Orcd Mierobiol Itntnunol 1992: 7: 89-95
The differential effects of polymorphonuclear leukocyte secretion on human natural killer cell activity
A. Lala\ R. A. Lindemann^ K. T. Miyasaici' Sections of 'Oral Biology and ^Hospital Dentistry and Dentai Research Institute, UCLA School of Dentistry, Los Angeles, California, USA
Lala A, Lindetnann RA, Miyasakl KT. The dijferetitial effects of polytnorplwtiuclear leukocyte secretion on hutnan natural killer cell activity. Orat Microbiol Itntnutwl 1992: 7: 89-95. We studied the differential effects of polymorphonuclear leukocyte (PMN) secretion on human natural killer cell (NK) and lytnphokine-activated killer cell (LAK) activity. Supematant fiuids frotn PMN stitnulated by serum-opsonized zytnosan (SOZ), n-fortnyhnethiotiylleucylphenylalanine (FMLP) and interleukin-8 (lL-8) were incubated with periphetal blood lymphocytes (PBL) for 1 d and 4 d. Supernates frotn unstimulated PMN and PMN induced by FMLP and IL-8 deereased NK atid LAK cytotoxicity in a dose-dependent fashion against K562 and M14 targets, respectively. Only the suppression caused by supernates from unstitnulated PMN was ablated by incubation of PMN with indotnetliacin. Secretions frotn PMN stitnulated by SOZ increased both NK and LAK cytotoxicity and induced PBL proliferatioti synergistically with IL-2. This enhancing factor was heat-labile, nondialyzable (MWCO 3500), and not blocked by anti-interferon-y. Anaerobic conditions did not infiuence the modulatory activity of PMN supernates, indicating that oxygen tnetabolites were not involved. We conclude that PMN release factors that tnodulate //; vitro NK and LAK activities.
Natutal killer (NK) cells are peripheral blood lytnphoeytes with spontaneous cytotoxicity agaitist tumor cells, viruses, and baeteria. Unlike eytotoxic T lymphocytes, NK cells lack CD3 and Tcell leceptor tnolecules (21) and are not restricted by the tnajor histocotnpatibility complex in exerting eytotoxic effects (22, 26, 28, 29). NK cells are reeeptive to aetivation by interleukin-2 (IL-2) and mediate a wider spectrum of target killing when activated (11, 15). NK cells activated by IL-2 constitute the tnajority of lytnphokine-activated killer (LAK) cells (29). Additionally, baeteria and bacterial products activate NK cells (17, 34) and NK cells can kill certain gratn-negative bacteria by an extracellular, nonphagocytic mechanism (10). NK cells increase locally at diseased sites in experimental gingivitis (41), ean kill gingival fibroblasts when activated by dental plaque (32), and exhibit increased cytotoxicity and the capacity to kill NK-resistant targets after exposure to periodontopathic bacteria or their lipopolysaccharides (LPS) (17, 18). NK
cells are an important source of interferon-}' (IFNy), which they elaborate when stimulated with IL-2 or hydrogen peroxide (24, 35), suggesting that NK cells may be important compotients of the itntnunoregulatory circuitry involved in the pathogenesis of periodontal diseases. Polytnorphonuclear leukocytes (PMN) play an itnportant protective role it! periodontal disease, usually attributed to antitnicrobial activities (36). The interaction between PMN atid lytnphoid cells has received little attentioti in relationship to the periodontal diseases. "Unstitnulated" PMN interact with NK cells; and to be specific, decrease NK cytotoxicity (33). We believe that PMN may protect against periodontal destruction iti a second way, by interaetion with lytnphoid cells and regulation of either their cytotoxic activity or their cytokine production. PMN are itnportant secretory cells (13). They contain a tnixture of pritnary (azurophil), secotidary (specific), and tertiary (secretory) granules, which per-
Key words: polymorphonuclear leukocyte; NK cell; LAK cell; cytotoxicity; secretion Dr. Robert A. Lindemann, UCLA School of Dentistry, 13-089 CHS, Los Angeles, CA 900241668, USA Accepted for publication May 21, 1991
mits the differential release of numerous biologically active products, depending on the stitnulus (1,2, 39, 40). We hypothesized that secretion resulting from exposure to different stimuli tnay have differential regulatory effects on lytnphocyte activity. In this study, we initiated an examination of the interplay between PMN and lytnphocytes and report how PMN secretions affected the cytotoxicity of peripheral blood lymphocytes against standard NK and LAK targets. i\/iateriai and metiiods Blood ceil isolation
Heparinized whole blood was diluted l.i in Dulbecco's phosphate-buffered saline (PBS) and centrifuged (400 g) for 30 min on a density gradient of FicollPaque (Phartnacia-LKB Biotechnologies, Piseataway, NJ) to obtain an upper layer enriched in peripheral blood tnononuclear cells and a red blood cell pellet. The peripheral blood tnononuclear cells were incubated in tissue
90
Lala et al.
culture dishes (Costar; Cambridge, MA) at a concentration of 2 x 10'' per ml of "cotnplete tnedium" consisting of RPMI 1640 containing 10% hutnan AB serum, penicillin/streptomycin/fungizone, 25 mM HEPES buffer and 5 pgl ml of polymyxin B sulfate (Sigma Chemical, St. Louis, MO). After 1-2 h incubation, non-adherent cells were eluted with warm PBS and resuspended in complete medium, and are hereafter referred to as peripheral blood lymphocytes (PBL). PMN were obtained by diluting the red blood cell pellet 1:4 in complete medium. PMN were separated from red cells by layering the diluted RBC pellet on a 62% isotonic Percoll gradient (Pharmacia-LKB Biotechnologies). Residual red eells were lysed by hypotonic shock with distilled water as necessary. The isolated cells were washed free of Percoll and resuspended in complete medium. Cells were greater than 95% PMN and 98% viable as assessed by Wright's stain and trypan blue dye exclusion, respectively.
Target ceii preparation
Stimuiation of PiVIN secretion
PMN, 2x 10'" cells/ml, were incubated in microcentrifuge tubes with either FMLP, SOZ, or IL-8 at 37°C, with 4 rpm end-over-end rotation. After 15 min the reaction was terminated by chilling for 5 min on ice and microcentrifugation at 11,000 g for 1 min. The supernate was collected and frozen at — 85°C. Stimulation of PMN secretion was also performed in an anaerobic chamber (Forma Scientific, Marietta, OH) under an atmosphete containing 5% CO,, 10%. Hj, and 85% Nj (20). To verify that the systetn was functioning anaerobically, superoxide produetion was tneasured using a cytochrome c (Sigma) reduction assay (38). Superoxide dismutase (Sigma) was added to the supernate at a concentration of 0.5 mg/ ml, to serve as control. The reduction of cytochrome c, 0.1 mM, was determined by measuring the ehatige in absorbance at 550 nm spectrophotometrically. Total cell-associated myeioperoxidase was detertnined by sonieatioti, 10 s, of whole cells in 0.5%) cetyltritnethylammonium brotnide (Aldrieh Chemical, Milwaukee, Wl). Total cellassociated lysozyme and lactate dehydrogenase were determined by 10 s sonieation of whole cells in 0.1% Triton X-100 (Fisher Scientific, Santa Clara, CA).
The NK-sensitive human erythroleukemia cell line, K562, and the NK-resistant melanoma cell line, UCLA SO-M14 (Ml4) (42) were used as targets in the cytotoxicity assays. The M14 target cell line is a LAK-sensitive target. Target cells (5 X lOMn 1 tnl of RPMI 1640 with 10%) fetal calf serum) were labelled with 250 p.C\ of "Cr-sodium chromate, 0.2 iViicroenzyme assays ^g/ml, for 1 h at 37 C. Target cells were Supernatant myeioperoxidase, lysowashed 3 titnes in RPMI 1640 with 10% zyme and lactate dehydrogenase activifetal calf serum. ties, used to assess azurophil granule release, total granule release, and PMN lysis, respectively, were analyzed in Reagents and cytokines microplate format (4, 6, 8, 19).
Zymosan A (Sigma) was pre-opsonized with lO'Mi normal human serum, washed Treatment of PBL witii PiWN supernates with PBS, and resuspended in complete medium. Reagents and cytokines were PMN were stimulated by varying conused at the following concentrations in centrations of FMLP, SOZ, and IL-8 cotnplete tnedium: indomethaein (Sig- for 15 min. Then 400 p\ of supernate ma), 10 /ig/ml; thromboxane Bj (TXBj; from each condition was incubated with Sigma), 5-10 //g/ml; IL-2 (Amgen; PBL at a concentration of 2.5 x 10' per Thousand Oaks, CA) 25 U/ml; poly- ml in a total volume of 2 tnl. PBL were clonal rabbit anti-human IFN-y anti- incubated with PMN supernates for serum (Genzyme, Boston, MA), 100 either 24 h and 4 d for the NK assay NU/tnl; n-formylmethionylleucylphen- and LAK assay, respectively. PBL only ylalanitie(FMLP, Sigma), 10-^-10-'M; served as eontrol in NK assay. In the IL-8 (Genzytne), 0.1-10 ng/ml; and LAK assay, IL-2 (25 U/ml) was added serum pre-opsonized zytnosan (SOZ), to PBL in the presence of supernates 0.02-2 mg/ml. The FMLP stock was from PMN stimulated as described initially solubilized in dimethyl sulfox- above. PBL incubated with IL-2 only served as control. After the incubation ide at 10-' M.
period, PBL were washed twice with PBS resuspended in complete medium. Cytotoxicity assay
Target cells (5x10') were mixed with PBL (at ratios of 50:1, 25:1, and 12.5;1) in round-bottotn tnicrotiter plates containing 200 //I of cotnplete medium. The plates were centrifuged at 65 g for 4 tnin to initiate cell-to-cell contact and then incubated for 4 h at 37°C in an hutnidified incubator with 5%o COj. At the end of the reaction, 100 p\ of supematant fiuid was harvested frotn each well and counted for "Cr released from target cells. Each assay was perfortned in triplicate. Cytotoxicity was defttied as the percentage of speeific ^'Cr released or [(experimental release-spontaneous release)/(tnaximal release-spontaneous release)] X 100%. Experimental release was defined as the eounts of "Cr released from target cells caused by effector cells as tneasured by counts per minute. Maxitnal release was the "Cr released frotn target cells induced by 2% Nonidet P-40 detergent. Spontaneous release was the counts of "Cr from target cells incubated in medium alone. Lymphoproiiferation assay
PBL were cultured in tubes for 6 d with IL-2 or IL-2 plus supernatant fluids from PMN at 37 C in a hutnidified incubator with 5% CO2. After the iticubation period, PBL were washed free of PMN supernates. PBL (10' cells/well in 200 /;1 complete mediutn) were seeded in tnicrotiter wells and each well was pulsed with 1 (iCi of [methyl-'H]thymidine (New England Nuclear, Boston, MA) for 5 h under similar incubation and collected onto filter paper with a tnultiple satnple harvester (Catnbridge Technology, Cambridge, MA). Each assay was done in triplicate. Calculations
The cytotoxicity at 3 effector-to-target ratios (50:1, 25:1, 12.5:1) was converted into lytic units (LU 30%) (30). LU 30% arc defined as the nutnber of cells required to cause a 3O'M) target lysis and are expressed per lO*" PBL. The paired /-test (two-tailed) was then applied to determine the significance of LU 30%) values.
Effects of PMN secretion on NK activity Results PiUIN secretion in response to FiUILP and SOZ
of 10-' M for FMLP and 10 ng/tnl for IL-8. In all experitnents, PBL without supernatant fluids frotn PMN served as a control. Additionally, to detertnine any direct effect of SOZ, FMLP and IL-8 on NK activity, these agents were incubated with PBL in eotnplete tnediutn under sitnilar conditions atid NK cytotoxicity retnained unchanged (data not shown).
Under our assay conditions, up to 3%) of azurophil granule cotnponents were released in the presence of FMLP at 10" nM as assessed by supernatant levels of MPO (Fig. 1 A). About 3%) of all granules were released at 10^ nM of FMLP as assessed by lysozyme activity. A higher percentage of azurophil gratiule release relative to the percentage of all Effects of secretions on NK and LAK eell granules released was observed at cytotoxicity 10' nM, suggesting a dissociation bePBL iticubated with unstitnulated PMN tween the two events. SOZ produced supernates exhibited slightly decreased about 4% enzytne release at 0.2 nig/ml NK and LAK cytotoxicity for both sub(Fig. IB). Basal levels of enzytne release jects tested (Table 1). A tnore signifieant occurred iti the absetice of exogetious reduction in NK and LAK cytotoxicity secretagogues (Fig. IA, B). In no inwas observed when PBL were incubated stance was tnore than 0.6%) of cytoplaswith FMLP- and lL-8-stitnulated PMN mic lactate dehydrogenase released. secretions (Table 1). In contrast, SOZstimulated PMN secretion increased NK and LAK eytotoxieity (Table 1). Effects of PMN secretions, using varying doses of secretagogues, on NK cytotoxicity
Fig. 2 shows the effects of secretagogueinduced (PMN) supernatant fluids on subsequent NK activity after 24 h. SOZstitnulated supernates induced a dosedepetident iticrease in cytotoxicity (Fig. 2A). Peak cytotoxicity occurred at a concentration of 0.2 tng/tnl. PMN supernatant stitnulated with FMLP (concentrations ranging frotn 10"' to 10^' M) and lL-8 (concentratiotis ranging from 0.1-10 ng/tnl) decreased cytotoxicity in a dose-dependent tnanner (Fig. 2B, C). The peak decrease in cytotoxicity was observed at a concentration
iVIodulation of iympiioproiiferation by PiUiN secretion
Supertiate collected frotn both stitnulated and unstitnulated PMN were tested for their effect on PBL proliferation induced by IL-2. Cotnpared with the control containing PBL-I-IL-2, only supernate collected from PMN stitnulated by SOZ-HL-2 had an effect on proliferation (Fig. 3). SOZ-stitnulated PMN secretion in eotijunction with IL2 induced a significant increase in 'HTdR uptake. At 6 d, this inciease in proliferation coiticided with a signifi-
—
7 B
ASb
65 4 -
^^.
3-
v; -'.
HJ UJ CC — L zU
o
CC JJ
^'•. ^
m
0 -
10
100 1000
[FMLP], nM
.
I1
1-
1
'
2-
0
0.002 0.02
0.2
91
cant increase in LAK cytotoxicity against M14 target cells (Fig. 3). Effect of anaerobic conditions
All reagents and PMN were equilibrated for 2 h in an anaerobic chatnber (19). After anaerobic equilibration, half the PMN were stimulated in the anaerobic chatnber and the remaining PMN were retnoved frotn the chatnber and stitnulated aerobically using the standard secretagogues. Supernates from PMN were collected and superoxide production was measured. A 94%) redttction in superoxide anion production was observed under anaerobic conditions using PMN suspended at of 6x lO"^ cells/ml (data not showti). Despite this tnarked ditninution in the reduction of dioxygen, anaerobically stitnulated PMN seeretions had the satne effect on the PBL cytotoxicity as their aerobic counterparts (Table 2). Role of PGEj and tfiromboxane in NK suppression
As PMN secrete prostaglandin Ei (PGE2), which is known to inhibit NK activity (14), we tested the possible role of PGEi in the unstitnulated and FMLP- and IL-8-induced cytotoxic inhibition by using indomethacin, which blocks the production of PGEi. Indotnethacin was added to the PMN before stitnulation. Ten pglmX of indotnethacin reversed the suppression in NK cytotoxicity caused by the utistimulated PMN secretions (Table 3). However, the NK inhibition caused by FMLP- and IL-8stimulated PMN secretions was unaffected by indotnethaein. PMN also produce large amounts of TXB, when stitnulated (23). As indotnethacin blocks both PGE, and TXB,, we detertnined whether throtnboxane rather than PGET tnay be responsible for the inhibitive effeet of unstitnulated PMN supernate. TXB,, incubated at a concentration of 5-10 //g/tnl, had no significant effect on NK cytotoxicity. The cytotoxicity of PBL against K562 targets when incubated with TXB, ranged frotn 3.2 + 0.3 to 3.5±0.3 LU 30%), and the cytotoxicity demonstrated by PBL alone was 2.9 ±1.3 LU 30%.
2
[ZYMOSAN], mg/mi
Fig. I. PMN secretion in response to (A) FMLP and (B) sertim-opsonized-zymosan. Values are expressed in terms of percentage of total cell enzyme. Markers tised include myeioperoxidase (MPO) and lysozyme. Bars and vertical lines represent the means and SD of triplicate assay, respectively.
Roie of t F N - / in NK activation
We initiated studies aitned at determitiitig whether the stimulatory effect of SOZ-stimulated PMN secretions on NK
Lala et al.
92
5 -
o CO
5 n
10-
4 3-
O X
3 -
6-
2-
o
2 -
4-
1-
1 -
o .0
None
0.02 [SOZ],
0.2
None
2
10--9
10-8
1-7
None
[FMLP], M
mg/ml
0.1
1
[IL-8],
ng/ml
10
Fig. 2. The effect of (A) SOZ-, (B) FMLP- and (C) IL-8-stimulated PMN secretion on PBL cytotoxicity after 24 h incubation against K562 targets. Points and vertical lines represent the mean and SD of triphcate assay, respectively.
cytotoxicity was direct or indirect. One potential indirect mechanism may involve the stimulation of IFN-y release by T cells or NK cells (31). IFN-y is a potent up-regulator of NK-induced
cytotoxicity (12). PBL alone atid PBL cultured with IFN-y served as controls. NK activity in PBL cultured with IFN)' showed a significant increase in cytotoxicity over PBL alone (PBL = 2.8±
Table I. The effects of PMN secretions on NK and LAK cytotoxicity Cytotoxicity (LU 30%±SD)" Subject 2
Subject 1 Condition PBL only PBL + unstim. sup.'' PBL + SOZ-stim. sup.' PBL + FMLP-stim. sup.' PBL + IL-8-stim. sup.'
NK
LAK
NK
LAK
19.5 + 2.3 15.8 + 0.8 38.8 + 9.6 9.8 + 1.3 9.1+2.1
45.5 + 5.8 39.0 + 4.1 60.3 + 3.2 26.7 + 2.8 22.5 + 3.0
21.5 + 2.4 17.6+1.3 39.7 + 6.8 13.0 + 0.9 10.8 + 2.9
50.8 + 9.8 39.3 + 2.0 62.9 + 9.6 30.0 + 7.9 29.4 + 9.1
"Values represent the mean and SD of these experiments. ''PBL cytotoxicity against K562 (NK targets) and M14 (LAK targets) cells after 1 d (NK) and 4 d (LAK) incubations with a 1:5 dilution of supernates from unstimuiated PMN. PBL were washed free of supernates prior to the assay. "PBL cytotoxicity as above after incubation with supernates from PMN stimulated with SOZ, 0.2 mg/ml; FMLP, 10 ' M; or IL-8, 1 ng/ml.
0.5 LU 30%; PBL + IFN-)' = 8.2±0.3 LU 30'^). This increase in cytotoxicity was completely reversed by the addition of anti-lFN-)' antiserum (100 NU/ml) during culture. Anti-IFN-)'was added to the PBLcultutes incubated with SOZ-indticed PMN secretions. This antibody had no effect on the increased cytotoxicity caused by the SOZ stimulation (PBL + SOZ = 5.1 ±0.9 LU 30%; PBL + SOZ + anti-IFN-7 = 5.1 +0.7 LU 30%). Characterization of SOZ-stimulated PMN secretion
The dialysis (3500 MWCO, against complete medium) of PMN secretions
120 •
'„ 100
T g
Table 2. The effects PMN secretions generated under aerobic and anaerobic conditions on NK cytotoxicity Cytotoxicity (LU30%±SD)'' Conditions
Aerobic
Anaerobic
PBL only PBL + unstim. sup'' PBL + FMLP-stim. sup.' PBL + SOZ-stim. sup.' PBL+IL-8-stim. sup.'
2.8 + 0.5 2.1+0.6 1.7 + 0.2
Not determined 2.3 + 0.7 1.2 + 0.3
5.1+0.9
5.2+1.0
1.3 + 0.4
1.2 + 0.3
"Values represent the mean and SD of 3 experiments. ''PBL cytotoxicity against K562 (NK targets) cells after 1 d incubation with a 1:5 dilution of supernates from unstimulated PMN. PBL were washed free of supernates prior to the assay. TBL cytotoxicity as above after incubation with supernates from PMN stimulated with SOZ, 0.2 mg/ml, FMLP, 10"' M; or IL-8, 1 ng/ml. . •
t
T
T
6040-
g Q,
20 •
Ftg. 3. The effect of differentially stimulated PMN secretion on PBL cytotoxicity and proliferation after 6 d culture with IL-2 against M14 targets. The supernates used were derived from unstimulated PMN and from PMN stimulated with I-MLP, SOZ, and IL8. Bars and vertical lines represent the mean and SD of triplicate assay, respectively. *Significantly different from control at P
The differential effects of polymorphonuclear leukocyte secretion on human natural killer cell activity
A. Lala\ R. A. Lindemann^ K. T. Miyasaici' Sections of 'Oral Biology and ^Hospital Dentistry and Dentai Research Institute, UCLA School of Dentistry, Los Angeles, California, USA
Lala A, Lindetnann RA, Miyasakl KT. The dijferetitial effects of polytnorplwtiuclear leukocyte secretion on hutnan natural killer cell activity. Orat Microbiol Itntnutwl 1992: 7: 89-95. We studied the differential effects of polymorphonuclear leukocyte (PMN) secretion on human natural killer cell (NK) and lytnphokine-activated killer cell (LAK) activity. Supematant fiuids frotn PMN stitnulated by serum-opsonized zytnosan (SOZ), n-fortnyhnethiotiylleucylphenylalanine (FMLP) and interleukin-8 (lL-8) were incubated with periphetal blood lymphocytes (PBL) for 1 d and 4 d. Supernates frotn unstimulated PMN and PMN induced by FMLP and IL-8 deereased NK atid LAK cytotoxicity in a dose-dependent fashion against K562 and M14 targets, respectively. Only the suppression caused by supernates from unstitnulated PMN was ablated by incubation of PMN with indotnetliacin. Secretions frotn PMN stitnulated by SOZ increased both NK and LAK cytotoxicity and induced PBL proliferatioti synergistically with IL-2. This enhancing factor was heat-labile, nondialyzable (MWCO 3500), and not blocked by anti-interferon-y. Anaerobic conditions did not infiuence the modulatory activity of PMN supernates, indicating that oxygen tnetabolites were not involved. We conclude that PMN release factors that tnodulate //; vitro NK and LAK activities.
Natutal killer (NK) cells are peripheral blood lytnphoeytes with spontaneous cytotoxicity agaitist tumor cells, viruses, and baeteria. Unlike eytotoxic T lymphocytes, NK cells lack CD3 and Tcell leceptor tnolecules (21) and are not restricted by the tnajor histocotnpatibility complex in exerting eytotoxic effects (22, 26, 28, 29). NK cells are reeeptive to aetivation by interleukin-2 (IL-2) and mediate a wider spectrum of target killing when activated (11, 15). NK cells activated by IL-2 constitute the tnajority of lytnphokine-activated killer (LAK) cells (29). Additionally, baeteria and bacterial products activate NK cells (17, 34) and NK cells can kill certain gratn-negative bacteria by an extracellular, nonphagocytic mechanism (10). NK cells increase locally at diseased sites in experimental gingivitis (41), ean kill gingival fibroblasts when activated by dental plaque (32), and exhibit increased cytotoxicity and the capacity to kill NK-resistant targets after exposure to periodontopathic bacteria or their lipopolysaccharides (LPS) (17, 18). NK
cells are an important source of interferon-}' (IFNy), which they elaborate when stimulated with IL-2 or hydrogen peroxide (24, 35), suggesting that NK cells may be important compotients of the itntnunoregulatory circuitry involved in the pathogenesis of periodontal diseases. Polytnorphonuclear leukocytes (PMN) play an itnportant protective role it! periodontal disease, usually attributed to antitnicrobial activities (36). The interaction between PMN atid lytnphoid cells has received little attentioti in relationship to the periodontal diseases. "Unstitnulated" PMN interact with NK cells; and to be specific, decrease NK cytotoxicity (33). We believe that PMN may protect against periodontal destruction iti a second way, by interaetion with lytnphoid cells and regulation of either their cytotoxic activity or their cytokine production. PMN are itnportant secretory cells (13). They contain a tnixture of pritnary (azurophil), secotidary (specific), and tertiary (secretory) granules, which per-
Key words: polymorphonuclear leukocyte; NK cell; LAK cell; cytotoxicity; secretion Dr. Robert A. Lindemann, UCLA School of Dentistry, 13-089 CHS, Los Angeles, CA 900241668, USA Accepted for publication May 21, 1991
mits the differential release of numerous biologically active products, depending on the stitnulus (1,2, 39, 40). We hypothesized that secretion resulting from exposure to different stimuli tnay have differential regulatory effects on lytnphocyte activity. In this study, we initiated an examination of the interplay between PMN and lytnphocytes and report how PMN secretions affected the cytotoxicity of peripheral blood lymphocytes against standard NK and LAK targets. i\/iateriai and metiiods Blood ceil isolation
Heparinized whole blood was diluted l.i in Dulbecco's phosphate-buffered saline (PBS) and centrifuged (400 g) for 30 min on a density gradient of FicollPaque (Phartnacia-LKB Biotechnologies, Piseataway, NJ) to obtain an upper layer enriched in peripheral blood tnononuclear cells and a red blood cell pellet. The peripheral blood tnononuclear cells were incubated in tissue
90
Lala et al.
culture dishes (Costar; Cambridge, MA) at a concentration of 2 x 10'' per ml of "cotnplete tnedium" consisting of RPMI 1640 containing 10% hutnan AB serum, penicillin/streptomycin/fungizone, 25 mM HEPES buffer and 5 pgl ml of polymyxin B sulfate (Sigma Chemical, St. Louis, MO). After 1-2 h incubation, non-adherent cells were eluted with warm PBS and resuspended in complete medium, and are hereafter referred to as peripheral blood lymphocytes (PBL). PMN were obtained by diluting the red blood cell pellet 1:4 in complete medium. PMN were separated from red cells by layering the diluted RBC pellet on a 62% isotonic Percoll gradient (Pharmacia-LKB Biotechnologies). Residual red eells were lysed by hypotonic shock with distilled water as necessary. The isolated cells were washed free of Percoll and resuspended in complete medium. Cells were greater than 95% PMN and 98% viable as assessed by Wright's stain and trypan blue dye exclusion, respectively.
Target ceii preparation
Stimuiation of PiVIN secretion
PMN, 2x 10'" cells/ml, were incubated in microcentrifuge tubes with either FMLP, SOZ, or IL-8 at 37°C, with 4 rpm end-over-end rotation. After 15 min the reaction was terminated by chilling for 5 min on ice and microcentrifugation at 11,000 g for 1 min. The supernate was collected and frozen at — 85°C. Stimulation of PMN secretion was also performed in an anaerobic chamber (Forma Scientific, Marietta, OH) under an atmosphete containing 5% CO,, 10%. Hj, and 85% Nj (20). To verify that the systetn was functioning anaerobically, superoxide produetion was tneasured using a cytochrome c (Sigma) reduction assay (38). Superoxide dismutase (Sigma) was added to the supernate at a concentration of 0.5 mg/ ml, to serve as control. The reduction of cytochrome c, 0.1 mM, was determined by measuring the ehatige in absorbance at 550 nm spectrophotometrically. Total cell-associated myeioperoxidase was detertnined by sonieatioti, 10 s, of whole cells in 0.5%) cetyltritnethylammonium brotnide (Aldrieh Chemical, Milwaukee, Wl). Total cellassociated lysozyme and lactate dehydrogenase were determined by 10 s sonieation of whole cells in 0.1% Triton X-100 (Fisher Scientific, Santa Clara, CA).
The NK-sensitive human erythroleukemia cell line, K562, and the NK-resistant melanoma cell line, UCLA SO-M14 (Ml4) (42) were used as targets in the cytotoxicity assays. The M14 target cell line is a LAK-sensitive target. Target cells (5 X lOMn 1 tnl of RPMI 1640 with 10%) fetal calf serum) were labelled with 250 p.C\ of "Cr-sodium chromate, 0.2 iViicroenzyme assays ^g/ml, for 1 h at 37 C. Target cells were Supernatant myeioperoxidase, lysowashed 3 titnes in RPMI 1640 with 10% zyme and lactate dehydrogenase activifetal calf serum. ties, used to assess azurophil granule release, total granule release, and PMN lysis, respectively, were analyzed in Reagents and cytokines microplate format (4, 6, 8, 19).
Zymosan A (Sigma) was pre-opsonized with lO'Mi normal human serum, washed Treatment of PBL witii PiWN supernates with PBS, and resuspended in complete medium. Reagents and cytokines were PMN were stimulated by varying conused at the following concentrations in centrations of FMLP, SOZ, and IL-8 cotnplete tnedium: indomethaein (Sig- for 15 min. Then 400 p\ of supernate ma), 10 /ig/ml; thromboxane Bj (TXBj; from each condition was incubated with Sigma), 5-10 //g/ml; IL-2 (Amgen; PBL at a concentration of 2.5 x 10' per Thousand Oaks, CA) 25 U/ml; poly- ml in a total volume of 2 tnl. PBL were clonal rabbit anti-human IFN-y anti- incubated with PMN supernates for serum (Genzyme, Boston, MA), 100 either 24 h and 4 d for the NK assay NU/tnl; n-formylmethionylleucylphen- and LAK assay, respectively. PBL only ylalanitie(FMLP, Sigma), 10-^-10-'M; served as eontrol in NK assay. In the IL-8 (Genzytne), 0.1-10 ng/ml; and LAK assay, IL-2 (25 U/ml) was added serum pre-opsonized zytnosan (SOZ), to PBL in the presence of supernates 0.02-2 mg/ml. The FMLP stock was from PMN stimulated as described initially solubilized in dimethyl sulfox- above. PBL incubated with IL-2 only served as control. After the incubation ide at 10-' M.
period, PBL were washed twice with PBS resuspended in complete medium. Cytotoxicity assay
Target cells (5x10') were mixed with PBL (at ratios of 50:1, 25:1, and 12.5;1) in round-bottotn tnicrotiter plates containing 200 //I of cotnplete medium. The plates were centrifuged at 65 g for 4 tnin to initiate cell-to-cell contact and then incubated for 4 h at 37°C in an hutnidified incubator with 5%o COj. At the end of the reaction, 100 p\ of supematant fiuid was harvested frotn each well and counted for "Cr released from target cells. Each assay was perfortned in triplicate. Cytotoxicity was defttied as the percentage of speeific ^'Cr released or [(experimental release-spontaneous release)/(tnaximal release-spontaneous release)] X 100%. Experimental release was defined as the eounts of "Cr released from target cells caused by effector cells as tneasured by counts per minute. Maxitnal release was the "Cr released frotn target cells induced by 2% Nonidet P-40 detergent. Spontaneous release was the counts of "Cr from target cells incubated in medium alone. Lymphoproiiferation assay
PBL were cultured in tubes for 6 d with IL-2 or IL-2 plus supernatant fluids from PMN at 37 C in a hutnidified incubator with 5% CO2. After the iticubation period, PBL were washed free of PMN supernates. PBL (10' cells/well in 200 /;1 complete mediutn) were seeded in tnicrotiter wells and each well was pulsed with 1 (iCi of [methyl-'H]thymidine (New England Nuclear, Boston, MA) for 5 h under similar incubation and collected onto filter paper with a tnultiple satnple harvester (Catnbridge Technology, Cambridge, MA). Each assay was done in triplicate. Calculations
The cytotoxicity at 3 effector-to-target ratios (50:1, 25:1, 12.5:1) was converted into lytic units (LU 30%) (30). LU 30% arc defined as the nutnber of cells required to cause a 3O'M) target lysis and are expressed per lO*" PBL. The paired /-test (two-tailed) was then applied to determine the significance of LU 30%) values.
Effects of PMN secretion on NK activity Results PiUIN secretion in response to FiUILP and SOZ
of 10-' M for FMLP and 10 ng/tnl for IL-8. In all experitnents, PBL without supernatant fluids frotn PMN served as a control. Additionally, to detertnine any direct effect of SOZ, FMLP and IL-8 on NK activity, these agents were incubated with PBL in eotnplete tnediutn under sitnilar conditions atid NK cytotoxicity retnained unchanged (data not shown).
Under our assay conditions, up to 3%) of azurophil granule cotnponents were released in the presence of FMLP at 10" nM as assessed by supernatant levels of MPO (Fig. 1 A). About 3%) of all granules were released at 10^ nM of FMLP as assessed by lysozyme activity. A higher percentage of azurophil gratiule release relative to the percentage of all Effects of secretions on NK and LAK eell granules released was observed at cytotoxicity 10' nM, suggesting a dissociation bePBL iticubated with unstitnulated PMN tween the two events. SOZ produced supernates exhibited slightly decreased about 4% enzytne release at 0.2 nig/ml NK and LAK cytotoxicity for both sub(Fig. IB). Basal levels of enzytne release jects tested (Table 1). A tnore signifieant occurred iti the absetice of exogetious reduction in NK and LAK cytotoxicity secretagogues (Fig. IA, B). In no inwas observed when PBL were incubated stance was tnore than 0.6%) of cytoplaswith FMLP- and lL-8-stitnulated PMN mic lactate dehydrogenase released. secretions (Table 1). In contrast, SOZstimulated PMN secretion increased NK and LAK eytotoxieity (Table 1). Effects of PMN secretions, using varying doses of secretagogues, on NK cytotoxicity
Fig. 2 shows the effects of secretagogueinduced (PMN) supernatant fluids on subsequent NK activity after 24 h. SOZstitnulated supernates induced a dosedepetident iticrease in cytotoxicity (Fig. 2A). Peak cytotoxicity occurred at a concentration of 0.2 tng/tnl. PMN supernatant stitnulated with FMLP (concentrations ranging frotn 10"' to 10^' M) and lL-8 (concentratiotis ranging from 0.1-10 ng/tnl) decreased cytotoxicity in a dose-dependent tnanner (Fig. 2B, C). The peak decrease in cytotoxicity was observed at a concentration
iVIodulation of iympiioproiiferation by PiUiN secretion
Supertiate collected frotn both stitnulated and unstitnulated PMN were tested for their effect on PBL proliferation induced by IL-2. Cotnpared with the control containing PBL-I-IL-2, only supernate collected from PMN stitnulated by SOZ-HL-2 had an effect on proliferation (Fig. 3). SOZ-stitnulated PMN secretion in eotijunction with IL2 induced a significant increase in 'HTdR uptake. At 6 d, this inciease in proliferation coiticided with a signifi-
—
7 B
ASb
65 4 -
^^.
3-
v; -'.
HJ UJ CC — L zU
o
CC JJ
^'•. ^
m
0 -
10
100 1000
[FMLP], nM
.
I1
1-
1
'
2-
0
0.002 0.02
0.2
91
cant increase in LAK cytotoxicity against M14 target cells (Fig. 3). Effect of anaerobic conditions
All reagents and PMN were equilibrated for 2 h in an anaerobic chatnber (19). After anaerobic equilibration, half the PMN were stimulated in the anaerobic chatnber and the remaining PMN were retnoved frotn the chatnber and stitnulated aerobically using the standard secretagogues. Supernates from PMN were collected and superoxide production was measured. A 94%) redttction in superoxide anion production was observed under anaerobic conditions using PMN suspended at of 6x lO"^ cells/ml (data not showti). Despite this tnarked ditninution in the reduction of dioxygen, anaerobically stitnulated PMN seeretions had the satne effect on the PBL cytotoxicity as their aerobic counterparts (Table 2). Role of PGEj and tfiromboxane in NK suppression
As PMN secrete prostaglandin Ei (PGE2), which is known to inhibit NK activity (14), we tested the possible role of PGEi in the unstitnulated and FMLP- and IL-8-induced cytotoxic inhibition by using indomethacin, which blocks the production of PGEi. Indotnethacin was added to the PMN before stitnulation. Ten pglmX of indotnethacin reversed the suppression in NK cytotoxicity caused by the utistimulated PMN secretions (Table 3). However, the NK inhibition caused by FMLP- and IL-8stimulated PMN secretions was unaffected by indotnethaein. PMN also produce large amounts of TXB, when stitnulated (23). As indotnethacin blocks both PGE, and TXB,, we detertnined whether throtnboxane rather than PGET tnay be responsible for the inhibitive effeet of unstitnulated PMN supernate. TXB,, incubated at a concentration of 5-10 //g/tnl, had no significant effect on NK cytotoxicity. The cytotoxicity of PBL against K562 targets when incubated with TXB, ranged frotn 3.2 + 0.3 to 3.5±0.3 LU 30%), and the cytotoxicity demonstrated by PBL alone was 2.9 ±1.3 LU 30%.
2
[ZYMOSAN], mg/mi
Fig. I. PMN secretion in response to (A) FMLP and (B) sertim-opsonized-zymosan. Values are expressed in terms of percentage of total cell enzyme. Markers tised include myeioperoxidase (MPO) and lysozyme. Bars and vertical lines represent the means and SD of triplicate assay, respectively.
Roie of t F N - / in NK activation
We initiated studies aitned at determitiitig whether the stimulatory effect of SOZ-stimulated PMN secretions on NK
Lala et al.
92
5 -
o CO
5 n
10-
4 3-
O X
3 -
6-
2-
o
2 -
4-
1-
1 -
o .0
None
0.02 [SOZ],
0.2
None
2
10--9
10-8
1-7
None
[FMLP], M
mg/ml
0.1
1
[IL-8],
ng/ml
10
Fig. 2. The effect of (A) SOZ-, (B) FMLP- and (C) IL-8-stimulated PMN secretion on PBL cytotoxicity after 24 h incubation against K562 targets. Points and vertical lines represent the mean and SD of triphcate assay, respectively.
cytotoxicity was direct or indirect. One potential indirect mechanism may involve the stimulation of IFN-y release by T cells or NK cells (31). IFN-y is a potent up-regulator of NK-induced
cytotoxicity (12). PBL alone atid PBL cultured with IFN-y served as controls. NK activity in PBL cultured with IFN)' showed a significant increase in cytotoxicity over PBL alone (PBL = 2.8±
Table I. The effects of PMN secretions on NK and LAK cytotoxicity Cytotoxicity (LU 30%±SD)" Subject 2
Subject 1 Condition PBL only PBL + unstim. sup.'' PBL + SOZ-stim. sup.' PBL + FMLP-stim. sup.' PBL + IL-8-stim. sup.'
NK
LAK
NK
LAK
19.5 + 2.3 15.8 + 0.8 38.8 + 9.6 9.8 + 1.3 9.1+2.1
45.5 + 5.8 39.0 + 4.1 60.3 + 3.2 26.7 + 2.8 22.5 + 3.0
21.5 + 2.4 17.6+1.3 39.7 + 6.8 13.0 + 0.9 10.8 + 2.9
50.8 + 9.8 39.3 + 2.0 62.9 + 9.6 30.0 + 7.9 29.4 + 9.1
"Values represent the mean and SD of these experiments. ''PBL cytotoxicity against K562 (NK targets) and M14 (LAK targets) cells after 1 d (NK) and 4 d (LAK) incubations with a 1:5 dilution of supernates from unstimuiated PMN. PBL were washed free of supernates prior to the assay. "PBL cytotoxicity as above after incubation with supernates from PMN stimulated with SOZ, 0.2 mg/ml; FMLP, 10 ' M; or IL-8, 1 ng/ml.
0.5 LU 30%; PBL + IFN-)' = 8.2±0.3 LU 30'^). This increase in cytotoxicity was completely reversed by the addition of anti-lFN-)' antiserum (100 NU/ml) during culture. Anti-IFN-)'was added to the PBLcultutes incubated with SOZ-indticed PMN secretions. This antibody had no effect on the increased cytotoxicity caused by the SOZ stimulation (PBL + SOZ = 5.1 ±0.9 LU 30%; PBL + SOZ + anti-IFN-7 = 5.1 +0.7 LU 30%). Characterization of SOZ-stimulated PMN secretion
The dialysis (3500 MWCO, against complete medium) of PMN secretions
120 •
'„ 100
T g
Table 2. The effects PMN secretions generated under aerobic and anaerobic conditions on NK cytotoxicity Cytotoxicity (LU30%±SD)'' Conditions
Aerobic
Anaerobic
PBL only PBL + unstim. sup'' PBL + FMLP-stim. sup.' PBL + SOZ-stim. sup.' PBL+IL-8-stim. sup.'
2.8 + 0.5 2.1+0.6 1.7 + 0.2
Not determined 2.3 + 0.7 1.2 + 0.3
5.1+0.9
5.2+1.0
1.3 + 0.4
1.2 + 0.3
"Values represent the mean and SD of 3 experiments. ''PBL cytotoxicity against K562 (NK targets) cells after 1 d incubation with a 1:5 dilution of supernates from unstimulated PMN. PBL were washed free of supernates prior to the assay. TBL cytotoxicity as above after incubation with supernates from PMN stimulated with SOZ, 0.2 mg/ml, FMLP, 10"' M; or IL-8, 1 ng/ml. . •
t
T
T
6040-
g Q,
20 •
Ftg. 3. The effect of differentially stimulated PMN secretion on PBL cytotoxicity and proliferation after 6 d culture with IL-2 against M14 targets. The supernates used were derived from unstimulated PMN and from PMN stimulated with I-MLP, SOZ, and IL8. Bars and vertical lines represent the mean and SD of triplicate assay, respectively. *Significantly different from control at P
