- Email: [email protected]
Cytotoxicity of bovine leucocytes for parainfluenza type-3 virus-infected cells
Veterinary Immunology and lmmunopathology, 31 ( 1992 ) 115-127
115
Elsevier Science Publishers B.V., Amsterdam
Cytotoxicity of bovine leucocytes for parainfluenza type-3 virus-infected cells H.E.L. Bradford, B.M. Adair, M.S. McNulty and J.C. Foster Veterinary Research Laboratories, Stormont, Belfast BT4 3SD, UK (Accepted 8 March 1991 )
ABSTRACT Bradford, H.E.L., Adair, B.M., McNulty, M.S. and Foster, J.C., 1992. Cytotoxicity of bovine leukocytes for parainfluenza type-3 virus-infected cells. Vet. Immunol. Immunopathol., 31: 115-127. The cytotoxic effect of bovine neutrophils, alveolar macrophages, monocytes and lymphocytes for parainfluenza type-3 (PI-3) virus-infected cells in ~lchromium-release assays is described. Specific lysis of virus-infected target cells with PI-3 virus antibody and complement was first observed 8 h after infection coincident with the appearance of haemadsorption-positive cells. Specific lysis increased rapidly reaching a peak 18-24 h after infection. This increase was parallelled by the increase in the percentage of cells with surface haemagglutinin. Target cells were subsequently used in 5tchromium-release assays between 18 and 20 h after virus infection. Antibody-independent killing of PI-3 virus-infected cells was observed with neutrophils, alveolar macrophages and lymphocytes. Levels of specific lysis up to 30% for neutrophils and 68% for alveolar macrophages were observed, although there was considerable variation in activity from animal to animal. Lymphocyte preparations showed levels of cytotoxicity up to 20% in some cases while monocytes had low killing ability. Addition of PI-3 virus-specific antibodies enhanced killing by neutrophils, monocytes and lymphocytes but inhibited killing by alveolar macrophages. Complement, particularly guinea pig complement, was cytotoxic for virus-infected but not for uninfected cells, and also considerably enhanced the cytotoxic effect of neutrophils and lymphocytes. ABBREVIATIONS ADCC, antibody-dependent cell-mediated cytotoxicity; EMEM, Eagle's minimum essential medium; E:T ratio, effector to target ratio; FBS, foetal bovine serum; GBK, Georgia bovine kidney; HI, haemagglutination inhibition; IBR, infectious bovine rhinotracheitis; MOI, multiplicity of infection; PBS, phosphate-buffered saline; PI-3, parainfluenza type-3.
INTRODUCTION
The cytotoxic interaction between leucocytes and virus-infected cells is thought to play an important role in recovery from virus infection. While the mechanisms operating against infectious bovine rhinotracheitis (IBR) virusinfected cells have been extensively investigated (Babiuk et al., 1988 ), little is known of the interactions between lymphoid effectors and other bovine respiratory viruses. © 1992 Elsevier Science Publishers B.V. All rights reserved 0165-2427/92/$05.00
1 16
H.E.L. BRADFORD ET AL.
Parainfluenza type-3 (PI-3) virus, plays an important role in the pathogenesis of pneumonia in young calves (Bryson et al., 1978), and severe lung lesions can be produced by experimental inoculation of PI-3 virus alone into calves (Bryson et al., 1979 ). Following initial multiplication in the upper respiratory tract, the virus multiplies in the lung, growing to high titres in pneumocytes and alveolar macrophages (Hesse and Toth, 1983; Liggitt et al., 1985 ), and by adversely affecting the function of these cells may pave the way for secondary bacterial infection (Lopez et al., 1976 ). Sequential pathological studies in infected calves (Bryson et al., 1979 ) have demonstrated a large influx of mononuclear and polymorphonuclear cells to the site of the virus infection; however, it is not clear what role individual cells play in the removal of virus and virus-infected cells from the animal. As a first step to elucidating the mechanisms involved it is important to know what interactions are possible. This paper investigates the cytotoxic capabilities of bovine neutrophils, alveolar macrophages, monocytes and lymphocytes for PI-3 virusinfected target cells in vitro. MATERIALS AND METHODS
Cells and virus Georgia bovine kidney (GBK) cells were grown in Eagle's minimal essential medium (EMEM) containing 10% foetal bovine serum (FBS) and maintained in medium with 2% FBS. The PI-3 virus used in these experiments was isolated during investigations into the aetiology of outbreaks of pneumonia in young calves (Bryson et al., 1978). The virus (strain 125 ) was grown in GBK cell monolayers in 75 cm 2 plastic flasks (Costar, Northumbria Biologicals, Northumberland, U K ) and virus titres were determined by endpoint titration in GBK cells in 96-well flat-bottomed microtitre plates.
Antiserum and complement The PI-3 virus antiserum used was a convalescent serum from a calf taken 4 weeks after intratracheal inoculation with PI-3 virus. This serum had a titre of 1/ 1024 as determined by haemagglutination inhibition (HI) test using 0.8% guinea pig erythrocytes. The serum was heat-inactivated at 56 °C for 30 min before use in cytotoxic assays. Guinea pig, rabbit and baby rabbit complement were obtained from Sera-Lab, Crawley Down, Sussex, UK.
Animals A batch of ten conventional, cross-bred calves aged between 6 weeks and 6 months were used as donor animals for effector cells. All had antibodies to
CYTOTOXICITY OF BOVINE LEUCOCYTES
117
PI-3 virus. PI-3 HI titres varied from 1/ 16 to 1/ 128 during the course of the experiments.
Haemadsorption GBK cells ( 106) were infected in suspension with PI-3 virus at a MOI of 1. Virus-infected cells were diluted to 105 m l - ~and added to 4-well multidishes (Costar, Northumbria Biologicals, Northumberland, UK) at 105 cells per well. At various times after infection, the medium was removed and the cells washed with 1 ml of warm 0.01 M phosphate-buffered saline (PBS) pH 7.2. This was removed and replaced with 1 ml of 1.0% guinea pig erythrocytes in PBS and dishes were placed at 37 °C for 30 min to allow haemadsorption to take place. The erythrocytes were removed and the cells washed as above with PBS and cultures examined using an inverted phase-contrast microscope. The mean percentage of cells with three or more erythrocytes attached was determined in six microscope fields.
Im munofluorescence GBK cells ( 106 ) were infected with PI-3 virus as above, and 5 × 105 cells were added to each of 2 wells in a 4-well multidish, each of which contained a 9 mm circular coverslip. Twenty hours after infection, coverslips containing infected cells were removed, gently washed in warm PBS and stained without fixation for 40 min at 37 °C with a bovine antiserum to PI-3 virus conjugated with fluorescein isothiocyanate. After gentle washing with warm PBS, coverslips were inverted in PBS onto a microscope slide and examined under ultraviolet light.
Effector cells Alveolar macrophages were obtained by bronchoalveolar lavage. A 0.6 cm diameter polypropylene tube was passed intranasally through the trachea and lodged in a bronchus. Aliquots (50 ml) of 0.9% sodium chloride solution were infused and withdrawn through the tube by means of a 50 ml syringe connected to the end. The recovered lavage fluid was filtered through sterile gauze and centrifuged at 150Xg for 30 rain at 4°C. The cell pellet was washed 3 times by dilution with saline and centrifugation. After the third wash, the cells were suspended in RPMI- 1640 medium containing 200 gg m l - 1 of gentamycin and 2.5 gg m l - 1 of amphotericin B. Heparinised blood was layered over a column of Ficoll-Hypaque (Pharmacia, Milton Keynes, U K ) and centrifuged at 3 0 0 0 × g for 30 min. Mononuclear cells were recovered from the plasma-Ficoll interface, washed twice
l 18
H . E i . B R A D F O R D ET AL.
by dilution in RPMI-1640 m e d i u m without serum and centrifugation, and counted. Monocytes were separated by the m e t h o d of Goddeeris et al. (1986). Briefly, 75 cm 2 plastic cell culture flasks were pre-coated with gelatin by adding 10 ml of a suspension of 2% gelatin (Sigma Chemicals, Poole, Dorset, U K ) and incubated for 2 h at 37°C. The gelatin solution was then removed and the flasks allowed to dry at 37 ° C. Autologous plasma ( 10 ml), recovered from heparinised blood by centrifugation, was added to each flask and incubated for 1 h at 37°C. The plasma was then removed and each flask rinsed with RPMI- 1640 without serum. Mononuclear cell suspension ( 15 ml) containing 5 × 106 cells ml-~ were added to each flask and incubated for 1 h at 37 ° C. Non-adherent cells were removed by pipette and recovered by centrifugation. Adherent cells were detached by adding 10 ml of 10 m M EDTA in calcium- and magnesium-free Hanks' basal salt solution to each flask and incubating for 5 min at room temperature. The cells were removed by pipette and recovered by centrifugation. Neutrophils were recovered from the erythrocyte pellet from Ficoll-Hypaque by lysis of the erythrocytes with 10 ml distilled water for 30 s. Isotonicity was restored by addition of 2.7% saline to one-third of the volume. After centrifugation the cell pellets, which consisted of 98% neutrophils, were washed by dilution in RPMI- 1640 without serum and centrifugation, and the cells counted. 51C h r o m i u m - r e l e a s e assays
Cytotoxicity assays were performed in flat-bottomed 96-well microtitre plates using PI-3 virus-infected GBK cells as targets. GBK cells growing in 25 sq. cm plastic cell culture flasks were infected with PI-3 virus at a MOI of 1. Two millilitres of virus suspension, which was inoculated onto the cells for 1 h at 37°C. Control flasks were mock-infected with EMEM without serum. After virus absorption, the inoculum was removed and the cells washed with EMEM without serum. Sodium 5~chromate ( 5 0 / t C i ) (Amersham International, Amersham, U K ) in 5 ml EMEM without serum were added to each flask and the flasks incubated overnight at 37 ° C. The cell monolayer was then washed 3 times by addition and removal of EMEM without serum to remove excess 51chromium, and the cells trypsinized and counted in a haemocytometer. The cell concentration was adjusted to 105 m l - ~and 100/tl were added to each well of a microtitre plate (i.e. 104 cells per well). Effector cells in 50/~1 vols. were added to each well at various effector to target ( E : T ) cell ratios, e.g. I00: 1, 50: 1, 10: 1, 3 wells for each determination. Heat-inactivated PI-3 virus antiserum (convalescent bovine serum, HI titre 1/1024) a n d / o r complement (50 #1 vols.) diluted in EMEM without serum were added to some assays. Where no antiserum or complement was
CYTOTOX1CITY OF BOVINE LEUCOCYTES
119
added, this was replaced with a 50 gl vol of EMEM without serum. EMEM without serum ( 100 gl) was added to 6 wells for spontaneous release determination and plates were incubated at 37°C in 5% CO2 for 18 h. Supernatants were collected using a multichannel pipette and counted in triplicate in a gamma scintillation counter (Compugamma, Pharmacia, Milton Keynes, U K ) . Means and standard deviations of all counts were determined; standard deviations in all cases were < 5% of the mean. Total release was determined by addition of 100/tl of 3% Triton X-100 to 6 wells containing only target cells, and percent specific lysis calculated using the formula: % specific lysis =
mean CPM in test s a m p l e - m e a n spontaneous release mean total C P M - m e a n spontaneous release
Non-specific esterase stain
Cytospin preparations of mononuclear ceils were fixed with acetone-formaldehyde and stained for non-specific esterase activity by the method of Koski et al. (1976), using alpha-naphthyl acetate (Sigma Chemicals, Poole, Dorset, U K ) as substrate. RESULTS Target cells
The percentage of GBK cells haemadsorbing guinea pig erythrocytes increased rapidly after infection with virus reaching almost 100% after 18-24 h (Fig. 1 ). Specific lysis of target cells with antiserum to PI-3 virus ( 1:50 dilution) and guinea pig complement ( 1 : 5 dilution) began to increase from 8 h after infection rising to 45% by 12 h (Fig. 2). By 24 h after infection 85% specific lysis was observed. From these studies 18-24 h post-infection was
'°°I oo
2o
I
H O U R S POST-INFKCTION
Fig. 1. Haemadsorptionby PI-3 virus-infectedGBK cellsat various times after virus infection.
120
r~
N
H . E i . B R A D F O R D ET AL.
80
20
ob
I
--5 I0 15 20 H O U R S POST-INFECTION
25
Fig. 2. Increase in lysis of PI-3 virus-infected GBK cells with virus-specific antibody ( 1 : 50 dilution ) and guinea pig complement ( 1 : 5 dilution) at different times after virus infection.
Fig. 3. PI-3 virus-infected GBK cells 24 h after virus infection, showing fluorescent virus antigens on the surface of the cells.
chosen as the o p t i m u m time for use of targets. Direct immunofluorescence of live virus-infected GBK cells at this time showed 95-100% of cells with fluorescent viral antigens on the cell membrane (Fig. 3 ). Heat-inactivated PI-3 virus antiserum alone failed to lyse target cells infected for 18 h with PI-3 virus. A n t i b o d y - c o m p l e m e n t lysis of target cells at 8 h was greatest at low antiserum and complement dilutions, and decreased with increasing antis e r u m - c o m p l e m e n t dilution. There was no lysis of uninfected cells even with high concentrations of antiserum and complement.
Neutrophil cytotoxicity Cytotoxicity by neutrophils alone was extremely variable from one animal to another. All the animals used as source of neutrophils for these experi-
CYTOTOXICITY
121
OF BOVINE LEUCOCYTES
~ZO
--0- CALF3 --..&..-CA/,F• ---m- P.i/~
...........-'""'"
~,o I~
& .........
. .......
t
4
,p ....... 50:1 I0:I £FF£CTOIL'TARGET RATIO
o
• ................ 100:I
Fig. 4. Antibody-independent neutrophil cytotoxicity for PI-3 virus-infected target cells. Different symbols show typical results with neutrophils from three different animals. I00
~ 80 80
~
""O'" E:T 10:I l : Y 100:1
2O
o t,h o
, ,/5o
"~. ...................9
t / t oo
1=1-3 ANTISERUM DILUTION
t/~oo
Fig. 5. Antibody-dependent neutrophil cytotoxicity for PI-3 virus-infected target cells. PI-3 antiserum HI titre 1/1024. ments had circulating PI-3 virus antibodies. However, levels of neutrophil cytotoxicity were unrelated to antibody titre. Levels of specific lysis varied from almost zero to 30%. Results with three animals are shown in Fig. 4. It was often the case that lysis was greater at low E: T ratios. Addition of PI-3 virus antiserum enhanced lysis of target cells particularly at high E: T ratios. Typical values are shown in Fig. 5. Levels of specific lysis observed were always m u c h greater than for neutrophils alone and often approached 100%.
Cytotoxicity by mononuclear cells There was variation between animals in the degree of specific lysis observed with alveolar macrophages. Values were often higher at an E : T ratio of 50:1 than at 100:1 and varied from 25 to 68% specific lysis. A typical result is shown in Fig. 6. Addition of PI-3 virus antiserum had the effect of inhibiting the level of cytotoxicity seen in the cells on their own (Fig. 6 ), by up to 60% in some cases. Peripheral blood mononuclear cells adhering to plasma-gelatin-coated flasks were > 95% esterase-positive. This population on their own had low (up to 5% specific lysis) cytotoxic activity. E n h a n c e m e n t of cytotoxicity was ob-
122
H.E.L. BRADFORDET AL. I
I NO P I - 3 S E R U M 4- P 1 - 3
SERUM
60 u3
•..a
40
!2o 0
10:1
50:1
100:1
EFFECTOR .. TARGET RATIO
Fig. 6. Alveolar macrophage cytotoxicity and inhibition by PI-3 virus antiserum. PI-3 antiserum HI titre 1/1024, used at 1 : 10 dilution.
30
2o t...>
i. t.u a. to
10
,,.t
o NONE 1/10 1 / 5 0 1/100 1 / 2 0 0 P I - 3 A N T I S E R U M DILUTION
Fig. 7. Antibody-dependent monocyte cytotoxicity by adherent mononuclear cells from peripheral blood. E: T ratio 100: 1. PI-3 antiserum HI titre 1 / 1024.
40
•" 0 "
E:T tO:f
E:7' tO0:l
0........... o
d,o
,/5o
, roo
S E R U M DILUTION
Fig. 8. A D C C by non-adherent mononuclear cells from peripheral blood. PI-3 antiserum HI titre 1/1024. In the absence o f PI-3 virus-specific antibodies, specific lysis values up to 20% were observed at an E: T ratio of 100: I.
123
CYTOTOX1CITY OF BOVINE LEUCOCYTES
served following addition of PI-3 virus antiserum and specific lysis values up to 25% were observed at high E:T ratios (Fig. 7). The non-adherent cells (lymphocytes) from plasma-gelatin-coated flasks were < 3% esterase-positive. These cells on their own had up to 20% specific lysis at an E: T ratio of 100: 1, although values of 5-10% were more common. Presence of antibody boosted this level to 40%. Typical results are shown in Fig. 8.
Complement cytotoxicity Complement on its own had considerable lytic activity in some experiments. The complement source was found to be important. Guinea pig complement on its own had the highest activity, giving rise to up to 70% lysis of infected targets in some assays. Lower lytic values were observed with rabbit complement and baby rabbit complement gave lowest values of all. The activity was almost totally limited to virus-infected cells and only low lytic values (highest 5%) were observed with uninfected cells. The effect of baby rabbit complement on neutrophil cytotoxicity is shown in Fig. 9. The greatest enhancement of lytic activity was observed at high ~80 ~ 60
-A-
~40
. o
COMPLgM£NT
4LON£ ""0"" JF:T I0:I ---lib E:T I00:1
- ..................... "
7s
I/to
"
t/so
___
I/Too
COMPLEMgNT DILUTION Fig. 9. Effect of complement on neutrophil cytotoxicity.
~ 80 6O
--A- ¢OKPIJ~NT 4LON£ •"0"" £:T 10:1 Jr:T I00:I
20
o 1?s
1,ho
,;50
COMPL~MXNT DILUTION
I/;oo
Fig. 10. Complement enhancement of lymphocyte cytotoxicity.
124
H.E.L. BRADFORD ET AL.
complement dilution and an E: T ratio of 50: 1. Complement also had a considerable enhancing effect on the cytotoxic activity of non-adherent mononuclear cells. A typical result is shown in Fig. 10. Complement alone in this experiment produced 24% specific lysis at the lowest dilution tested ( 1 : 5 ). With non-adherent cells alone 21% lysis at an E:T ratio of 100:1 and 6% at 10:1 was observed. Complement and non-adherent cells together had much higher values. Greatest enhancement was observed at an E: T ratio of 100: 1. With adherent mononuclear cells no enhancing activity was observed, although levels of lysis with complement alone were high (up to 25%) in these experiments. DISCUSSION
The interaction between IBR virus, which is recognised as a major viral agent involved in the 'shipping fever' complex in calves, and bovine lymphoid cells has been extensively investigated (Rouse and Babiuk, 1978; Babiuk et al., 1988 ). In vitro cytotoxic activity for IBR virus-infected cells has been demonstrated using 51chromium_release assays for neutrophils, lymphocytes, and monocytes, and it is assumed that these cells also play similar roles in recovery from IBR virus infection in vivo (Rouse and Babiuk, 1978; Rouse et al., 1977). In Northern Ireland the viruses most commonly associated with respiratory disease outbreaks in young calves are PI-3 and respiratory syncytial viruses (Bryson et al., 1978), and severe disease has been reproduced experimentally with field isolates of both agents (Bryson et al., 1979, 1983). PI-3 virus infection in the lung is associated with a large influx of mononuclear and polymorphonuclear leucocytes (Bryson et al., 1979), yet little is known of the role of these cells in recovery from infection with these viruses. The purpose of this work was to investigate the in vitro interactions between PI-3 virus-infected cells and bovine leucocytes, as a first step to the elucidation of the sequence of events taking place during recovery from PI-3 virus infection in the host animal. In all experiments blood samples and alveolar macrophages were recovered from the calves during the early part of the morning to minimize the possibility of variation due to functional changes occurring during the day. Neutrophil cytotoxicity for PI-3 virus-infected cells has not been described previously, although neutrophils have been shown to be cytotoxic for other virus-infected targets (Rouse et al., 1977; Smith and Sheppard, 1982 ). From the present study it is clear that neutrophils can kill PI-3 virus-infected cells by one of several mechanisms. In the absence of antibody, cytotoxic activity up to 30% was observed. This type of killing (antibody-independent cytotoxicity) has not been described with other bovine respiratory viruses and could be an important means of anti-viral defence, possibly playing a role in the
CYTOTOXICITY OF BOVINE LEUCOCYTES
12 5
period immediately following the appearance of viral antigens on the cell surface, prior to the development of virus-specific antibodies. Antibody-independent killing by neutrophils was described by Lukacs et al. (1985) using 5~chromium-labelled chicken erythrocytes as targets and was shown to be dependent on the state of activation of the neutrophils. In the present study antibody-independent killing by neutrophils was greatest at low E:T ratios. This effect was also observed by Dunkley et al. (1974) in Slchromium release assays using mouse lymphocytes and tumor cell targets, and was thought to be due either to the limited capacity of the culture medium to provide nutrients or oxygen at the highest cell concentrations over the 18 h period of the assay, or to shielding of target cells as a result of physical crowding in the most dense cultures. Antibody-dependent cell-mediated cytotoxicity (ADCC) by neutrophils represents a second mechanism of interaction with PI-3 virus-infected targets. Killing by the antibody-dependent mechanism was always much greater than antibody-independent killing and levels of lysis of almost 100% were seen in some cases. Although not previously reported with PI-3 virus, this mechanism appears to operate with many membrane-bound viruses and is assumed to be an important mechanism in recovery from virus infection (Rouse and Babiuk, 1978 ), although this has yet to be demonstrated in vivo. Neutrophil cytotoxicity was also enhanced in the presence of complement and this represents a further mechanism by which neutrophils interact with, and kill PI-3 virus-infected cells. Although this has not been described previously with PI-3 virus, this finding was not unexpected since glycoproteins expressed on the surface of some virus-infected cells can bind complement components, in some cases eliciting cytotoxicity (Smith et al., 198 l; Bielefeldt Ohmann and Babiuk, 1988 ). In the experiments described here, complement on its own was highly cytotoxic for PI-3 virus-infected cells, and this has also been described for IBR virus (Bielefeldt Ohmann and Babiuk, 1988 ) and RS virus (Kaul et al., 1984). Alveolar macrophages also exhibited antibody-independent killing of PI-3 virus-infected cells (up to 68% lysis), and this probably represents an important defence mechanism in the lung in the early stages following infection, particularly since, as with neutrophils, killing activity by these cells can also be enhanced by exposure to lymphokine preparations such as interferon (LeBlanc, 1989 ). However, unlike neutrophils, alveolar macrophage cytotoxicity was inhibited in the presence of PI-3 virus antibodies. This observation was unexpected, since alveolar macrophages, like neutrophils, express receptors for the Fc portion of the IgG molecule. Since alveolar macrophages are highly permissive for PI-3 virus (Hesse and Toth, 1983 ), it is possible that a significant proportion of the cells become infected during the 18-24 h period during which the macrophages are in contact with the infected target cells. The presence of PI-3 virus antibodies could then inhibit macrophage-target
126
H.E.L. BRADFORD ET AL.
cell interaction by binding to viral antigens expressed on the macrophage membranes. Alternatively, the effect could be brought about through ingestion by the macrophages of virus-antibody complexes which in some cases can inhibit macrophage activity (Astry and Jakab, 1984). Inhibition of alveolar macrophage cytotoxicity by virus-specific antibodies was also reported by others (Probert et al., 1977; Stott et al., 1978 ). Adherent mononuclear cells from peripheral blood had generally low lytic ability (below 5% ) compared to alveolar macrophages and neutrophils. However, presence of PI-3 virus antibodies boosted this activity to around 25%. Non-adherent lymphocytes which had up to 5% monocyte contamination were capable on their own of up to 20% lysis at high E : T cell ratios, and this activity was enhanced considerably following addition of PI-3 antiserum or complement. Campos et al. (1982), also using PI-3 virus-infected targets, demonstrated levels of specific lysis up to 36% using unfractionated peripheral blood mononuclear cell preparations in the absence of PI-3 virus-specific antibodies (natural cytotoxicity). These preparations contained up to 9% phagocytic cells, presumably monocytes, and lytic activity was dramatically reduced after their removal. The purpose of this study was to investigate whether the cytotoxic mechanisms which have been shown to operate in vitro against other virus-infected cells, particularly IBR virus (Rouse and Babiuk, 1978) can also be demonstrated using PI-3 virus-infected targets. In particular, we feel that antibodyindependent killing by neutrophils and macrophages has been greatly underestimated in host-defense against virus infection and could be particularly important following activation by virus-induced lymphokines during the early stages of infection. Complement enhancement of cytotoxic activity could also be important at this time. Studies with individual effector cell preparations are continuing in vivo and in vitro and the results will be reported in due course.
REFERENCES Astry, C.L. and Jakab, G.J., 1984. Influenzavirus induced immune complexessuppressalveolar macrophagephagocytosis.J. Virol., 50: 287-292. Babiuk, L.A., Lawman, M.J.P. and Griebel, P., 1988. Immunosuppression by bovine herpes virus l and other selected herpes viruses. In: S. Spector (Editor), Virus-InducedImmunosuppression. Plenum Press, pp. 14l- 17 I. Bielefeldt Ohmann, H. and Babiuk, L.A., 1988. Induction of receptors for complement and immunoglobulinsby herpes viruses of various species.Virus Res. 9:335-342. Bryson, D.G., McFerran, J.B., Ball, H.J. and Neill, S.D., 1978. Observations on outbreaks of respiratory disease in housed calves ( 1) Epidemiological,clinicaland microbiologicalfindings. Vet. Rec., 103: 485-489. Bryson, D.G., McNulty, M.S., Ball, H.J., Neill, S.D., Connor, T.J. and Cush, P.F., 1979. The
CYTOTOXICITY OF BOVINE LEUCOCYTES
127
experimental reproduction of pneumonia in calves by intranasal inoculation of parainfluenza type 3 virus. Vet. Rec., 105: 566-573. Bryson, D.G., McNulty, M.S., Logan, E.F. and Cush,P.F., 1983. Respiratory syncytial virus pneumonia in young calves: clinical and pathological findings. Am. J. Vet. Res., 44:16481655. Campos, M., Rossi, C.R. and Lawman, M.J.P., 1982. Natural cell-mediated cytotoxicity of bovine mononuclear cells against virus-infected cells. Infect. Immun., 36: 1054-1059. Dunkley, M., Miller, R.G. and Shortman, K., 1974. A modified 5~Crrelease assay for cytotoxic lymphocytes. J. Immunol. Methods, 6:39-51. Goddeeris, B.M., Baldwin, C.L., Ole-Moi Yoi, O. and Morrison, W.I., 1986. Improved methods for purification and depletion of monocytes from bovine peripheral blood mononuclear cells. Functional evaluation of monocytes in responses to lectins. J. Immunol. Methods 89:165173. Hesse, R.A. and Toth, T.E., 1983. Effects of bovine parainfluenza-3 virus on phagocytosis and phagosome-lysosome fusion of cultured bovine alveolar macrophages. Am. J. Vet. Res., 44: 1901-1907. Kaul, T.N., Faden, H., Baker, R. and Ogra, P.L., 1984. Virus induced complement activation and neutrophil mediated cytotoxicity against respiratory syncytial virus (RSV). Clin. Exp. Immunol., 56: 501-508. Koski, I.R., Poplack, D.G. and Blaese, R.M., 1976. A non-specific esterase stain for identification of monocytes and macrophages. In: D.R. Bloom and J.R. David (Editors), "In Vitro" methods in cell mediated and tumour immunity. Academic Press, New York, 1976, pp. 359362. LeBlanc, D.A., 1989. Macrophage activation for cytolysis ofvirally infected target cells. J. Leuk. Biol., 45: 345-352. Liggitt, D., Huston, L., Silflow, R., Evermann, J. and Trigo, E., 1985. Impaired function of bovine alveolar macrophages infected with parainfluenza-3 virus. Am. J. Vet. Res., 46:17401744. Lopez, A., Thompson, R.G. and Savan, M., 1976. The pulmonary clearance of pasteurella hemolytica in calves infected with bovine parainfluenza-3 virus. Can. J. Comp. Med., 40: 385391. Lukacs, K., Roth, J.A. and Kaeberle, M.L., 1985. Activation of neutrophils by antigen-induced lymphokine, with emphasis on antibody-independent cytotoxicity. J. Leuk. Biol., 38: 557572. Probert, M., Stott, E.J. and Thomas, L.H., 1977. Interactions between calf alveolar macrophages and parainfluenza-3 virus. Infect. Immun., 15: 576-585. Rouse, B.T. and Babiuk, L.A., 1978. Mechanisms of recovery from herpes virus infections - - a review. Can. J. Comp. Med., 42: 414-427. Rouse, B.T., Wardley, R.C., Babiuk, L.A. and Mukkur, T.K.S., 1977. The role of neutrophils in anti-viral defence - - "in vitro" studies on the mechanism of antiviral inhibition. J. Immunol., 118: 1957-1961. Smith, J.W. and Sheppard, A.M., 1982. Activity of rabbit monocytes, macrophages and neutrophils in antibody-independent cellular cytotoxicity of herpes simplex virus-infected corneal cells. Infect. Immun., 36: 685-690. Smith, T.F., Mclntosh, K., Fishout, M. and Henson, P.M., 1981. Activation of complement by cells infected with respiratory syncytial virus. Infect. Immun., 33: 43-48. Stott, E.J., Probert, M. and Thomas, L.H., 1978. Cytotoxicity of alveolar macrophages for virusinfected cells. Nature, 255: 710-712.
115
Elsevier Science Publishers B.V., Amsterdam
Cytotoxicity of bovine leucocytes for parainfluenza type-3 virus-infected cells H.E.L. Bradford, B.M. Adair, M.S. McNulty and J.C. Foster Veterinary Research Laboratories, Stormont, Belfast BT4 3SD, UK (Accepted 8 March 1991 )
ABSTRACT Bradford, H.E.L., Adair, B.M., McNulty, M.S. and Foster, J.C., 1992. Cytotoxicity of bovine leukocytes for parainfluenza type-3 virus-infected cells. Vet. Immunol. Immunopathol., 31: 115-127. The cytotoxic effect of bovine neutrophils, alveolar macrophages, monocytes and lymphocytes for parainfluenza type-3 (PI-3) virus-infected cells in ~lchromium-release assays is described. Specific lysis of virus-infected target cells with PI-3 virus antibody and complement was first observed 8 h after infection coincident with the appearance of haemadsorption-positive cells. Specific lysis increased rapidly reaching a peak 18-24 h after infection. This increase was parallelled by the increase in the percentage of cells with surface haemagglutinin. Target cells were subsequently used in 5tchromium-release assays between 18 and 20 h after virus infection. Antibody-independent killing of PI-3 virus-infected cells was observed with neutrophils, alveolar macrophages and lymphocytes. Levels of specific lysis up to 30% for neutrophils and 68% for alveolar macrophages were observed, although there was considerable variation in activity from animal to animal. Lymphocyte preparations showed levels of cytotoxicity up to 20% in some cases while monocytes had low killing ability. Addition of PI-3 virus-specific antibodies enhanced killing by neutrophils, monocytes and lymphocytes but inhibited killing by alveolar macrophages. Complement, particularly guinea pig complement, was cytotoxic for virus-infected but not for uninfected cells, and also considerably enhanced the cytotoxic effect of neutrophils and lymphocytes. ABBREVIATIONS ADCC, antibody-dependent cell-mediated cytotoxicity; EMEM, Eagle's minimum essential medium; E:T ratio, effector to target ratio; FBS, foetal bovine serum; GBK, Georgia bovine kidney; HI, haemagglutination inhibition; IBR, infectious bovine rhinotracheitis; MOI, multiplicity of infection; PBS, phosphate-buffered saline; PI-3, parainfluenza type-3.
INTRODUCTION
The cytotoxic interaction between leucocytes and virus-infected cells is thought to play an important role in recovery from virus infection. While the mechanisms operating against infectious bovine rhinotracheitis (IBR) virusinfected cells have been extensively investigated (Babiuk et al., 1988 ), little is known of the interactions between lymphoid effectors and other bovine respiratory viruses. © 1992 Elsevier Science Publishers B.V. All rights reserved 0165-2427/92/$05.00
1 16
H.E.L. BRADFORD ET AL.
Parainfluenza type-3 (PI-3) virus, plays an important role in the pathogenesis of pneumonia in young calves (Bryson et al., 1978), and severe lung lesions can be produced by experimental inoculation of PI-3 virus alone into calves (Bryson et al., 1979 ). Following initial multiplication in the upper respiratory tract, the virus multiplies in the lung, growing to high titres in pneumocytes and alveolar macrophages (Hesse and Toth, 1983; Liggitt et al., 1985 ), and by adversely affecting the function of these cells may pave the way for secondary bacterial infection (Lopez et al., 1976 ). Sequential pathological studies in infected calves (Bryson et al., 1979 ) have demonstrated a large influx of mononuclear and polymorphonuclear cells to the site of the virus infection; however, it is not clear what role individual cells play in the removal of virus and virus-infected cells from the animal. As a first step to elucidating the mechanisms involved it is important to know what interactions are possible. This paper investigates the cytotoxic capabilities of bovine neutrophils, alveolar macrophages, monocytes and lymphocytes for PI-3 virusinfected target cells in vitro. MATERIALS AND METHODS
Cells and virus Georgia bovine kidney (GBK) cells were grown in Eagle's minimal essential medium (EMEM) containing 10% foetal bovine serum (FBS) and maintained in medium with 2% FBS. The PI-3 virus used in these experiments was isolated during investigations into the aetiology of outbreaks of pneumonia in young calves (Bryson et al., 1978). The virus (strain 125 ) was grown in GBK cell monolayers in 75 cm 2 plastic flasks (Costar, Northumbria Biologicals, Northumberland, U K ) and virus titres were determined by endpoint titration in GBK cells in 96-well flat-bottomed microtitre plates.
Antiserum and complement The PI-3 virus antiserum used was a convalescent serum from a calf taken 4 weeks after intratracheal inoculation with PI-3 virus. This serum had a titre of 1/ 1024 as determined by haemagglutination inhibition (HI) test using 0.8% guinea pig erythrocytes. The serum was heat-inactivated at 56 °C for 30 min before use in cytotoxic assays. Guinea pig, rabbit and baby rabbit complement were obtained from Sera-Lab, Crawley Down, Sussex, UK.
Animals A batch of ten conventional, cross-bred calves aged between 6 weeks and 6 months were used as donor animals for effector cells. All had antibodies to
CYTOTOXICITY OF BOVINE LEUCOCYTES
117
PI-3 virus. PI-3 HI titres varied from 1/ 16 to 1/ 128 during the course of the experiments.
Haemadsorption GBK cells ( 106) were infected in suspension with PI-3 virus at a MOI of 1. Virus-infected cells were diluted to 105 m l - ~and added to 4-well multidishes (Costar, Northumbria Biologicals, Northumberland, UK) at 105 cells per well. At various times after infection, the medium was removed and the cells washed with 1 ml of warm 0.01 M phosphate-buffered saline (PBS) pH 7.2. This was removed and replaced with 1 ml of 1.0% guinea pig erythrocytes in PBS and dishes were placed at 37 °C for 30 min to allow haemadsorption to take place. The erythrocytes were removed and the cells washed as above with PBS and cultures examined using an inverted phase-contrast microscope. The mean percentage of cells with three or more erythrocytes attached was determined in six microscope fields.
Im munofluorescence GBK cells ( 106 ) were infected with PI-3 virus as above, and 5 × 105 cells were added to each of 2 wells in a 4-well multidish, each of which contained a 9 mm circular coverslip. Twenty hours after infection, coverslips containing infected cells were removed, gently washed in warm PBS and stained without fixation for 40 min at 37 °C with a bovine antiserum to PI-3 virus conjugated with fluorescein isothiocyanate. After gentle washing with warm PBS, coverslips were inverted in PBS onto a microscope slide and examined under ultraviolet light.
Effector cells Alveolar macrophages were obtained by bronchoalveolar lavage. A 0.6 cm diameter polypropylene tube was passed intranasally through the trachea and lodged in a bronchus. Aliquots (50 ml) of 0.9% sodium chloride solution were infused and withdrawn through the tube by means of a 50 ml syringe connected to the end. The recovered lavage fluid was filtered through sterile gauze and centrifuged at 150Xg for 30 rain at 4°C. The cell pellet was washed 3 times by dilution with saline and centrifugation. After the third wash, the cells were suspended in RPMI- 1640 medium containing 200 gg m l - 1 of gentamycin and 2.5 gg m l - 1 of amphotericin B. Heparinised blood was layered over a column of Ficoll-Hypaque (Pharmacia, Milton Keynes, U K ) and centrifuged at 3 0 0 0 × g for 30 min. Mononuclear cells were recovered from the plasma-Ficoll interface, washed twice
l 18
H . E i . B R A D F O R D ET AL.
by dilution in RPMI-1640 m e d i u m without serum and centrifugation, and counted. Monocytes were separated by the m e t h o d of Goddeeris et al. (1986). Briefly, 75 cm 2 plastic cell culture flasks were pre-coated with gelatin by adding 10 ml of a suspension of 2% gelatin (Sigma Chemicals, Poole, Dorset, U K ) and incubated for 2 h at 37°C. The gelatin solution was then removed and the flasks allowed to dry at 37 ° C. Autologous plasma ( 10 ml), recovered from heparinised blood by centrifugation, was added to each flask and incubated for 1 h at 37°C. The plasma was then removed and each flask rinsed with RPMI- 1640 without serum. Mononuclear cell suspension ( 15 ml) containing 5 × 106 cells ml-~ were added to each flask and incubated for 1 h at 37 ° C. Non-adherent cells were removed by pipette and recovered by centrifugation. Adherent cells were detached by adding 10 ml of 10 m M EDTA in calcium- and magnesium-free Hanks' basal salt solution to each flask and incubating for 5 min at room temperature. The cells were removed by pipette and recovered by centrifugation. Neutrophils were recovered from the erythrocyte pellet from Ficoll-Hypaque by lysis of the erythrocytes with 10 ml distilled water for 30 s. Isotonicity was restored by addition of 2.7% saline to one-third of the volume. After centrifugation the cell pellets, which consisted of 98% neutrophils, were washed by dilution in RPMI- 1640 without serum and centrifugation, and the cells counted. 51C h r o m i u m - r e l e a s e assays
Cytotoxicity assays were performed in flat-bottomed 96-well microtitre plates using PI-3 virus-infected GBK cells as targets. GBK cells growing in 25 sq. cm plastic cell culture flasks were infected with PI-3 virus at a MOI of 1. Two millilitres of virus suspension, which was inoculated onto the cells for 1 h at 37°C. Control flasks were mock-infected with EMEM without serum. After virus absorption, the inoculum was removed and the cells washed with EMEM without serum. Sodium 5~chromate ( 5 0 / t C i ) (Amersham International, Amersham, U K ) in 5 ml EMEM without serum were added to each flask and the flasks incubated overnight at 37 ° C. The cell monolayer was then washed 3 times by addition and removal of EMEM without serum to remove excess 51chromium, and the cells trypsinized and counted in a haemocytometer. The cell concentration was adjusted to 105 m l - ~and 100/tl were added to each well of a microtitre plate (i.e. 104 cells per well). Effector cells in 50/~1 vols. were added to each well at various effector to target ( E : T ) cell ratios, e.g. I00: 1, 50: 1, 10: 1, 3 wells for each determination. Heat-inactivated PI-3 virus antiserum (convalescent bovine serum, HI titre 1/1024) a n d / o r complement (50 #1 vols.) diluted in EMEM without serum were added to some assays. Where no antiserum or complement was
CYTOTOX1CITY OF BOVINE LEUCOCYTES
119
added, this was replaced with a 50 gl vol of EMEM without serum. EMEM without serum ( 100 gl) was added to 6 wells for spontaneous release determination and plates were incubated at 37°C in 5% CO2 for 18 h. Supernatants were collected using a multichannel pipette and counted in triplicate in a gamma scintillation counter (Compugamma, Pharmacia, Milton Keynes, U K ) . Means and standard deviations of all counts were determined; standard deviations in all cases were < 5% of the mean. Total release was determined by addition of 100/tl of 3% Triton X-100 to 6 wells containing only target cells, and percent specific lysis calculated using the formula: % specific lysis =
mean CPM in test s a m p l e - m e a n spontaneous release mean total C P M - m e a n spontaneous release
Non-specific esterase stain
Cytospin preparations of mononuclear ceils were fixed with acetone-formaldehyde and stained for non-specific esterase activity by the method of Koski et al. (1976), using alpha-naphthyl acetate (Sigma Chemicals, Poole, Dorset, U K ) as substrate. RESULTS Target cells
The percentage of GBK cells haemadsorbing guinea pig erythrocytes increased rapidly after infection with virus reaching almost 100% after 18-24 h (Fig. 1 ). Specific lysis of target cells with antiserum to PI-3 virus ( 1:50 dilution) and guinea pig complement ( 1 : 5 dilution) began to increase from 8 h after infection rising to 45% by 12 h (Fig. 2). By 24 h after infection 85% specific lysis was observed. From these studies 18-24 h post-infection was
'°°I oo
2o
I
H O U R S POST-INFKCTION
Fig. 1. Haemadsorptionby PI-3 virus-infectedGBK cellsat various times after virus infection.
120
r~
N
H . E i . B R A D F O R D ET AL.
80
20
ob
I
--5 I0 15 20 H O U R S POST-INFECTION
25
Fig. 2. Increase in lysis of PI-3 virus-infected GBK cells with virus-specific antibody ( 1 : 50 dilution ) and guinea pig complement ( 1 : 5 dilution) at different times after virus infection.
Fig. 3. PI-3 virus-infected GBK cells 24 h after virus infection, showing fluorescent virus antigens on the surface of the cells.
chosen as the o p t i m u m time for use of targets. Direct immunofluorescence of live virus-infected GBK cells at this time showed 95-100% of cells with fluorescent viral antigens on the cell membrane (Fig. 3 ). Heat-inactivated PI-3 virus antiserum alone failed to lyse target cells infected for 18 h with PI-3 virus. A n t i b o d y - c o m p l e m e n t lysis of target cells at 8 h was greatest at low antiserum and complement dilutions, and decreased with increasing antis e r u m - c o m p l e m e n t dilution. There was no lysis of uninfected cells even with high concentrations of antiserum and complement.
Neutrophil cytotoxicity Cytotoxicity by neutrophils alone was extremely variable from one animal to another. All the animals used as source of neutrophils for these experi-
CYTOTOXICITY
121
OF BOVINE LEUCOCYTES
~ZO
--0- CALF3 --..&..-CA/,F• ---m- P.i/~
...........-'""'"
~,o I~
& .........
. .......
t
4
,p ....... 50:1 I0:I £FF£CTOIL'TARGET RATIO
o
• ................ 100:I
Fig. 4. Antibody-independent neutrophil cytotoxicity for PI-3 virus-infected target cells. Different symbols show typical results with neutrophils from three different animals. I00
~ 80 80
~
""O'" E:T 10:I l : Y 100:1
2O
o t,h o
, ,/5o
"~. ...................9
t / t oo
1=1-3 ANTISERUM DILUTION
t/~oo
Fig. 5. Antibody-dependent neutrophil cytotoxicity for PI-3 virus-infected target cells. PI-3 antiserum HI titre 1/1024. ments had circulating PI-3 virus antibodies. However, levels of neutrophil cytotoxicity were unrelated to antibody titre. Levels of specific lysis varied from almost zero to 30%. Results with three animals are shown in Fig. 4. It was often the case that lysis was greater at low E: T ratios. Addition of PI-3 virus antiserum enhanced lysis of target cells particularly at high E: T ratios. Typical values are shown in Fig. 5. Levels of specific lysis observed were always m u c h greater than for neutrophils alone and often approached 100%.
Cytotoxicity by mononuclear cells There was variation between animals in the degree of specific lysis observed with alveolar macrophages. Values were often higher at an E : T ratio of 50:1 than at 100:1 and varied from 25 to 68% specific lysis. A typical result is shown in Fig. 6. Addition of PI-3 virus antiserum had the effect of inhibiting the level of cytotoxicity seen in the cells on their own (Fig. 6 ), by up to 60% in some cases. Peripheral blood mononuclear cells adhering to plasma-gelatin-coated flasks were > 95% esterase-positive. This population on their own had low (up to 5% specific lysis) cytotoxic activity. E n h a n c e m e n t of cytotoxicity was ob-
122
H.E.L. BRADFORDET AL. I
I NO P I - 3 S E R U M 4- P 1 - 3
SERUM
60 u3
•..a
40
!2o 0
10:1
50:1
100:1
EFFECTOR .. TARGET RATIO
Fig. 6. Alveolar macrophage cytotoxicity and inhibition by PI-3 virus antiserum. PI-3 antiserum HI titre 1/1024, used at 1 : 10 dilution.
30
2o t...>
i. t.u a. to
10
,,.t
o NONE 1/10 1 / 5 0 1/100 1 / 2 0 0 P I - 3 A N T I S E R U M DILUTION
Fig. 7. Antibody-dependent monocyte cytotoxicity by adherent mononuclear cells from peripheral blood. E: T ratio 100: 1. PI-3 antiserum HI titre 1 / 1024.
40
•" 0 "
E:T tO:f
E:7' tO0:l
0........... o
d,o
,/5o
, roo
S E R U M DILUTION
Fig. 8. A D C C by non-adherent mononuclear cells from peripheral blood. PI-3 antiserum HI titre 1/1024. In the absence o f PI-3 virus-specific antibodies, specific lysis values up to 20% were observed at an E: T ratio of 100: I.
123
CYTOTOX1CITY OF BOVINE LEUCOCYTES
served following addition of PI-3 virus antiserum and specific lysis values up to 25% were observed at high E:T ratios (Fig. 7). The non-adherent cells (lymphocytes) from plasma-gelatin-coated flasks were < 3% esterase-positive. These cells on their own had up to 20% specific lysis at an E: T ratio of 100: 1, although values of 5-10% were more common. Presence of antibody boosted this level to 40%. Typical results are shown in Fig. 8.
Complement cytotoxicity Complement on its own had considerable lytic activity in some experiments. The complement source was found to be important. Guinea pig complement on its own had the highest activity, giving rise to up to 70% lysis of infected targets in some assays. Lower lytic values were observed with rabbit complement and baby rabbit complement gave lowest values of all. The activity was almost totally limited to virus-infected cells and only low lytic values (highest 5%) were observed with uninfected cells. The effect of baby rabbit complement on neutrophil cytotoxicity is shown in Fig. 9. The greatest enhancement of lytic activity was observed at high ~80 ~ 60
-A-
~40
. o
COMPLgM£NT
4LON£ ""0"" JF:T I0:I ---lib E:T I00:1
- ..................... "
7s
I/to
"
t/so
___
I/Too
COMPLEMgNT DILUTION Fig. 9. Effect of complement on neutrophil cytotoxicity.
~ 80 6O
--A- ¢OKPIJ~NT 4LON£ •"0"" £:T 10:1 Jr:T I00:I
20
o 1?s
1,ho
,;50
COMPL~MXNT DILUTION
I/;oo
Fig. 10. Complement enhancement of lymphocyte cytotoxicity.
124
H.E.L. BRADFORD ET AL.
complement dilution and an E: T ratio of 50: 1. Complement also had a considerable enhancing effect on the cytotoxic activity of non-adherent mononuclear cells. A typical result is shown in Fig. 10. Complement alone in this experiment produced 24% specific lysis at the lowest dilution tested ( 1 : 5 ). With non-adherent cells alone 21% lysis at an E:T ratio of 100:1 and 6% at 10:1 was observed. Complement and non-adherent cells together had much higher values. Greatest enhancement was observed at an E: T ratio of 100: 1. With adherent mononuclear cells no enhancing activity was observed, although levels of lysis with complement alone were high (up to 25%) in these experiments. DISCUSSION
The interaction between IBR virus, which is recognised as a major viral agent involved in the 'shipping fever' complex in calves, and bovine lymphoid cells has been extensively investigated (Rouse and Babiuk, 1978; Babiuk et al., 1988 ). In vitro cytotoxic activity for IBR virus-infected cells has been demonstrated using 51chromium_release assays for neutrophils, lymphocytes, and monocytes, and it is assumed that these cells also play similar roles in recovery from IBR virus infection in vivo (Rouse and Babiuk, 1978; Rouse et al., 1977). In Northern Ireland the viruses most commonly associated with respiratory disease outbreaks in young calves are PI-3 and respiratory syncytial viruses (Bryson et al., 1978), and severe disease has been reproduced experimentally with field isolates of both agents (Bryson et al., 1979, 1983). PI-3 virus infection in the lung is associated with a large influx of mononuclear and polymorphonuclear leucocytes (Bryson et al., 1979), yet little is known of the role of these cells in recovery from infection with these viruses. The purpose of this work was to investigate the in vitro interactions between PI-3 virus-infected cells and bovine leucocytes, as a first step to the elucidation of the sequence of events taking place during recovery from PI-3 virus infection in the host animal. In all experiments blood samples and alveolar macrophages were recovered from the calves during the early part of the morning to minimize the possibility of variation due to functional changes occurring during the day. Neutrophil cytotoxicity for PI-3 virus-infected cells has not been described previously, although neutrophils have been shown to be cytotoxic for other virus-infected targets (Rouse et al., 1977; Smith and Sheppard, 1982 ). From the present study it is clear that neutrophils can kill PI-3 virus-infected cells by one of several mechanisms. In the absence of antibody, cytotoxic activity up to 30% was observed. This type of killing (antibody-independent cytotoxicity) has not been described with other bovine respiratory viruses and could be an important means of anti-viral defence, possibly playing a role in the
CYTOTOXICITY OF BOVINE LEUCOCYTES
12 5
period immediately following the appearance of viral antigens on the cell surface, prior to the development of virus-specific antibodies. Antibody-independent killing by neutrophils was described by Lukacs et al. (1985) using 5~chromium-labelled chicken erythrocytes as targets and was shown to be dependent on the state of activation of the neutrophils. In the present study antibody-independent killing by neutrophils was greatest at low E:T ratios. This effect was also observed by Dunkley et al. (1974) in Slchromium release assays using mouse lymphocytes and tumor cell targets, and was thought to be due either to the limited capacity of the culture medium to provide nutrients or oxygen at the highest cell concentrations over the 18 h period of the assay, or to shielding of target cells as a result of physical crowding in the most dense cultures. Antibody-dependent cell-mediated cytotoxicity (ADCC) by neutrophils represents a second mechanism of interaction with PI-3 virus-infected targets. Killing by the antibody-dependent mechanism was always much greater than antibody-independent killing and levels of lysis of almost 100% were seen in some cases. Although not previously reported with PI-3 virus, this mechanism appears to operate with many membrane-bound viruses and is assumed to be an important mechanism in recovery from virus infection (Rouse and Babiuk, 1978 ), although this has yet to be demonstrated in vivo. Neutrophil cytotoxicity was also enhanced in the presence of complement and this represents a further mechanism by which neutrophils interact with, and kill PI-3 virus-infected cells. Although this has not been described previously with PI-3 virus, this finding was not unexpected since glycoproteins expressed on the surface of some virus-infected cells can bind complement components, in some cases eliciting cytotoxicity (Smith et al., 198 l; Bielefeldt Ohmann and Babiuk, 1988 ). In the experiments described here, complement on its own was highly cytotoxic for PI-3 virus-infected cells, and this has also been described for IBR virus (Bielefeldt Ohmann and Babiuk, 1988 ) and RS virus (Kaul et al., 1984). Alveolar macrophages also exhibited antibody-independent killing of PI-3 virus-infected cells (up to 68% lysis), and this probably represents an important defence mechanism in the lung in the early stages following infection, particularly since, as with neutrophils, killing activity by these cells can also be enhanced by exposure to lymphokine preparations such as interferon (LeBlanc, 1989 ). However, unlike neutrophils, alveolar macrophage cytotoxicity was inhibited in the presence of PI-3 virus antibodies. This observation was unexpected, since alveolar macrophages, like neutrophils, express receptors for the Fc portion of the IgG molecule. Since alveolar macrophages are highly permissive for PI-3 virus (Hesse and Toth, 1983 ), it is possible that a significant proportion of the cells become infected during the 18-24 h period during which the macrophages are in contact with the infected target cells. The presence of PI-3 virus antibodies could then inhibit macrophage-target
126
H.E.L. BRADFORD ET AL.
cell interaction by binding to viral antigens expressed on the macrophage membranes. Alternatively, the effect could be brought about through ingestion by the macrophages of virus-antibody complexes which in some cases can inhibit macrophage activity (Astry and Jakab, 1984). Inhibition of alveolar macrophage cytotoxicity by virus-specific antibodies was also reported by others (Probert et al., 1977; Stott et al., 1978 ). Adherent mononuclear cells from peripheral blood had generally low lytic ability (below 5% ) compared to alveolar macrophages and neutrophils. However, presence of PI-3 virus antibodies boosted this activity to around 25%. Non-adherent lymphocytes which had up to 5% monocyte contamination were capable on their own of up to 20% lysis at high E : T cell ratios, and this activity was enhanced considerably following addition of PI-3 antiserum or complement. Campos et al. (1982), also using PI-3 virus-infected targets, demonstrated levels of specific lysis up to 36% using unfractionated peripheral blood mononuclear cell preparations in the absence of PI-3 virus-specific antibodies (natural cytotoxicity). These preparations contained up to 9% phagocytic cells, presumably monocytes, and lytic activity was dramatically reduced after their removal. The purpose of this study was to investigate whether the cytotoxic mechanisms which have been shown to operate in vitro against other virus-infected cells, particularly IBR virus (Rouse and Babiuk, 1978) can also be demonstrated using PI-3 virus-infected targets. In particular, we feel that antibodyindependent killing by neutrophils and macrophages has been greatly underestimated in host-defense against virus infection and could be particularly important following activation by virus-induced lymphokines during the early stages of infection. Complement enhancement of cytotoxic activity could also be important at this time. Studies with individual effector cell preparations are continuing in vivo and in vitro and the results will be reported in due course.
REFERENCES Astry, C.L. and Jakab, G.J., 1984. Influenzavirus induced immune complexessuppressalveolar macrophagephagocytosis.J. Virol., 50: 287-292. Babiuk, L.A., Lawman, M.J.P. and Griebel, P., 1988. Immunosuppression by bovine herpes virus l and other selected herpes viruses. In: S. Spector (Editor), Virus-InducedImmunosuppression. Plenum Press, pp. 14l- 17 I. Bielefeldt Ohmann, H. and Babiuk, L.A., 1988. Induction of receptors for complement and immunoglobulinsby herpes viruses of various species.Virus Res. 9:335-342. Bryson, D.G., McFerran, J.B., Ball, H.J. and Neill, S.D., 1978. Observations on outbreaks of respiratory disease in housed calves ( 1) Epidemiological,clinicaland microbiologicalfindings. Vet. Rec., 103: 485-489. Bryson, D.G., McNulty, M.S., Ball, H.J., Neill, S.D., Connor, T.J. and Cush, P.F., 1979. The
CYTOTOXICITY OF BOVINE LEUCOCYTES
127
experimental reproduction of pneumonia in calves by intranasal inoculation of parainfluenza type 3 virus. Vet. Rec., 105: 566-573. Bryson, D.G., McNulty, M.S., Logan, E.F. and Cush,P.F., 1983. Respiratory syncytial virus pneumonia in young calves: clinical and pathological findings. Am. J. Vet. Res., 44:16481655. Campos, M., Rossi, C.R. and Lawman, M.J.P., 1982. Natural cell-mediated cytotoxicity of bovine mononuclear cells against virus-infected cells. Infect. Immun., 36: 1054-1059. Dunkley, M., Miller, R.G. and Shortman, K., 1974. A modified 5~Crrelease assay for cytotoxic lymphocytes. J. Immunol. Methods, 6:39-51. Goddeeris, B.M., Baldwin, C.L., Ole-Moi Yoi, O. and Morrison, W.I., 1986. Improved methods for purification and depletion of monocytes from bovine peripheral blood mononuclear cells. Functional evaluation of monocytes in responses to lectins. J. Immunol. Methods 89:165173. Hesse, R.A. and Toth, T.E., 1983. Effects of bovine parainfluenza-3 virus on phagocytosis and phagosome-lysosome fusion of cultured bovine alveolar macrophages. Am. J. Vet. Res., 44: 1901-1907. Kaul, T.N., Faden, H., Baker, R. and Ogra, P.L., 1984. Virus induced complement activation and neutrophil mediated cytotoxicity against respiratory syncytial virus (RSV). Clin. Exp. Immunol., 56: 501-508. Koski, I.R., Poplack, D.G. and Blaese, R.M., 1976. A non-specific esterase stain for identification of monocytes and macrophages. In: D.R. Bloom and J.R. David (Editors), "In Vitro" methods in cell mediated and tumour immunity. Academic Press, New York, 1976, pp. 359362. LeBlanc, D.A., 1989. Macrophage activation for cytolysis ofvirally infected target cells. J. Leuk. Biol., 45: 345-352. Liggitt, D., Huston, L., Silflow, R., Evermann, J. and Trigo, E., 1985. Impaired function of bovine alveolar macrophages infected with parainfluenza-3 virus. Am. J. Vet. Res., 46:17401744. Lopez, A., Thompson, R.G. and Savan, M., 1976. The pulmonary clearance of pasteurella hemolytica in calves infected with bovine parainfluenza-3 virus. Can. J. Comp. Med., 40: 385391. Lukacs, K., Roth, J.A. and Kaeberle, M.L., 1985. Activation of neutrophils by antigen-induced lymphokine, with emphasis on antibody-independent cytotoxicity. J. Leuk. Biol., 38: 557572. Probert, M., Stott, E.J. and Thomas, L.H., 1977. Interactions between calf alveolar macrophages and parainfluenza-3 virus. Infect. Immun., 15: 576-585. Rouse, B.T. and Babiuk, L.A., 1978. Mechanisms of recovery from herpes virus infections - - a review. Can. J. Comp. Med., 42: 414-427. Rouse, B.T., Wardley, R.C., Babiuk, L.A. and Mukkur, T.K.S., 1977. The role of neutrophils in anti-viral defence - - "in vitro" studies on the mechanism of antiviral inhibition. J. Immunol., 118: 1957-1961. Smith, J.W. and Sheppard, A.M., 1982. Activity of rabbit monocytes, macrophages and neutrophils in antibody-independent cellular cytotoxicity of herpes simplex virus-infected corneal cells. Infect. Immun., 36: 685-690. Smith, T.F., Mclntosh, K., Fishout, M. and Henson, P.M., 1981. Activation of complement by cells infected with respiratory syncytial virus. Infect. Immun., 33: 43-48. Stott, E.J., Probert, M. and Thomas, L.H., 1978. Cytotoxicity of alveolar macrophages for virusinfected cells. Nature, 255: 710-712.