mm in diameter (Koh-i-Noor Hardt- muth , Czech Republic) for 1 h at 37 ~ at a maximum humidity and in 5 % CO2. After incubation the cells were separated by suction, the dish carefully rinsed, the cells were centrifuged and the cell adher- ence performed once more. Nonadherent cells were washed three times with MEM medium.
1994 ANTIBACTERIAL AC'qavrI'Y OF HUMAN MONONUCLEAR LEUKOCYTES 429
Determination of monocytes in cell suspension. Aliquots of ceils before and after adherence were used to make smears. The specimens were stained for unspecific esterase characteristic for monocytes according to Lojda (1958).
Bacteria. The antibacterial assay was carried out with a clinical isolate of S. aureus or E. coli 083 freshly transferred on a solid medium. A suspension of bacteria in physiological saline was pre- pared and its concentration was determined by absorbance measurement.
Antibacterial assay. The assay was performed according to Nencioni et al. (1983) using polystyrene test tubes (10 mL, GAMA, Czech Republic). Bacteria, cells and sera were diluted by the RPMI-1640 medium (Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Prague) supplemented with L-glutamine (2 retool/L), HEPES (2 retool/L) and 10 % FCS. The medium contained no antibiotics. A volume of 0.1 mL bacterial suspension (cell concentration 10/~L) was supplied with 0.1 mL medium or 0.1 mL dilute autologous serum inactivated for 30 min at 56 ~ The test tubes were centrifuged for 10 min at 1300g and 4 ~ Thereafter 0.1 mL cell suspension was added, mostly at a cell concentration 1/nL (106/mL) so that the effector-to-target (E:T) ratio was 100 ." 1. Control test tubes were supplied with 0.1 mL medium instead of the cell suspension. The test tubes were then centrifuged (5 min, 500g, 4 *C) and then incubated for 2 h at 37 ~ a maximum humidity and 5 % CO2. In contrast to the original method we used 10-fold lower concentrations of both cells and bacteria, the fmal volume and E: T ratio being the same. Each sample was assayed in parallel in two test tubes. The pellet after incubation was intensively shaken, the volume was made up to 2 mL and 50 ~tL volumes from each test tube were inoculated on 2 Petri dishes with a solid medium. The dishes were incubated overnight and the number of the colony-forming units (CFU) reflecting the number of surviving bacteria was determined. The antibacterial activity was expressed as an antibacte- rial index (% AI) which defines the per cent reduction of surviving bacteria in the sample as compared with the control count.
% AI = 100 - 100 x (CFUe/CFUc)
where CFUe is the count in the experimental tube and CFUc is the count in the control tube.
When this assay is performed in the presence of antibodies the antibacterial index is denoted % AI-Ab.
Electron microscopy. Cell suspensions were fixed in 1 % paraformaldehyde-glutaraldehyde solution buffered in 0.2 mol/L cacodylate buffer (pH 7.2) and posffixed in 1% buffered OsO 4 Dehy- drated cells were embedded in a Vestopal W polyester resin (Serva, Germany) and ultrathin sections were observed in a Tesla BS 500 transmission electron microscope (Brno, Czech Republic). Ultra- structural characterization of monocytes and lymphocytes was accomplished by using the monograph of M. Bessis (1973). Lymphocytes containing granules were described as granular lymphocytes.
RESULTS
The assay of antibacterial activity of mononuclear leukocytes against S. aureus was carried out by the method described by Nencioni et al. (1983), albeit with 10-fold lower concentrations of both cells and bacteria. The results were comparable with the data obtained at the higher concentrations.
The original method used for demonstrating antibacterial cytotoxicity against G - bacteria achieves optimum results at high values of the E : T ratio (100:1, 200:1 or more). We tested E : T ratios of 200 : 1, 100 �9 1, 10: 1 and in some cases also 1 : 1 (Table I). No perceptible differences were found between the results obtained with E : T -- 200 : 1 and 100 : 1 whereas further lowering of the ratio reduced substantially the efficiency of the antibacterial action. All further assays were therefore performed at E ." T = 100: 1.
The antibacterial activity of mononuclear leukocytes against S. aureus can be demonstrated in the absence of antibodies (% AI) but it usually increases in the presence of antibodies (% AI-Ab). Heat-inactivated serum from the cell donor was used as a source of antibodies. The serum was first diluted 10-1 in order to at least partially approximate in our system the conditions obtaining in the organism in which the antibodies against S. aureus are present in considerable amounts. However, control experiments showed that inactivated serum diluted 10-1 exhibits often a considerable antibac- terial activity. For this reason we tested various dilutions of the sera with regard to their intrinsic activ- ity and also with regard to their possible participation in the antibacterial cytotoxicity of mononuclear
430 L PROKF~OV, A, et al. Voi. 39
leukocytes. The dilution of the serum reduces its antibacterial activity (Table II). At a 10 -4 dilution the activity is negligible but this dilution still substantially enhances the antibacterial activity of the cells. In all further assays we therefore used a serum dilution of 10 -4.
Examination of the 28 healthy donors un- der def'mitely established experimental condi-
Table I. Antibacterial activity of mononuclear leukocytes tions is summarized in Table III. In all donors against S. a u r e u s at effector-to-target (E : T) ratios 100 : 1 and 10:1 a we demonstrated the antibacterial action of
the cells even in the absence of antibodies. In 20 donors this activity, i.e. % AI, was in-
%AI %AI-Ab creased in the presence of antibodies (A), in Donor 8 donors the antibodies lowered the % AI
100:1 10 :1 100:1 10:1 (a). The authors who described the antibacte- H A N 85 62
K U N 68 51 - - rial cytotoxicity against G - bacteria have ENG 62 28.4 77 4 2 . 2 maintained that at least part of this activity is HRU 72 51 87 85 due to T cells and that even T cells alone KAL 89 17.0 72 72 without macrophages can exert this action C R H 42.0 -39 .0 77 41.0 (Lowell 1980c; Kleinman and Hunt 1982; ROS 68 -6.4 74 45.6 Tagliabue et al. 1985a,b; Markham et al. LIS 49.0 13.0 72 60 BED 35.5 26.8 76 64 1985) . W e endeavoured t o a t least tentatively ROT 37.7 24.6 42.1 48.7 determine what is the situation with G § PIC 14.5 -7 .9 62 34.2 staphylococci by comparing the antibacterial MEL 48.8 23.3 68 63 activity of mononuclear leukocytes from pe- CITY 31.8 -6.9 57 -4.7 ripheral blood with the activity of these cells HAN 41.1 0 36.4 16.3 after monocyte depletion. Table IV docu-
ments that the antibacterial activity against S. Mean 53.1 16.9 66.7 47.3 aureus after monocyte depletion ( < 1 %
monocytes in the cell suspension) disappears aAutologous serum added in 10-1 dilution.
or is strongly reduced. Negative values of the indices indicate that the bacteria grow better
in the presence of cells than in in the controls. For comparison, in 4 out of the 6 donors under study we assayed with the same cells also the antibacterial activity against G - E. coli. Depletion of monocytes causes a lowering of antibacterial activity also against E. coli but a considerable part of the activity remains intact. One may conclude that with the G + S. aureus the antibacterial activity of mononuclear leukocytes is mostly due to monocytes whereas with G - E. coli also lymphocytes play their role.
Table II. Effect of serum dilution on antibacterial activ-
ity of the serum itself and on the antibacterial activity of mononuclear leukocytes a
Table I lL Antibacterial activity of mononuclear leuko-
cytes from healthy donors against S. a u r e u s a
Quanti ty A B
% AI mean 46.4 49 range 9 - 76 20 - 74
% AI-Ab mean 65.7 26.5
range 3 3 - 9 2 5 - 4 9
Number of individuals 20 8
aE : T = 100 : 1 (means from 8 experiments). % AI = 45. a E : T = 100: 1, dilution of sera 10-4; A - individuals in
which serum enhances % AI, B - individuals in which serum decreases % AI.
A morphological proof was obtained for the antibacterial activity of monocytes against S. aureus. The pellet containing the cells and the bacteria after incubation was analyzed electronopti- cally. Fig. 1 documents a considerable activation of monocytes and the presence of phagocytized staphylococci. The surface of monocytes carries formations that could be ghosts of dead bacteria. The
1994 ANTIBACTERIAL ACTIVrFY OF HUMAN MONONUCLEAR LEUKOCYTES 4 3 1
extraceHular killing of S. aureus cells by monocytes cannot thus be completely excluded. The ghosts were found on the surface of monocytes in the presence as well as in the absence of antibodies. If the
Table IV. Antibacterial activity of mononuclear leukocytes against S. aureus and E. coli
(comparison of activity of original cell suspension with that of monocyte-depleted suspension) a
Donor Original suspension Monocyte-depleted suspension
ghosts signify extracellular killing the ADCC mechanisms cannot be solely responsible but other cyto- toxic mechanisms independent of antibodies have to play a role. In addition to the unequivocal proof of activated and staphylococci-killing monocytes the electronoptic examination revealed isolated cells such as that shown in Fig. 2. The cell has large granular lymphocyte (LGL) morphology and its surface again carries a formation resembling the ghost of an extracellulary killed bacterium. It is thus possible that also NK cells take part in the killing of S. aureus.
DISCUSSION
We demonstrated the killing of S. aureus cells by mononuclear leukocytes from the peripheral blood of healthy donors. This activity disappears or is strongly reduced after the depletion of mono- cytes by adherence, and is thus due mostly to monocytes not to lymphocytes. Electronoptical proof has been provided that monocytes do not act merely as a source of mediators activating lymphocytes but that they themselves phagocytize staphylococci. After incubation with staphylococci the monocytes exhibit maximum activation and we cannot exclude the possibility that they kill S. aureus also extracel- lulary, both in the absence and in the presence of antibodies. Kleinman and Hunt (1982) who described the killing of the G- Acinetobacter calcoaceticus and E. coli by purified lymphocytes failed to prove a similar killing of a highly virulent S. aureus. In their opinion lymphocytes killed nonpathogenic or mildly pathogenic bacteria but not strongly pathogenic ones. We assume that the differences in the sus- ceptibility to the action of lymphocytes are not related to the pathogenicity of the bacteria but reflect differences in cell wall composition. It is noteworthy that all bacteria that were proven to be killed by lympocytes were G- (Lowell et al. 1979, 1980a-c; Nencioni et al. 1983; Tagliabue et al. 1983, 1985a,b; Markham et al. 1985). There are substantial differences in the cell wall composition between G + and G- bacteria, and it is therefore not surprising that the same antibacterial mechanisms are not effective
432 L. PROKE~OVJ~ eta/. Vol. 39
against the G + S. aureus and against G- bacteria. Both lymphocytes and monocytes can kill G- bac- teria whereas S. aureus is killed in the first place by monocytes. However, one should bear in mind the
Fig. 1. Eiectronoptic picture of cells after incubation with S. aureus during antibacterial assay. Light arrows indicate phagocytized bacteria, dark arrows show formations on monocyte surfaces which could represent ghosts of extracellularly killed bacteria; top: antibacterial assay in the absence of antibodies (18 000x), bottom: antibacterial assay in the presence of antibodies (12 000x).
possible action of NK cells since Garcfa-Pefiarrubia et al. (1989) demonstrated antibacterial activity of human NK cells not only against G- bacteria but also against Staphylococcus epidermidis. Our electronoptic pictures also revealed isolated LGL whose surface carried formations resembling ghosts of dead bacteria. However, the loss of antibacterial activity after depletion of ~idherent cells does not attest to any sizable role of NK cells. Still, the outcome may be dependent on an insufficient sensitivity of our assay system. The above authors who demonstrated convincingly the NK activity against several bacterial species worked with highly purified NK cells. It is clear that monocytes play a major role in
1994 ANTIBACI'ERIAL ACI'IVrYY OF HUMAN MONONUCLEAR LEUKOCYTES 433
the antibacterial activity of mononuclear leukocytes against S. aureus but the participation of NK cells in the process is likely.
n
�9 ~ , . . , . ~ .,~ ~. '~,.~; ~ ,
Fig. 2. Electrorioptic picture of LGL from an antibacterial assay performed in the presence of antibodies (15 000x). The arrow indicates a ghost at cell surface, presumably a remnant of an extracellularly killed bacterium.
The antibacterial activity of mononuclear leukocytes against S. aureus increases in most cases in the presence of autologous sera. The serum antibodies probably play a part during opsonization or during ADCC. It is difficult to explain the "antibacterial activity" of inactivated sera in the absence of the cells. The decrease of CFU number could possibly be caused by agglutination of staphylococci by less diluted sera. The effect disappears at a lower concentration of antibodies which is, however, still sufficient for opsonization or ADCC. Sera from normal donors contain as a rule high levels of anti- staphylococcal antibodies. The total antibody levels determined with the aid of the antigens of staphy- lococcal sonicate are not in correlation with the degree of the potentiation effect of the serum during the antibacterial assay (results not shown). The potentiation includes apparently the action of a small subpopulation of antibodies of a certain specificity and isotype whose level is not revealed during test- ing of total antistaphylococcal antibodies. In a smaller number of donors the autologous serum does not enhance the antibacterial action of the cells. These donors can be assumed to lack antibodies of appropriate properties. Yet even in donors in which the % AI was not increased in the presence of the serum the addition of the serum was not without effect but usually markedly decreased the % AI. One may hypothesize that in the absence of suitable antibodies other antibodies bind to the bacterial surface and actually protect the bacteria against the defence mechanisms of the host via an effect similar to immunologic enhancement. This points to the fact that not all antibodies against staphylococci, which are so copious in human serum, have a protective role but some of them may even interfere with the protective response of the host.
REFERENCES
BESSlS M.: Living Blood Cells and Their Ultrastructure. Springer Verlag, Berlin 1973. B~VUM A.: Separation of lymphocytes, granulocytes, and monocytes from human blood using iodinated density gradient
media. Methods Enzymol. 108, 88-102 (1984). G~r P., Kos'r'~R ET., Km1~'Y R.O., M c D o w ~ T.D., B A ~ u R ~ r A . D . : Antibacterial activity of human
natural killer cells. J.E.~.Med. 169, 99-113 (1989). KL~MAN R., HUNT M.D.: Killing of pathogenic and nonpathogenic bacteria by lymphocytes from normal individuals.
Fed.Proc. 41, 832 (1982) LOJDA Z.: Azo-coupling Reaction in ttistochemical Enzyme Determination. (In Czech) SZdN, Prague 1958.
434 L PROKE~OVfi, eta/. Vol. 39
LOWELL G.H., SMITH L.F., ARTENSTEIN M.S., NASH G.S., MACDERMO'rr Jr. 1LP.: Antibody-dependent cell-mediated antibacte- rial activity of human mononuclear cells I. K lymphocytes and monocytes are effective against meningococci in coop- eration with human immune sera. J.Exp.Med. 150, 127-137 (1979).
LOWELL G.H., MACDERMOTr R.P., SUMMERS P.L., REEDER A.A., BERTOXqCH M3., FORMAL S.B.: Antibody dependent cell- mediated antibacterial activity:. K lymphocytes, monocytes, and granulocytes are effective against Shigella. Zlmmunol. 125, 2778- 2784 (1980c).
MARKHAM R.B., PIER G.B., OOELLNER JJ., MIZEL S.B.: In vitro T cell-mediated killing ofPseudomonas aeruginosa. II. The role of macrophages and T cell subsets in T cell killing. Zlmmunol. 134, 4112-4117 (1985).
NENCIONI L., VILLA L., BORASCHI D., BER~ B., TAGLIABUE A.: Natural and antibody-dependent cell-mediated activity against Salmonella typhimurium by peripheral and intestinal lymphoid cells in mice. ZImmunol. 130, 903-907 (1985).
TAGLIABUE A., NENCIONI L., VILLA L., KEREN D.F., LOWELL G.H., BORASCHI D.: Antibody~ependent cell-mediated antibacte- rial activity of intestinal iymphocytes with secretory IgA. Nature 306, 184-186 (1983).
TAGLIABUE A., BORASCHI D., VILLA L., KEREN D.F., LOWELL G.H., RAPPUOU IL, NENCIONI L.: IgA-dependent cell-mediated activity against enteropathogenic bacteria: Distribution, specificity, and characterization of the effector cells. Zlmmunol. 133,988- 992 (1984).
TAGLIABUE A., NENCIONI L., CAFFARENA A., VILLA L., BORASCHI D., C.AZZOLA G., ~AVALIERI S.: Cellular imunity against Salmolella (yphi after live oral vaccine. Clin.Exp.lmmunol. 62, 242 (1985a).
TAGLIABUE A., VILLA L., BORASCHi D., PERI G., DE GORI V., NENCIONI L.: Natural anti-bacterial activity against Salmonella typhi by human T4 + lymphocytes armed with IgA antibodies. Zlmmunol. 135, 4178-4182 (1985b).
