The phenotype and phagocytic activity of macrophages during maedi-visna virus infection

Veterinary immunokqy and Veterinary

ELSEVIER

immunopathology

Immunology and lmmunopathology 51 (1996) 113-126

The phenotype and phagocytic activity of macrophages during maedi-visna virus infection W.C. Lee ‘, P. Bird 2, I. McConnell 3, N.J. Watt, B.A. Blacklaws 3.* Deparhnent of Veterinary Pathology, University of Edinburgh, Summerball, Edinburgh, EH9 IQH. UK Accepted

10 July 1995

Abstract Macrophages from maedi-visna virus (MVV) infected sheep have been shown to have an activated phenotype from sites of lesions in vivo. Here we have looked at the direct effect of virus infection on macrophage phenotype and activity in vitro by flow cytometry. There was no significant difference in the expression of several surface markers (CD4, CD8, MHC Class I, MHC Class II, lymphocyte function associated antigen(LFA)-1 and LFA-3) on monocyte-derived macrophages (MDM) by 5 days post MVV infection. In contrast the phagocytic activity of MVV-infected MDM for the yeast Candida utilis and erythrocytes was decreased by 5 days p.i. although the surface binding of erythrocytes was not affected. Interestingly, an activated phenotype was seen on alveolar macrophages (AM) from sheep with maedi &n-face expression of MHC Class I, Class II and LFA-1 was increased), but there was no difference in the binding and phagocytosis of erythrocytes by these cells. However the binding and phagocytosis of the bacterium, Pasteurella hemolyticu was increased with AM from MVV-infected sheep without lesions. Similarly there was no significant difference in the phagocytic and erythrocyte rosetting activity between fresh monocytes from MVV-infected and uninfected control sheep. Therefore the phenotype of macrophages taken from sites of lesions caused by MVV does not correspond to a direct effect by the virus on these cells or to particular activities of the macrophages. Keywords: Maedi-visna virus; Macrophage phenotype; Macrophage phagocytosis;

* Corresponding author. ’ Department of Veterinary Medicine, National Chung Hsing University,

Sheep

250 Kuo Kuang Road, Taichung, Taiwan, ROC. * Therapeutic Antibody Centre, Sir William DUM School of Pathology, University of Oxford, South Parks Road, Oxford, OX1 3RE, UK. 3 Current address: Department of Clinical Veterinary Medicine, University of Cambridge, Madingley Road, Cambridge, CB3 OES, UK.

0165.2427/%/$15.00

0 1996 Elsevier Science B.V. All rights reserved

SSDI 0165-2427(95)05508-8

1. Abbreviations

AM, alveolar macrophages; FSC, forward scatter; LFA, lymphocyte function associated antigen; MDM, monocyte-derived macrophages; MF, mean fluorescence intensity; MVV, maedi-visna virus; SSC, side scatter.

2. Introduction

Maedi-visna virus (MVV), a lentivirus, causes a persistent infection of sheep. It replicates slowly in the host so that most infections are subclinical but there is development of progressive inflammatory lesions in many organs (Narayan and Clements, 1989). Although MVV does not cause the gross immunodeficiency seen with HIV or SIV, all the lentiviruses infect monocyte/ macrophages which play a central role in viral persistence (Gendelman et al., 1989). The surface molecule expression of macrophages is often used as an important indicator of their functional activity. It has been shown that MHC Class II expressed on the surface of alveolar (Lujan et al., 1993) and synovial fluid (Harkiss et al.. 1991) macrophages from MVV-infected sheep is significantly increased compared with uninfected sheep. In addition a significant increase in spontaneous fibronectin release and neutrophil chemotactic factor by alveolar macrophages (AM) from sheep affected by MVV interstitial pneumonitis has also been described (Cordier et al., 1990) suggesting that macrophages are activated at the sites of lesions caused by MVV. However, concurrent bacterial infections are often found in field studies in MVV-infected sheep with interstitial pneumonia (Markson et al., 1983) and a decrease in the delayed type hypersensitivity response to purified protein derivative has been reported in MVV-infected sheep (Myer et al., 19881, both suggesting that some macrophage functions may be impaired during MVV infection in vivo. There is therefore a clear need to understand the direct effects of MVV infection on macrophage function. The effect of infection by another lentivirus, HIV, on macrophages has been studied in more depth, both in vivo and in vitro. In contrast to MVV-infected sheep, most macrophage populations studied during AIDS show decreased surface antigen expression (Roy et al., 1987; Bray et al., 1993) and reduced functional abilities (Smith et al., 1984; Wahl et al., 1989). These defects are often attributed to the depletion of CD4+ T cells caused by this virus (Bowen et al., 1986) and indeed when the monocytic cell line U937 is infected in vitro with HIV, most surface antigens show increased expression apart from MHC Class II antigens (Petit et al., 1987). We have therefore undertaken a study of the changes in surface molecule expression and phagocytic activity of cultured monocyte derived macrophages (MDM) infected with MVV in vitro and tried to relate this to similarly analyzed monocytes and AM taken directly from MVV-infected sheep. In this way we can show the direct effects of the virus on an important macrophage function (phagocytosis) and whether the changes seen in vivo are due to the virus alone.

W.C. Lee et al. / Veterinary

Immutwlogy

and Immunopathology

51 (1996) 113-126

115

3. Materials and methods 3.1. Monocyte/macrophage

culture

Peripheral blood mononuclear cells (PBMC) were obtained from heparinized blood by taking buffy coat cells and separating them over Ficoll-Hypaque (Blacklaws et al., 1994). PBMC were harvested at the interface, washed three times (500 X g, 5 min) with phosphate-buffered saline (PBS) and then finally suspended in RPMI-1640 medium (Gibco) containing 2 mM r_-glutamine, 20 mM HEPES, 100 U benzyl penicillin per ml, 100 pg streptomycin per ml, and 5 X 10m5 M 2-mercaptoethanol (RPMI) with 10% FCS and 10% lamb serum (LS) (RPMI/lO% FCS/lO% LS). Monocytes were obtained by adhering PBMC (2.5 X lo6 per well of 24-well plates or 2.5 X 10’ per 25 cm2 flask) in gelatin coated tissue culture plasticware (Jones et al., 1989) for 1 h or overnight at 37°C 5% CO,. Non-adherent cells were washed off with warm RPMI/2% FCS and fresh RPMI/lO% FCS/lO% LS was added to monocytes for continuous culture and replaced every 5 days. After 5-7 days culture, over 95% of these adherent cells were non-specific esterase positive (MDM). To harvest MDM, cells were treated with 5 mM EDTA in PBS for 2-5 min until cells were in suspension. AM were collected from freshly excised lungs by bronchoalveolar lavage (BAL) with cold Hank’s buffered saline. The cells were pelleted by centrifugation at 700 X g, 4°C for lo-20 min, washed with RPMI/2% FCS three times and then used. 3.2. Virus growth and infection Skin cell monolayers were prepared and infected as in Bird et al. (1993) with MVV strain EVl (Sargan et al., 1991) at 0.1 TCID,, per cell in Dulbecco’s modified Eagles medium with 2 mM L-glutamine and 2% FCS. When extensive cytopathic effects were seen, the virus containing cell supematant was harvested, clarified and stored at - 70°C. MDM were cultured for 5-7 days and then infected with EVl at 1 TCID,, per cell, in RPMI/2% FCS for the specified times. 3.3. Indirect immunojluorescent

staining for flow cytometry

A panel of monoclonal antibodies which included SBU-T4 (anti-CD4), SBU-T8 (anti-CD8) (Maddox et al., 1985), VPM8 (anti-Ig-L chain), VPM19 (anti-MI-K Class I) (Hopkins and Dutia, 1990), VPM36 (anti-MHC Class II DQ), VPM38 (anti-MHC Class II DR) (Dutia et al., 1990), VPM65 (anti-CD14) (Gupta et al., 1993) FlO-150-39 (anti-CD1 la, i.e. lymphocyte function associated antigen(LFA)-1) (Gupta et al., 1993) L180-1 (anti-LFA-3) (Hunig, 1985) and ST197 (anti-T19) (Mackay et al., 1989) were used in this study. All washes used PBS with 0.1% BSA and 0.01% sodium azide (PBA) three times. Briefly, 1 X lo5 macrophages or 1 X lo6 PBMC per tube were stained with monoclonal antibody for 40 min at 0°C washed and then developed with anti-mouse IgG, F(ab’)2FITC conjugate (1:50, Dakopatts Ltd) or R-phycoerythrin conjugate (1:50, Serotec) for 20 min at 0°C. Five thousand or 10 000 cells gated using forward scatter (FSC) and side scatter (SSC) on macrophage or lymphocyte areas were analyzed by

116

W.C. Lee et al./Veterinary

Immunology and Immunopathology 51 (19961113-126

flow cytometry. Normal mouse serum (NM& 1500) stained cells were set up as background staining controls. These gave the same results as controls which used a non-specific mouse IgG,, (UPClO, Sigma) isotype matched monoclonal for SBU-T4 and SBU-T8. Cells were analyzed on a Becton Dickinson FACScan and the percentage of positive cells and mean channel number (mean fluorescence intensity; MF) were determined using Consort 30 Version F and Lysys Version 1.62 software. 3.4. FITC-labelled particles and opsonization Particles for phagocytosis experiments were FITC labelled at the following concentrations: a 5% suspension of sheep red blood cells (RBC), 4 X 10’ yeast (Candida utilis) (Sigma) per ml and 3-4 X lo8 CFU gram-negative bacterium Pasteurella hemolytica (serotype A6) (k’m dl y provided by Dr. J.E. Phillips, Department of Veterinary Pathology, University of Edinburgh) per ml. Particles were washed three times in PBS, then incubated with FITC at a final concentration of 100 pg ml- ’ in PBS for 30 min at 37°C (RBC) or 0°C (yeast and bacteria) (Buschmann and Winter, 1989; Tuijnman et al., 1990). The FITC labelled RBC and bacteria were washed twice in PBS and then opsonized with rabbit anti-sheep RBC antiserum (1:40; kindly provided by Dr. J. Hopkins, Department of Veterinary Pathology, University of Edinburgh) or rabbit anti-P. hemolytica (A6) serum (1: 10 dilution; kindly given by Dr. W. Donachie; Moredun Research Institute, Edinburgh) at the original particle concentrations. A stock of FITC-labelled, opsonized bacteria in 15% glycerol, RPM1 with no antibiotics (OD,,a = OS), was stored at - 70°C and used in phagocytic assays with fresh monocytes and AM from normal control and MVV-infected sheep. 3.5. Phagocytic assays The following concentrations of FITC-labelled particles were used in phagocytic and rosetting assays: RBC, 1% (v/v); C. utilis, 1 X lo8 per ml; P. hemolytica, 3-4 X lo* per ml. Washed MDM in 24well plates were fed with 250 ~1 RPMI/2% FCS containing 50 ~1 of FITC particles (above concentrations) and incubated for 1 h at 37°C. Non-ingested particles were removed with PBS. Any remaining RBC on the outside of MDM were hypotonically lysed with distilled water for 30 s and immediately neutralized with 2 X PBS. The cells were harvested, washed with PBA once and immediately analyzed by flow cytometry. MDM which had not been fed with FITC-labelled particles were used as background controls. AM and PBMC were washed in RPMI/2% FCS; 1 X 10’ AM or 1 X lo6 PBMC were pelleted and then re-suspended in 50 ~1 of FITC-labelled particles (see above). After 1 h incubation at 37°C non-ingested RBC were hypotonically lysed with distilled water for 30 s and neutralized with 2 X PBS. After another three washes with PBA, the cells were analyzed by flow cytometry. 3.6. Erythrocyte rosette assay AM or MDM were washed in cold PBA. Cell pellets were re-suspended in 50 ~1 of 1% FITC-labelled opsonized RBC and then incubated at 0°C for 30 min. The samples

W.C. Lee et (11./Veterinary

Immunology

and Immunopathology

51 (1996) 113-126

117

were re-suspended in PBA and immediately analyzed by flow cytometry. For erythrocyte rosetting by monocytes, PBMC were washed in PBA and then stained with the monoclonal antibody, VPM65, followed by the R-phycoerythrin anti-mouse Ig conjugate. After three washes with cold PBA, the cells were incubated with FITC-labelled opsonized RBC at 0°C for 30 min, re-suspended in cold PBA, and then analyzed by flow cytometry immediately. PBMC alone, PBMC fed with FITC-labelled particles, and PBMC stained with monoclonal antibody without FITC-labelled particles were used as background controls. 3.7. Detection of viral antigen by flow cytometry Infected and uninfected MDM were cultured and harvested as described above. To detect internal viral antigen, cells were fixed in 80% methanol in PBS at room temperature for 5 min followed by three washes with PBA before blocking with PBA/2% normal rabbit serum/2% normal sheep serum for 30 min at 0°C. The cells were then incubated with monoclonal antibody 415 (anti-MVV gag ~15) (Houwers and Schaake, 1987) at 0°C for 40 min, washed and developed with FITC-conjugated, anti-mouse IgG F(ab’)2 as above. 3.8. Sheep

MVV seropositive sheep from naturally infected flocks were of mixed breed (Texel, Blackface and Blau de Maine crosses) and age (all over 3 years old) (Watt et al., 1992). All experimentally infected sheep (Finnish Landrace Crosses and Texel crosses) had been infected with MVV (subcutaneous infection) for more than 2 years. Controls were seronegative experimental sheep (Finnish Landrace and Texel crosses), all over 2 years old.

4. Results 4.1. Monocyte-derived-macrophage

phenotypic

analysis

Cultured MDM were variable in size, so during flow cytometry were gated in a large area by forward and side scatter which excluded lymphocytes and cell debris. With monocyte maturation there is an increase not only in size, but also in autofluorescence of the cells and therefore the phenotype of MDM may not be correctly reflected using percentage positive cells particularly where molecules were weakly expressed. Hence, the mean fluorescence intensity (MF) of specific molecules with the MF of negatively stained (NMS) cells subtracted, was used to express the following data. Monocytes were cultured then infected with MVV on Days 7,9, 10, and 11. Antigen expression on MVV and mock infected MDM was then analyzed on Day 12. The results for uninfected and MDM Day 5 p.i. are shown in Table 1. Expression of all the surface molecules studied including CD4, CD8, MHC Class I and MHC Class II on infected MDM was not significantly altered up to 5 days p.i. (data analyzed by a Wilcoxon test).

Table 1 Surface molecule exoression

on MVV and mock infected monocvte Monocyte-derived

Surface molecule

’

macrophage

Mock infected

5 days p.i.

10.9* 29.7k

3.1 6.9 70.8+ 7.3 10.4+ 14.4 37.4 * 10.3 51.4* 6.0

CD4 CD8 MHC Class I MHC Class II DR LFA- I LFA-3

derived macronhages

7.3+ 6.0 23.4* 4.3 63.7+ 8.5 1.2f 10.2 46.2f 13.9 45.8 * 14.9

’ Cultured MDM on Day 7 were infected with MVV, I TCID,, per cell, and the phenotype analyzed on Day 12. Data (meanf SD) are expressed as mean fluorescence intensity (MF) with negative control fluorescence (NMS control) subtracted. Data were collected from five separate experiments.

After 7-8 days p.i., some surface molecules, especially MHC Class I which was highly expressed on cultured MDM, showed greatly decreased expression (data not shown). It was difficult to obtain enough cells for flow cytometric analysis at these late times p.i. (7-8 days) as the cells were productively infected by this time and decreased surface molecule expression could have been due to cell death. 4.2. Erythrocyte

rosetting and phagocytic

activity of MVV-infected

MDM

In order to see if macrophage phenotype correlated to altered biological activity, rosetting and phagocytosis of opsonized erythrocytes by MVV-infected MDM was studied using a flow cytometric assay. In the early phase of viral infection (up to 3 days p.i.>, both MVV-infected and uninfected cells showed very high activity, with nearly W-95% of cultured MDM having erythrocyte rosetting and phagocytic activity (Fig. 1 and Table 2). Experiments which analyzed Fc receptor expressed on the surface of MDM in productive infection (after 5 days p.i.) using erythrocyte rosetting did not show any difference between mock and MVV-infected MDM (Figs. lC, ID). However, the phagocytic capacity of MVV-infected MDM by 5 and 7 days p.i. was dramatically decreased to 67% and 47% positive cells, respectively, also reflected in a decreased mean fluorescence intensity (Figs. lG, 1H). There was a statistically significant difference between MVV-infected and mock-infected MDM at these time points (P < 0.05) (Table 2). 4.3. Binding and phagocytic utilis

activity of MVV-infected

MDM for P. hemolytica

and C.

The gram-negative bacterium, P. hemolytica, is a common pathogen seen in the upper respiratory tract of sheep. Therefore binding and phagocytosis of opsonized P. hemolytica by MDM was examined. There was no apparent impairment of binding and phagocytosis of opsonized bacteria by MDM infected with MVV by 5 and 7 days p.i. (Table 2). However, after washing, many externally bound bacteria were detected on the MDM by UV microscopy (data not shown) and so the assay could not differentiate

W.C. Lee et ul./

Veterinary

Immunology

FLUORESCENCE

and Immunoputhalngy

51 (19%)

113-1.26

119

c

INTENSITY

Fig. I. Comparison of erythrocyte rosetting and phagocytosis by MVV- or mock-infected MDM. MDM were cultured and infected with MVV on Days 5 (D, H), 7 (C, G) and 9 (B, F) respectively. A and E were mock-infected controls. Resetting (A-D) and phagocytosis (E-H) for RBC were assayed on Day 12. The unshaded areas ( K!) are negative control MDM without RBC and the shaded areas (W) are MDM binding or ingesting RBC. MFs of RBC fed MDM in each profile are: A = 58.8; B = 58.1; C = 57.2; D = 55.7; E = 64.6: F = 62.8; G = 49.6; H = 34.9.

between bound or ingested bacteria, unlike the erythrocyte rosetting and phagocytosis assay above. It is therefore likely that any fall in phagocytic activity for opsonized bacteria may be masked by the steady Fc receptor levels already shown by erythrocyte rosetting. In a similar manner to RBC, a defect in phagocytosis of yeast (C. urilis) by MVV-infected MDM was seen, in particular, by 5 days after virus infection (Fig. 2). At 7 days post infection, incubation with yeast for 15 min gave the lowest percentage of cells with phagocytosed particles (data not shown). The number of cells positive for yeast then increased up to a 30 min incubation but was unchanged after this. Therefore the defect in phagocytosis was not overcome by longer incubation times.

Table 2 F’hagocytic activity of cultured monocyte-derived Target particles

infected with maedi-visna

virus

Time p.i. (days)

RBC(N=S) P. hrmolyticu

macrophages

(N = 3) ’

Mock infected

3

5

I

95.7k3.0

95.2+ 3.7

66.9 i 14.0 =

47.4*

91.3+

87.8 i 3.0

87.3 &I 3.6

89.2 k 3.0

’

1.6

11 .o a

’ Data are expressed as percentage positive cells with phagocytic activity by setting a 1% gate on negative controls (no RBC or P. hemolytica). * Bound and ingested bacteria. a Indicates a statistically significant difference (P < 0.05) between uninfected and MVV-infected MDM using a Mann-Whihley two-tail test.

120

W.C. Lee rt d./

Veterinury Immum~lopy wtd Immunoputholo~y

FLUORESCENCE INTENSI IY

51 f 1996) 113-126

-

Fig. 2. Phagocytosis of yeast C. utrlis by MDM. MDM were cultured and infected with MVV on Days 5 (D). 7 (C), 9 (B), respectively. A is a mock-infected control. Phagccytosis was analyzed on Day 12. The unshaded area (0) represents negative control, mock-infected MDM which had not been fed FITC-labelled yeast. The shaded area ( n ) shows MVV-infected or mock-infected MDM fed with FITC-labelled targets. Figures on the top right (parentheses) are percentage positive cells using a I% gate from the control.

4.4. Detection

of viral antigen by flow cytometry

In order to study the relationship between phenotype or phagocytic function and viral replication in MDM, viral antigen (gag ~15) expressed within the cytoplasm was assayed. Internal viral antigen could be detected as early as 3 days p.i. (6.7% cells positive, data not shown) and by 4 and 6 days p.i.. 30% and 65% of MDM were positive for internal ~1.5 (Fig. 3). These figures are an underestimate of the number of infected cells expressing viral antigen at these times as the positively stained population

A(30.0%)

FLUORESCENCE

INTENSITY b

Fig. 3. Expression of viral antigen in MDM. Viral antigen (~q p 15) m the cytoplasm of MDM was detected by flow cytometry as described in Materials and methods. Negative staining with an irrelevant mouse IgG?, monoclonal antibody is shown by the unshaded area (0 ) with staining for R(IR pl5 shown by the shaded area (m). A 1% gate was set on the negatively stained cells and the percentage cells staining positive for ~cl~ above this is shown in each panel in parentheses. A: 4 days pi.; B, 6 days p.i.

W.C. Lee et d/Veterinary Table 3 Surface molecule

exoression

Sheep groups

Immunology und Immunoputhology 51 (1996) 113-126

on BAL fluid cells from MVV-infected

and control sheer,

Alveolar macrophages MHC Class I

Control sheep 63.1 f 12.6 ’ (42. I-79.2) 3 (N= 10) Infected sheep 78.6& 15.1 without lung (50.5-101.5) lesions (N = 13) 89.4+ 13.3 a Infected sheep with lung (70.6- 104.1) lesions (N = 8)

121

Lymphocytes

MHC Class II MHC Class II LFA-1

LFA-3

CD4/CD8

51.2k 12.3 (34.7-67.8)

59.95 13.1 (35.9-75.5)

3.2k2.1 (1.9-6.5)

57.8k9.4 (43.2-68.1)

58.3i.ll.9 (27.3-75.2)

2.5*1.5 (0.4-5.4)

(DQ)

(DR)

45.7k 13.2 (19.5-61.5)

56.4 f 9.0 (45.7-69.5)

55.2+ 13.5 a (33.6-78.5)

67.5+ 12.7 a (45.6-89.0)

84.5+8.7 a.b (7 1.6-92.7)

84.9k 11.9 a.b 64.9k 10.5 a 61.2+ 11.0 (63.9-97.9) (47.1-76.0) (46.5-74.8)

0.7kO.3

’

a

(0.4- 1.4)

’ Ratio of alveolar lymphocytes in brochoalveolar lavage fluid. ’ Values (mean *SD) have been expressed as mean fluorescence intensity (MF) (MF with Mab-MF with NMS). r Range of mean fluorescence intensity. a Indicates a statistically significant difference (P < 0.05) between MVV-infected and uninfected sheep analyzed by Mann-Whitney two-tail test. b Indicates a statistically significant difference between MVV-infected sheep with pulmonary lesions and without pulmonary lesions analyzed by Mann-Whitney two-tail test.

overlapped with the negative control on the FACScan MDM were always negative for gag ~15.

4.5. Comparison

profile (Fig. 3B). Mock infected

of the AM phenotype from MW-infected

and uninfected

sheep.

In this study most lungs from experimentally infected sheep (13/16) did not show gross pulmonary lesions, whilst those from naturally infected sheep (5/5) showed mild to marked interstitial pneumonia. The lungs with lesions were larger, firmer and heavier than normal lungs. They had a reddish-pink hue with multiple grey foci or were a diffuse grey in colour. Control sheep were experimentally culled sheep without infection and showed no gross pulmonary lesions. As previously reported (Cordier et al., 1990; Lujan et al., 19931, AM from MVV-infected sheep, either with or without pulmonary lesions, showed significant increases in the surface expression of MHC Class II DQ and DR compared with uninfected control sheep (Table 3). Here the AM from lungs showing lesions also had a significantly increased MHC Class I and LFA-1 expression. However this was not seen on AM from lungs without gross lesions (Table 3). LFA-3 expression on AM was unaltered by MVV infection in vivo (Table 3). AM were negative for CD4, CD8 and T19 molecules (data not shown). Using FSC and SSC parameters to gate on small lymphocytes, the CD4 and CD8 T lymphocyte population was also studied in BAL. There was a statistically significant decrease in the CD4:CD8 ratio in MVV-infected sheep with lung lesions compared with uninfected sheep (Table 3).

Table 3 Binding and phagocytosis control sheep

of P. hrmolytrcu

Sheep groups

and RBC hy alveolar

Binding and phagocytosis

from MVV-infected

MF ’

Phagocytosis (RBC) MF

Erythrocyte rosette o/r L

46.7 * 12.4 70.8 k 20.0 ’

54.4+ 54.2f

54.5 f 9.8 50.8 + 6.4

41.8k

47.

( P. hrmolyticu) Control sheep (N = 8) Infected sheep without lung lesions (N = 5) Infected sheep with lung lesions (N = 8)

macrophages

16.1

13.7 6.1

I i 14.5

48.

and

I i 7.5

Data are expressed as hoth ’ mean fluorescence intensity (MF) (MF with FITC targets-MF without FITC targets) or ’ percentage of positive cells by setting a I % gatr on the negative control. a Indicates a statistically significant difference (P < 0.05) between MVV-infected and control sheep using a Mann-Whitney two-tail test.

4.6. Particle binding and phagocytosis

by AM

There was no significant difference in the RBC rosetting or phagocytosis activity of AM from sheep with maedi (lesions), MVV-infected sheep (no lung lesions) or controls (Table 4). However, with P. hemofytica the binding and phagocytic activity of AM was enhanced in MVV-infected sheep without lung lesions (MF 46.7 increased to 70.8). There was no difference in phagocytosis of P. hemolyticu between the control group and AM from sheep with MVV induced lung lesions (Table 4). 4.7. Rosetting und phagocytosis

qf erythrocytes

and P. hemolytica

by fresh nzonocytes

The phagocytic capacity of fresh monocytes during MVV infection was studied using a modified double staining method and flow cytometry. Rosetting and phagocytic

3

3r

ct s

0

.

ao-

8

;

iv, a I .

I

I

Binding 8 phagocyiosis P hemdytica

i

:* a,

.

02

:

Phagocytosis

Resetting

RBC

Fig. 4. Rosetting and phagocytic activity of monocytes. FITC lahelled particle binding and phagocytosis of CD14+ cells (VPM6S+ ) in blood were assayed as in the Materials and methods. A comparison of rosetting and phagocytic activity of monocytes from eight control sheep ( 0 ), nine experimental sheep without clinical symptoms ( n ) and five naturally infected sheep (0) is shown.

W.C. Lee et al./ Veterinwy

Immunology and Immurwpathology

51 (1996) 113-126

17-3

activities were found in fresh monocytes which were labelled with VPM65, anti-CD14. Compared with cultured MDM, fresh monocytes did not show a very strong phagocytic activity for opsonized RBC and bacteria and the activity seen varied greatly between individual sheep. When infected and uninfected sheep were compared there was no difference in the erythrocyte rosetting activity of monocytes from uninfected or MVVinfected sheep or the phagocytic activity of monocytes for opsonized RBC and P. hemofyticu (Fig. 4). Statistical analysis used a Mann-Whitney non-parametric test.

5. Discussion In this study we have looked at in vitro infected macrophages and macrophages from in vivo infections of sheep to see if changes in phenotype of the cells are indicative of changes in function, here the ability to bind and phagocytose three different types of particles. Opsonized sheep RBC and gram-negative bacteria bind through Fc receptors of the macrophages whilst yeast were used without antibody as they can bind efficiently to macrophages through the mannosyl fucosyl and p-glucan receptors (Speert, 19921. MDM which had been cultured for 7 days were used as they were considered to have differentiated into macrophages (by morphology, non-specific esterase staining and stable surface antigen expression). In vitro by 5 days p.i. MVV-infected MDM had a functional impairment of ingestion (for RBC and yeast) not reflected in the apparently normal phenotype and surface particle binding. Cells were productively infected by this time which may have caused disruption of the cellular cytoskeleton affecting ingestion of particles bound to surface receptors. Rearrangement of the cytoskeleton is known to occur after infection with HIV (Luftig and Lupo, 1994). However, the binding and phagocytosis of the opsonized bacterium, P. hemolyrica, was apparently unaffected at this time p.i. This may have been due to bound bacteria on the surface of the macrophages which were not washed off during the assay or alternatively the bacteria may be small enough to overcome the phagocytic defect shown by yeast and RBC. The BAL samples taken from MVV-infected sheep in this study showed similar results to those previously reported: a decreased CD4:CD8 lymphocyte ratio and the presence of AM with increased MHC Class II and LFA-1 expression (Watt et al., 1992; Lujan et al., 1993). As the severity of MVV induced pathology increased in the lung, the change in phenotype of the macrophages became more pronounced for MHC antigens and LFA-1 although LFA-3 expression did not change on the AM from any group. The actual level of detectable virus antigen is very low in MVV-infected lungs (Haase et al., 1977; Gendelman et al., 1985) and lesions caused by MVV are inflammatory in origin, being typified by mononuclear cell infiltrates which accumulate over time (Narayan and Clements, 1989). The interaction of lymphocytes with MVV-infected macrophages leads to the production of lentivirus IFN (Narayan et al., 1985) and it may be through the action of cytokines that antigen expression is increased. Indeed alveolar lymphocytes from MVV-infected sheep show increased MI-K Class II but decreased CD5 expression (Watt et al., 1992; Lujan et al., 19931 suggesting they are activated and so may be secreting cytokines which could influence macrophage phenotype.

Erythrocyte rosetting was unaffected in the AM from these animals suggesting that Fc receptor expression is unaffected in AM from MVV-infected sheep. Similarly AM from MVV-infected sheep showed no increase in phagocytosis of RBC. Therefore the activated phenotype expressed by these cells did not correlate to this activity. However there was a significant increase in the ability of AM from MVV-infected sheep without lung lesions to bind and phagocytose the opsonized bacterium, P. hemolytica. From the erytbrocyte rosetting ability of the cells, we know that Fc receptor expression was probably unaffected and so any increase in binding with P. hemolyrica may be due to other surface receptors for the bacteria, for example mannosyl fucosyl or LPS receptors 1 erence in particle size (RBC > bacteria) may also contribute to (Speert. 1992). Th e d’ff differences in activity. However the increase in binding and phagocytic function for P. hemolytica was not seen in AM from lungs with MVV lesions and it is these cells which show the most pronounced changes in phenotype. This again confirms that increased MHC Class II expression does not correlate with increases in phagocytic function of the AM. Pasteurellosis is sometimes found in MVV-infected sheep with interstitial pneumonia (Markson et al., 1983; Myer et al., 1988). The data presented here suggest that early events in AM interaction with bacteria (binding and internalization) are not affected in sheep with MVV induced lung lesions. Therefore later stages in the anti-bacterial function of AM (e.g. intracellular killing) or other factors involved in normal pulmonary function may be reduced in MVV-infected sheep with pasteurellosis. Normal pulmonary immunity involves local immunity (cell-mediated immunity, IgG, IgA), pulmonary secretions and physiological conditions in the lung (Kaltreider, 1976; Burrells, 1985). Perhaps a major contribution is the alveolar wall thickening seen in lungs affected by maedi which reduces transduction of plasma components into the respiratory system (Collie et al., 1993). This may lead to a failure in opsonization and allow bacteria entering the lower respiratory tract to replicate. There was no defect in the binding and phagocytic activity of monocytes from MVV-infected sheep for P. hemolytica in this study. This suggests that the low level of MVV in monocytes from infected sheep (Haase et al., 1977; Gendelman et al., 1985) does not interfere with their function. In addition, there may be no build-up of infected monocytes in the blood as circulating monocytes quickly enter tissues and differentiate into macrophages (Auger and Ross, 1992). In vitro infected macrophages have reduced phagocytic activity for RBC and yeast but normal surface molecule expression. This suggests that the phenotypic changes seen in vivo (increased MHC antigen and LFA-I expression on AM) are not directly caused by MVV but by inflammatory interactions at the site of lesions probably involving IFN-y. Activation of macrophages by IFN-y normally increases opsonic activity and this was seen for P. hemolytica in AM from lesion free infected sheep but not for AM from maedi positive sheep (although these cells are phenotypically highly activated). This could indeed indicate reduced phagocytic activity as a result of MVV infection as is seen in vitro. However, the percentage of productively infected AM may be low in vivo (Haase et al.. 1977: Gendelman et al., 1985) and so the lack of increased phagocytosis may be due to soluble viral products (for example the envelope glycoprotein) interfering with cellular activities. AM from cases of maedi show increased

W.C. Lee et al./ Veterinary Immunology and Immumpathology 51 (1996) 113-126

125

fibronectin and neutrophil chemotactic factor release (Cordier et al., 1990) but not the expected increase in phagocytosis considering the activation phenotype of the cells. This suggests that macrophages from sites with MVV lesions are activated for the induction of inflammation but not all other macrophage activities.

Acknowledgements

We would like to thank D. Allen and J. Williams for their help and support. B.A. Blacklaws and P. Bird were supported by MRC ADP Project Grant G890079 and later by Wellcome Trust Programme Grant 035157/Z/92/2/1.27. W.C. Lee was supported by the Ministry of Education, Taipei, Taiwan, ROC.

References Auger, M.J. and Ross, J.A., 1992. The biology of the macrophage, In: C.E. Lewis and J.O’D. McGee, (Editors), The Natural Immune System: The Macrophage, IRL Press, Oxford, pp. l-74. Blacklaws, B.A., Bird, P., Allen, D. and McConnell, I., 1994. Circulating cytotoxic T lymphocyte precursors in maedi-visna virus-infected sheep. J. Gen. Virol., 75: 1589- 1596. Bowen, D.L., Lane, H.C. and Fauci, AS., 1986. Immunological abnormalities in the acquired immunodeticiency syndrome. Prog. Allergy, 37: 207-223. Bray, D.H.. Squire, S.B., Kawana, A., Johnson, M.A. and Poulter, L.W., 1993. Antiretroviral treatment reverses HIV-induced reduction in the expression of surface-antigens on alveolar macrophages in AIDS patients. Clin. Exp. Immunol., 91: 13- 17. Burrells, C., 1985. Cellular and humoral elements of the lower respiratory tract of sheep. Immunological examination of cells and fluid obtained by bronchoalveolar lavage of normal lungs. Vet. Immunol. Immunopathol., 10: 225-243. Buschmann, H. and Winter, M., 1989. Assessment of phagocytic-activity of granulocytes using laser flow cytometry. J. Immunol. Methods, 124: 23 l-234. Collie, D.D.S., Watt, N.J., Warren, P.M., Begara, I. and Lujan, L., 1993. Effects on lung compliance, lung volume, and single-breath transfer factor for carbon monoxide in sheep with lentivirus-induced lymphoid interstitial pneumonia. Am. J. Vet. Res., 54: 454-462. Cordier, G., Cozon, G., Greenland, T., Rocher, F., Guiguen, F., Guerret, S., Brune, J. and Momex, J.F., 1990. In vivo activation of alveolar macrophages in ovine lentivirus infection. Clin. Immunol. Immunopathol., 55: 355-367. Dutia, B.M., Hopkins, J., Allington, M.P., Bujdoso, R. and McConnell, I., 1990. Characterization of monoclonal-antibodies specific for alpha-chains and beta-chains of sheep MHC class-II. Immunology, 70: 27-32. Gendelman, H.E., Narayan, 0.. Molineaux, S., Clement% J.E. and Ghotbi, Z., 1985. Slow, persistent replication of lentivituses: role of tissue macrophages and macrophage precursors in bone marrow. Proc. Natl. Acad. Sci. USA, 82: 7086-7090. Gendelman, H.E., Orenstein, J.M., Baca, L.M., Weiser, B., Burger, H., Kalter, D.C. and Meltzer, M.S., 1989. The macrophage in the persistence and pathogenesis of HIV infection. Aids, 3: 475-495. workshop monoclonal Gupta, V.K., McConnell, I. and Hopkins, J., 1993. Reactivity of the CDll/CDIS antibodies in the sheep. Vet. Immunol. Immunopathol., 39: 93- 102. Haase, A.T., Stowing, L., Narayan, O., Griffm, D.E. and Price, D.. 1977. Slow persistent infection caused by visna virus: role of host restriction. Science, 195: 175- 177. Harkiss, G.D., Watt, N.J., King, T.J., Williams, J. and Hopkins, J., 1991. Retroviral arthritis: phenotypic analysis of cells in the synovial-fluid of sheep with inflammatory synovitis associated with visna virus-infection. Clin. Immunol. Immunopathol., 60: 106- 117.

Hopkins. J. and Dutia, B.M., 1990. Monoclonal-antibodies to the sheep analogues of human-CD45 (leukocyte common antigen), MHC class-l and CD5differential expression after lymphocyte-activation in viva. Vet. Immunol. Immunopathol., 24: 33 I-346. Houwers, D.J. and Schaake, J., 1987. An improved ELISA for the detection of antibodies to ovine and caprine lentiviruses, employing monoclonal antibodies in a one-step assay. J. Immunol. Methods. 98: I5 I - 154. Hunig. T.. 1985. The cell surface molecule recognized by the erythrocyte receptor of T lymphocytes: identification and partial characterization using a monoclonal antibody. J. Exp. Med., 162: 890-901. Jones, B.M.. Nicholson. J.K.A.. Holman, R.C. and Hubbard, M., 1989. Comparison of monocyte separation methods using flow cytometric analysis. J. Immunol. Methods. 125: 41-47. Kaltreider. H.B.. 1976. Expression of immune mechanisms in the lung. Am. Rev. Resp. Dis.. 113:347-379. Luftig. R.B. and Lupo, L.D., 1994. Viral interactions with the host-cell cytoskeleton: the role of retroviral proteases. Trends Microbial., 2: 17% 182. Lujan. L.. Begara. I., Collie, D.D.S. and Watt, N.J.. 1993. Phenotypic analysis of cells in bronchoalveolar lavage fluid and peripheral-blood of maedi visna-infected sheep. Clin. Exp. Immunol.. 91: 272-276. Mackay. C.R., Beya, M.F. and Matzinger, P., 1989. yS T-cells express a unique surface moleculr appearing late during thymic development. Eur. J. Immunol., 19. 1477-14X3. Maddox, J.F.. Mackay. C.R. and Brandon, M.R., 1985. Surface antigens, SBU-T-l and SBU-TX. of sheep T lymphocyte subsets defined by monoclonal antibodies. Immunology. 55: 7.39-718. Markson. L.M.. Dawson. M. and Spence. J.B., 1983. Investigation of a flock heavily infected with maedi-visna virus. Vet. Rec., I 12: 267-271, Myer, M.S., Huchzermeyer. H.F.A.K.. York. D.F., Hunter. P.. Verwoerd. D.W. and Gamett. H.M., 1988. The possible mvolvement of immunosuppression caused by a lentivirus in the aetiology of jaagsiekte and pasteurellosis in sheep. Onderstepoort J. Vet. Res., 55: l27- 13.1. Narayan. 0. and Clements, J.E., 1989. Biology and pathogenesis of lentiviruses. J. Gen. Viral.. 70: 1617-1639. Narayan, 0.. Sheffer, D.. Clements, J.E. and Tennehoon, G.. 1985. Restricted replication of Ientiviruses: visna viruses induce a unique interferon during interaction between lymphocytes and infected macrophages. J. Exp. Med., 162: 1954- 1969. Petit, A.J.C., Terpstra. F.G. and Miedema, F.. 1987. Human immunodeliciency virus infection down-regulates HLA class-11 expression and induces differentiation in promonocytic U937 cells. J. Clin. Invest., 79: I 883- 1889. Roy, G.. Rojo, N. and Leyva-Cobian. F.. 1987. Phenotypic changes in monocytes and alveolar macrophages in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex (ARC). J. Clin. Lab. Immunol., 23: 135-141. Sargan, D.R., Bennet, I.D., Cousens, C.. Roy, D.J.. Blacklaws. B.A.. Dalziel, R.G.. Watt, N.J. and McConnell, 1.. 1991. Nucleotide sequence of EVI. a British isolate of maedi-visna virus. J. Gen. Virol., 72. 1893~1903. Smith, P.D.. Ohura, K., Masur. H.. Lane, H.C.. Fauci, AS. and Wahl. S.M., 1984. Monocyte function in the acquired immune-deficiency syndrome: defective chemotaxis. J. Clin, Invest., 74: 2121-2128. Speert. D.P., 1992. Macrophages in bacterial infection, In: C.E. Lewis and J.O’D. McGee, (Editors), The Natural Immune System: The Macrophage. IRL Press, Oxford, pp. 215-263. Tuijnman. W.B.. Van de Winkel, J.G.J. and Cape], P.J.A.. 1990. A flow cytometric rosetting assay for the analysis of IgG-Fc receptor interactions. J. Immunol. Methods, 127: 207-214. Wahl, S.M., Allen. J.B., Gartner. S., Orenstein, J.M.. Popovic, M., Chenoweth, D.E., Arthur, L.O.. Farrar, W.L. and Wahl. L.M., 1989. HIV-I and its envelope glycoprotein down-regulate chemotactic ligand receptors and chemotactic function of peripheral-blood monocytes. J. Immunol.. 142: 3553-3559. Watt. N.J., Macintyre. N.. Collie, D.. Sargan. D. and McConnell, 1.. 1992. Phenotypic analysis of lymphocyte populations in the lungs and regional lymphoid-tissue of sheep naturally infected with maedi-visna virus. Clin. Exp. Immunol.. 90: 204-208.