Changes of leukocyte phenotype and function in the ... - CiteSeerX

Journal of General Virology (2000), 81, 497–505. Printed in Great Britain ...................................................................................................................................................................................................................................................................................

Changes of leukocyte phenotype and function in the broncho-alveolar lavage fluid of pigs infected with porcine reproductive and respiratory syndrome virus : a role for CD8M cells Janneke N. Samsom, Tiny G. M. de Bruin, John J. M. Voermans, Janneke J. M. Meulenberg, Jan M. A. Pol and Andre T. J. Bianchi Department of Mammalian Virology, Institute for Animal Science and Health, PO Box 65, NL-8200 AB Lelystad, The Netherlands

Porcine reproductive and respiratory virus (PRRSV) primarily infects and destroys alveolar macrophages of the pig. The aim of the present study was to characterize the changes of leukocyte populations in the broncho-alveolar lavage fluid (BALF) of PRRSV-infected pigs. Piglets were inoculated intranasally with PRRSV strain LV ter Huurne. On various days post-infection the piglets were sacrificed and the lungs removed, washed semi-quantitatively and analysed by flow cytometry. The total number of recovered BALF cells increased approximately 10 times between day 10 and day 21 of infection and decreased thereafter. The number of small low-autofluorescent cells (SLAC), i.e. lymphocytic and monocytic cells, increased very strongly from day 2 until day 21 of infection ; in contrast, the number of large highly autofluorescent cells (LHAC), i.e. mostly macrophages, remained constant until day 14 of infection, increased slightly on day 21 and then decreased. On day 21 of infection in specific-pathogen-free piglets approximately 60 % of the SLAC consisted of CD2MCD8MCD4NγδTCRN cells, which were partly CD8MCD6M and partly CD8MCD6N. These phenotypes correspond to that of cytotoxic T-cells and natural killer cells respectively. From these results we can conclude that during a PRRSV infection the total number of BALF cells increases mainly due to an influx of lymphocytic cells with a cytolytic phenotype.

Introduction Porcine reproductive and respiratory syndrome virus (PRRSV) is a positive-strand RNA virus that belongs to the arterivirus family (Meulenberg et al., 1993 ; Conzelmann et al., 1993). An important characteristic of the arteriviruses is their strong tropism for cells that belong to the monocyte– macrophage lineage (Tong et al., 1977 ; Plagemann & Moening, 1992 ; Pol & Wagenaar, 1992 ; Voicu et al., 1994). PRRSV primarily infects alveolar macrophages of the pig which, in young piglets, may result in severe respiratory distress, whereas in sows infection often leads to reproductive failure (Hill, 1990 ; Lindhaus & Lindhaus, 1991 ; Cromwijk, 1991 ; Paton et al., 1991). Author for correspondence : Janneke Samsom. Present address : Free University, Department of Cell-Biology, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands. Fax j31 20 444 8080. e-mail jn.samsom.cell!med.vu.nl

0001-6480 # 2000 SGM

The effect of PRRSV infection on the lungs has been studied extensively. Reports on the pathogenesis of PRRSV in the lungs have revealed the presence of necrotic cell debris in the alveolar lumina and a thickening of the alveolar septa due to infiltration of mononuclear cells during the first week of infection (Pol et al., 1993 ; Collins et al., 1992 ; Halbur et al., 1993). The virus can be detected in macrophages in the alveolar septa and in the alveolar spaces (Pol et al., 1993 ; Halbur et al., 1994). However, even during the acute stage of an experimental infection no more than 2 % of the alveolar macrophages stain positive for PRRSV antigen (Duan et al., 1997 ; Mengeling et al., 1995). Analysis of the composition of the alveolar cell-population in the lungs of infected pigs yields variable results. Based on histological examination of haematoxylin–eosin-stained lung sections it has been reported that the alveolar lumina are filled with neutrophils and macrophages during the first days of infection (Rossow et al., 1994 ; Halbur et al., 1996). Similarly, microscopical examination of broncho-alveolar lavage fluid EJH

J. N. Samsom and others

(BALF) on day 7 of infection revealed that 35 % of the BALF cells were neutrophils, 50 % were macrophages and 15 % were lymphocytes (Zhou et al., 1992). However, others have reported that the percentage of neutrophils in the BALF-cell population remained low during a 56 day observation period after intranasal (i.n.) infection, whereas the percentage of macrophages decreased and the percentage of lymphocytes increased between days 0 and 28 of infection (Shibata et al., 1997). Furthermore, others report an increase in the number of macrophages in the broncho-alveolar lavage fluid of piglets that were intratracheally infected with PRRSV and sacrificed during the first days of infection (Van Reeth & Pensaert, 1997). These contradictory data may be due to a number of factors. Firstly, the porcine alveolar macrophage population is a very heterogeneous population and therefore relatively difficult to study microscopically (Choi et al., 1994). This problem may be avoided by using flow cytometry. With this technique it is possible to identify and quantify subsets of cells within a heterogeneous population by measuring size, granularity and presence of surface markers on the cells (Berndt & Mu$ ller, 1997). Secondly, a difference in microbiological status of the pigs may determine their susceptibility to PRRSV infection and may lead to variability in results. For example, it has been suggested that gnotobiotic piglets are more sensitive to PRRSV infection than specific-pathogen-free (SPF) piglets. Also, variation in the composition of the alveolar cell population in pigs with a different microbiological status may affect their response to PRRSV infection. Thirdly, the changes in the alveolar cell population may vary at different times of PRRSV infection. Lastly, most results have focused on relative changes in leukocyte populations, thus neglecting possible changes in absolute numbers. The aim of our study was to characterize the phenotypic changes of leukocyte populations in BALF of PRRSV-infected piglets using flow cytometry.

Methods

Animals and experimental setup First experiment. Gnotobiotic piglets were born by closed hysterectomy from F (Dutch LandraceiLarge White Yorkshire) sows. The " piglets were housed in stainless steel isolator units at a temperature of approximately 29n5 mC and were fed a commercial sterile milk substitute (Nutricia, Zoetermeer, the Netherlands). The animals were randomly allocated to an infected (n l 9) or a control (n l 9) group. At 7 days of age one group of piglets was infected with 0n5 ml per nostril of PRRSV strain ter Huurne at a TCID of 10&\ml. The other &! group of gnotobiotic piglets received uninfected culture supernatant in a similar manner and served as control piglets. On day 5 post-infection the piglets were sacrificed and the lungs removed and divided in two parts. The left lung-half was used for semi-quantitative bronchoalveolar lavage whereas the right half was used for immunohistochemistry. Second experiment. Dutch Landrace piglets aged 8 to 10 weeks were obtained from the SPF herd of the ID-DLO. After 1 week of acclimatization 20 piglets were infected with 0n5 ml per nostril of

EJI

PRRSV stain ter Huurne at a TCID of 10&\ml. Control SPF piglets &! received uninfected culture supernatant in a similar manner. On days 0, 2, 4, 7, 10, 14, 21, 28, 35 and 42 post-infection the piglets were sacrificed, and the lungs removed and washed semi-quantitatively. The experiments described in this study were performed according to regulations of the Animal Care Committee of the Institute.

Semi-quantitative broncho-alveolar lavage. Prior to the lavage the weight of each lung was determined. Thereafter, the lung was washed on ice with a fixed volume of ice-cold PBS. After gently squeezing the lung several times to spread the PBS over the alveoli, the BALF was recovered and its volume was determined. The BALF cells were collected by centrifugation at 300 g for 10 min at 4 mC. The cells were washed twice with ice-cold PBS and resuspended in a fixed volume of RPMI 1640 Dutch modification medium (ICN Biomedicals) containing 10 % heat-inactivated foetal bovine serum (Integro), 200 U\ml sodium penicillin-G (Yamanouchi), 0n2 mg\ml streptomycin (Biochemie), 0n3 mg\ml -glutamine (Flow Laboratories) and 5i10−& M β-mercaptoethanol (Sigma), hereafter referred to as medium. An aliquot of the cellsuspension was used for virus isolation and the total number of recovered cells per lung was determined. Thereafter, the cells were adjusted to a final concentration of 1i10( cells\ml in medium. A portion of the BALF cells was dispensed in aliquots of 50 µl per well in a 96-well V-bottomed microtitre plate (Nunc) for flow cytometric analysis. In addition, cytospin preparations were made and a portion of the remaining cells was used for the induction of cytokine release.

Flow cytometric analysis of BALF cells. BALF cells and PBMC were spun down in the V-bottomed microwell plates by centrifugation (230 g at 4 mC). The supernatant was discarded and the cells were incubated for 20 min on ice with various combinations of murine monoclonal antibodies (MAbs) directed against, or cross-reactive with, porcine leukocyte differentiation antigens (Fig. 1.) The MAbs used were : SWC1 (clone 76-6-7, IgM), SWC3 (clone 74-22-15, IgG ), MHC II (clone " MSA3, IgG2a), CD14 (clone MY4, Ig2b) (Coulter), CD11b (clone C25, IgG1), 517.2 ligand (517.2L) (clone CVI 517.2, IgG2b), CD3ε (clone ppt3, IgG1), CD2 (clone MSA4, IgG2a), CD4 (clone 74-12-4, IgG2b), CD8 (clone 295\33 IgG2a), CD5 (clone b53b7, IgG1), CD6 (clone a38b2, IgG1) and γδT-cell receptor (clone ppt16, IgG2b) ; all were diluted to optimal concentrations in PBS containing 2 % heat-inactivated foetal bovine serum and 0n01 % sodium azide (FACS buffer). After incubation the cells were washed three times with FACS buffer. The cells were incubated a second time for 20 min on ice with the appropriate fluorescein isothiocyanate (FITC)- or phycoerythrin (PE)-conjugated goat antimurine IgG isotype specific antibodies diluted in FACS buffer. After the second incubation the cells were washed three times and resuspended in FACS buffer. The cells were transferred to tubes and fluorescence was measured using a FACScan. For BALF cells a population of large highly autofluorescent cells and a population of small low-autofluorescent cells were distinguished and analysed separately.

Virus isolation from BALF cells. Aliquots of BALF cells in medium were frozen at k70 mC. After two freeze–thaw steps a virus isolation was performed as described in the O.I.E. Manual of Standards for Diagnostic Tests and Vaccines 1996, chapter X.12, Porcine Reproductive and Respiratory Syndrome, pp. 694–700.

Immunohistochemistry. A segment of the frontal lobe (A), cardiac lobe (B) and dorsal lobe (C) of the right lung-half was taken immediately after isolation of the lungs. The segments were rapidly frozen in liquid nitrogen and stored at k70 mC until use. Cryostat sections of all segments were made at 8 µm thickness at k20 mC, fixed in acetone for 10 min and air-dried. Sections and cytospins were stained according to the same procedure. In short, sections were incubated for 5 min in 0n06 %

CD8+ cells in the lungs of pigs with PRRS

Fig. 1. Characterization of swine leukocytes using differentiation antigens. Abbreviations : CD, cluster of differentiation ; SWC, swine workshop cluster ; γδTCR, T-cell receptor-γδ ; γδTCR, T-cell receptor-γδ ; MHC, major histocompatibility complex ; NK cell, natural killer cell. (j) indicates that the cell expresses the antigen ; (k) indicates that the cell does not express the antigen ; (j/k) indicates that the cell may or may not express the antigen. (hi) indicates that the cells express the antigen at a high density ; (low) indicates that the cells express the antigen at a low density.

Tris buffer supplemented with 2 % H O to eliminate endogenous # # peroxidase activity prior to immunoperoxidase staining. For the immunoperoxidase staining the sections were incubated for 20 min at room temperature with either anti-PRRSV-nucleoprotein antibody (SDOW 17) or anti-CD8 (clone 295\33, IgG2a), diluted to optimal concentrations in PBS containing 0n2 % (v\v) BSA (PBS–BSA). Control sections were stained with an irrelevant MAb of the same isotype. After incubation the sections were washed thoroughly with PBS–BSA and incubated for 20 min at room temperature with peroxidase-labelled rabbit anti-mouse Ig diluted 1\200 in PBS–BSA. The sections were washed and peroxidase activity was visualized by incubation with 0n17 mg\ml 3’3’-diaminobenzidine tetrahydrochloride (DAB) in 0n05 M Tris–HCl buffer (pH 7n6) supplemented with 0n02 % H O . The sections # # were counterstained with haematoxylin.

Statistical analysis. For experiment 1 results are expressed as dots for individual animals or as meanpSD for nine pigs per group. Analysis was performed with two-sample Student’s t-tests. In experiment 2 statistical analysis was performed using analysis of variance followed by a Fisher’s least significant difference test. It should be noted that the results are expressed as means of two pigs per timepoint. With larger groups even more significant differences might have been obtained. Values for uninfected animals are not shown but were comparable to the values on day 0.

Results Experimental design

Results of two different experiments will be described separately. In the first experiment the changes of leukocyte phenotype and function during PRRSV infection in one-weekold gnotobiotic piglets were assessed at one time-point during infection. To assess whether the observations were not exclusive for a PRRSV infection in young gnotobiotic piglets and to study the kinetics of these changes a second experiment was performed in 8 to 10-week-old SPF piglets. Characterization of BALF cells recovered from the lungs during a PRRSV infection in gnotobiotic or SPF piglets : number, phenotype and cell-associated virus Gnotobiotic piglets. In one-week-old gnotobiotic piglets the

total number of BALF cells isolated from the lung-halves of PRRSV-infected piglets (2n92p1n88i10() on day 5 of a PRRSV infection was not significantly different from that of control piglets (4n3p2n09i10(). EJJ

J. N. Samsom and others

Fig. 2. Characterization of BALF cells recovered from the lungs of SPF piglets with a PRRSV infection : number, phenotype and cell-associated virus. SPF piglets were inoculated i.n. with 105 TCID50 PRRSV strain ter Huurne. On various days post-infection the piglets were sacrificed and the lungs were removed and lavaged semi-quantitatively. The total number of BALF cells (#) recovered per lung was determined microscopically (A). The total number of LHAC (=) and SLAC ( ) were calculated for each individual piglet (A) using the total number of recovered BALF cells (A) and the percentages of SLAC ( ) and LHAC ( ) which had been determined by flow cytometry (B). The amount of cell-associated PRRSV was determined according to standard procedures and expressed in arbitrary units (C). Dots represent means of two animals per time-point ; values for uninfected animals are not shown but were comparable to the values on day 0. (*) indicates P 0n05 compared to t l 0.

To determine the phenotype of the BALF cells flow cytometric analysis was performed. As described by others multiple subpopulations of the BALF cells can be distinguished on the basis of size and granularity (Berndt & Mu$ ller, 1997). In our study the BALF cells were separated into two populations. Firstly, a population of small low-autofluorescent cells (SLAC) was distinguished which comprised small monocytic cells and lymphocytic cells. Secondly, a population of large highly autofluorescent cells (LHAC) was identified which mostly consists of large granular macrophages and intermediate-size but also granular macrophages. On day 5 of PRRSV infection in gnotobiotic piglets the percentage of SLAC was significantly higher in the BALF of PRRSV-infected (53n00p7n55 %) than in control (30n29 %p7n36 %) piglets. The percentage of LHAC was significantly lower in the BALF of PRRSV infected (36n83p5n13 %) than in control piglets (59n95p7n33 %). However, when expressed in numbers, the number of SLAC (1n54p1n03 i10() of PRRSV-infected piglets were slightly but not significantly different from the number of SLAC (1n22p0n42i10() in control piglets, due to the large variation in the total number of lavaged cells. Similarly, the number of LHAC (1n09p0n72i10() in BALF of PRRSVinfected pigs was not different from the number of LHAC (2n69p1n55i10() in control piglets. No differences in lung weight and volume of recovered BALF of infected and control piglets were observed. SPF piglets

During a PRRSV infection in 8 to 10-week-old SPF piglets the total number of BALF cells that was isolated decreased between day 0 and day 2 of infection, returned to its initial level between day 4 and 10 of infection, increased significantly to a peak at day 21 of infection and decreased thereafter (Fig. 2 A) FAA

During PRRSV infection in SPF piglets the percentage of SLAC steadily increased to a maximum of 50 % on day 14 of infection whereas the percentage of LHAC decreased from 90 % to 39 % on day 14 of infection (Fig. 2 B). When expressed in total numbers, the number of SLAC increased dramatically from day 2 until day 21 of infection whereas the total number of LHAC increased from day 14 until day 21 of infection. The amount of PRRSV that was isolated from the BALF cells reached a peak on day 7 of infection, decreased to undetectable levels on day 21 of infection and remained undetectable until day 42 of infection (Fig. 2 C). Flow cytometric analysis of SLAC in the BALF of gnotobiotic and SPF piglets Gnotobiotic piglets. On day 5 of infection the percentages of

SWC1+ SLAC and SWC3+ SLAC in PRRSV-infected piglets were significantly increased in comparison with the percentages of SWC1+ and SWC3+ in control piglets (Fig. 3). Since SWC1 is a common marker for monocytes, granulocytes and resting T-lymphocytes and SWC3 is a common marker for myeloid cells, staining of more specific myeloid and lymphoid markers was used to characterize the differences observed (Fig. 1). Analysis of macrophage and monocytic markers revealed that the percentage of both 517n2L+ and MHC II+ SLAC were significantly increased in PRRSV-infected piglets in comparison with controls. On the other hand, the number of CD14+ SLAC was only slightly but not significantly increased in comparison with controls (Fig. 3). Analysis of lymphoid markers revealed that the percentage of CD2+CD3− SLAC in PRRSV-infected piglets was higher than in control piglets (Fig. 3). Furthermore, the percentage of CD2+CD3+ SLAC in PRRSV-infected piglets was also slightly but significantly higher than in control piglets (Fig. 3). The

CD8+ cells in the lungs of pigs with PRRS

Fig. 3. Flow cytometric analysis of SLAC in the BALF of gnotobiotic piglets. One-week-old gnotobiotic piglets were inoculated i.n. with 0n5 ml 105 TCID50 PRRSV strain ter Huurne. At day 5 of infection control (#) and infected ( ) piglets were sacrificed and the lungs were removed and one lung-half washed semi-quantitatively. The BALF cells were stained with MAbs against various surface antigens, as indicated in Methods. The SLAC were selected and analysed. Dots represent individual animals and the bar indicates the mean value.

percentage of CD2−CD3+ cells was similar in both groups (Fig. 3). Double staining of SLAC for CD8 and CD4 revealed that the percentage of SLAC with the cytolytic phenotype CD8+CD4− was higher in PRRSV-infected piglets in comparison with controls but that the percentages of CD8−CD4+ and CD8+CD4+ SLAC were similar in both groups (Fig. 3). Cytolytic cells can be separated into natural killer cells with a CD8+CD6− phenotype, cytotoxic T-cells with a CD8+CD6+

phenotype and γδT-cells with a CD8+γδTCR+ phenotype (Fig. 1). Further analysis of the increased CD8+CD4− SLAC showed that significant percentages of CD8+CD6− and CD8+γδTCR− SLAC were present in the BALF of PRRSV-infected piglets whereas these were virtually absent in control piglets (Fig. 3). In addition, no CD8−CD6+, CD8+CD6+, CD8−γδTCR+ or CD8+γδTCR+ SLAC could be detected in the lungs of piglets in either of the groups (Fig. 3). FAB

J. N. Samsom and others

Fig. 4. Flow cytometric analysis of SLAC in the BALF of SPF piglets. SPF piglets were inoculated i.n. with 105 TCID50 PRRSV strain ter Huurne. On various days post-infection the piglets were sacrificed and the lungs removed and lavaged semiquantitatively. The BALF cells were stained with MAbs against various surface antigens, as indicated in Methods. The SLAC were selected and analysed. Dots represent means of two animals per time-point ; values for uninfected animals are not shown but were comparable to the values on day 0. (*) indicates P 0n05 compared to t l 0.

SPF piglets. During PRRSV infection in SPF piglets changes

in populations of SLAC in the BALF were comparable to those seen for gnotobiotic piglets. The percentages of SWC1+517n2L− and SWC1+CD14− SLAC slightly increased during the early stages of infection and remained constant thereafter. In contrast, the percentages of SWC1+517n2L+, SWC1+CD14+ and 517n2L+CD11b+ SLAC decreased slightly during the first 14 days of infection and remained almost constant during the following weeks (Fig. 4). The percentage of 517n2L−CD11b+ SLAC increased during the first 10 days of infection but returned to basal levels thereafter (Fig. 4). Analysis of lymphocytic markers revealed that CD2+CD3+ SLAC dramatically increased starting on day 10 of infection, peaked on day 21 and decreased thereafter (Fig. 4). In addition, the percentage of CD2+CD3− SLAC increased during the first week of infection and remained almost constant during the days following (Fig. 4). Most prominent was the change in the percentage of CD8+CD4− SLAC, which peaked on day 21 of FAC

infection at 40 % above the level of day 0 (Fig. 4). Further analysis of the CD8+ SLAC revealed that the percentage of CD8+CD6+ increased during the first 7 days of infection and reached a maximum of 20 % above base level. After day 7 of infection the percentage of CD8+CD6− SLAC also increased to a maximum of 20 % above base level (Fig. 4). The percentages of CD8+CD5+low SLAC, CD8+γδTCR− SLAC and CD3+γδTCR− SLAC followed a pattern similar to that of the CD8+CD4− SLAC (Fig. 4). Cytospin staining and histology

Cytospin preparations of BALF cells from gnotobiotic piglets isolated on day 5 of infection were stained with a MAb against the nucleocapsid protein of the European PRRSV strain ter Huurne. The mean percentage of LV-nucleocapsid-positive cells was 1n91p3n13 (n l 9), whereas no positive cells could be detected in control piglets.

CD8+ cells in the lungs of pigs with PRRS

Fig. 5. Number of CD8+ cells per mm2 tissue section in the lungs of PRRSV-infected gnotobiotic piglets. One-week-old gnotobiotic piglets were inoculated i.n. with 0n5 ml 105 TCID50 PRRSV strain ter Huurne. At day 5, of infection control ( ) and infected ( ) piglets were sacrificed and the lungs were removed. A segment of the frontal lobe (A), cardiac lobe (B) and dorsal lobe (C) of the right lung-half was cryosectioned and stained with a MAb against CD8. The number of CD8+ cells was assessed microscopically. Data are means (SD ; n l 9).

To substantiate our flow-cytometric findings, which suggest that an influx of CD8+ cells occurs in the alveolar lumen, histology was used to determine whether increased numbers of these cells could also be found in the lung tissue. Tissue sections from three different areas in the right lung-half of control and PRRSV-infected piglets were stained with an anti-CD8 MAb on day 5 of infection. In all three different areas of the lung the number of CD8+ cells\mm# was significantly higher in PRRSV-infected piglets than in control piglets (Fig. 5).

Discussion Our study demonstrates that during a PRRSV infection in pigs a strong influx of natural killer cells and cytotoxic Tlymphocytes occurs in the lungs. Support for this conclusion is provided by the following observations. Semi-quantitative lavage of the lungs of PRRSVinfected piglets revealed an increase in the total number of BALF cells between days 10 and 21 of infection. Subsequent flow cytometric analysis of BALF cells showed that the percentage of CD8+ cells increased dramatically during infection, with a peak between days 14 and 21 when approximately 30 % of the total number of BALF cells consisted of CD8+ cells. This increase in CD8+ BALF cells was observed in both gnotobiotic and SPF piglets with a PRRSV infection,

whereas it was not seen prior to infection or in control piglets. Further analysis of the CD8+ cells revealed that during the first days of infection mainly CD8+CD6− cells were detected in the lungs of gnotobiotic and SPF piglets. As described by others this phenotype corresponds to that of MHC-non-restricted cytolytic cells (Pauly et al., 1996). Double staining against CD8 in combination with other markers revealed that these cells were mostly CD2+, CD4−, CD5− or CD5+low but γδTCR−, indicating that the phenotype of the CD8+CD6− cells corresponds to a natural killer cell and not to the other cytolytic CD8+γδ+T-cells. After day 7 of infection an additional increase in the percentage of MHC-restricted cytotoxic T-cells in the lungs occurs. This is demonstrated by the presence of CD8+ cells which stained positive with anti-CD6 (Pauly et al., 1996) and by the absence of CD8+CD4+, CD8+γδTCR+ cells. The increase in the percentage of MHC-restricted cytotoxic T-cells in the lungs starting on day 7 of infection correlated precisely with a rapid decrease of the amount of BALF cellassociated PRRSV. This finding strongly suggests that the presence of cytolytic cells in the lungs during a primary infection is protective. Cytotoxic T-cells and natural killer cells are potent at lysis of infected cells and may thus prevent spread of the virus (Kimman et al., 1996). Furthermore, both types of cell have been shown to regulate cellular immunity via the production of interferon-γ (Trinchieri, 1995). Future research will focus on directly demonstrating the protective role of cytolytic cells during the host defence against PRRSV. It should be noted that there was no increase in the percentage of other types of lymphocyte, such as CD4+CD8−, CD4+CD8+ or other myeloid cells such as polymorphonuclear cells (data not shown), in the lungs of PRRSV-infected pigs. This finding seems to indicate a response to very selective chemotactic signal. Based on the findings that PRRSV causes apoptosis of infected macrophages (Suarez et al., 1996) and that PRRSVinfected pigs from the field often carry secondary infections, it has been suggested that infection with the virus may cause immunosuppression in its host due to a decrease in the number of macrophages in the lungs (Molitor et al., 1992). Our study clearly demonstrates that overall the number of macrophages in BALF does not decrease during a PRRSV infection in gnotobiotic or SPF piglets. However, the percentage of macrophages does decrease during PRRSV infection. Support for these conclusions is provided by the finding that semiquantitative broncho-alveolar lavage yielded either constant (gnotobiotic piglets) or increasing (SPF piglets) numbers of LHAC, i.e. macrophages, during infection. The percentage of alveolar macrophages decreased steadily during infection due to a large influx of SLAC, i.e. lymphocytes and monocytes. The latter finding agrees with Shibata et al. (1997) who performed a microscopical evaluation of BALF cells from PRRSV-infected pigs and demonstrated that the ratio of macrophages decreased during infection whereas the ratio of lymphocytes gradually increased and the ratio of neutrophils remained unchanged. It FAD

J. N. Samsom and others

should be noted that on day 2 of infection in SPF piglets the total number of BALF cells does decrease slightly. However, it is unlikely that this small change could cause continuous immunosuppression. In this study two infection models were used to assess whether differences in the microbiological status of piglets affects the changes in BALF cell population during PRRSV infection. In general, we did not identify different responses in the two models. The influx of CD8+CD6− natural killer cells during the first days of a PRRSV infection was seen in both models. Unfortunately, we cannot conclude that responses in the lungs in these two models do not differ because we have no data on PRRSV infection in gnotobiotic piglets during later times of infection. In summary, our findings demonstrate that during a PRRSV infection the total number of BALF cells increases mainly due to an influx of natural killer cells and cytotoxic T-lymphocytes. The number of macrophages in the lung does not decrease during infection. The authors wish to express their gratitude to Dr A. Saalmu$ ller for his comments and suggestions for Fig. 1. We also thank Dr M. Denyer for providing the C25 MAb and E. P. de Kluijver, G. Kok and R. Autar for technical assistance.

respiratory syndrome virus antigen in porcine lung. Journal of Veterinary Diagnostic Investigations 6, 254–257. Halbur, P. G., Paul, P. S., Meng, X. J., Lum, M. A., Andrews, J. J. & Rathje, J. A. (1996). Comparative pathogenicity of nine US porcine

reproductive and respiratory syndrome virus (PRRSV) isolates in a fiveweek-old cesarean-derived, colostrum-deprived pig model. Journal of Veterinary Diagnostic Investigations 8, 11–20. Hill, H. (1990). Overview and history of mystery swine disease (Swine Infertility Respiratory Syndrome). In Proceedings of the Mystery Swine Disease Committee Meeting (pp. 29–31), October 6 1990, Denver, Colorado. Madison, WI, USA : Livestock Conservation Institute. Kimman, T. G., De Bruin, T. G. M., Voermans, J. J. M. & Bianchi, A. T. J. (1996). Cell-mediated immunity to pseudorabies virus : cytolytic

effector cells with characteristics of lymphokine-activated killer cells lyse virus infected and glycoprotein gB- and gC-transfected L14 cells. Journal of General Virology 77, 987–990. Lindhaus, W. & Lindhaus, B. (1991). Ra$ tselhafte Schweinekrankheit. Der Praktische Tierarzt 5, 423–425. Mengeling, W. L., Lager, K. M. & Vorwald, A. C. (1995). Diagnosis of porcine reproductive and respiratory syndrome. Journal of Veterinary Diagnostic Investigations 7, 3–16. Meulenberg, J. J. M., Hulst, M. M., de Meijer, E. J., Moonen, P. L. J. M., de Besten, A., de Kluyver, E. P., Wensvoort, G. & Moormann, R. J. M. (1993). Lelystadvirus, the causative agent of porcine epidemic abortion

and respiratory syndrome (PEARS), is related to LDV and EAV. Virology 192, 62–72. Molitor, T. W., Leitner, G., Choi, C. S., Risdahl, J., Rossow, K. D. & Collins, J. E. (1992). Does SIRS virus cause immunosuppression?

References

Proceedings of the International Symposium on SIRS\PRRS\PEARS, p. 20.

Berndt, A. & Mu$ ller, G. (1997). Heterogeneity of porcine alveolar

Paton, D. J., Brown, I. H., Edwards, S. & Wensvoort, G. (1991). Blue ear

macrophages in experimental pneumonia. Veterinary Immunology and Immunopathology 57, 279–287.

Choi, C., Gustafson, K., Chinsakchai, S., Hill, H. & Molitor, T. (1994).

Heterogeneity of porcine alveolar macrophage subpopulations : immune functions and susceptibility to PEARS virus. Proceedings of the 13th IPVS Congress, Bangkok, Thailand, 26–30 June 1994, pp. 97. Collins, J. E., Benfield, D. A., Christianson, W. T., Harris, L., Hennings, J. C., Shaw, D. P., Goyal, S. M., McCullough, S., Morrison, R. B., Joo, H. S., Gorcyca, D. & Chladek, D. (1992). Isolation of swine infertility

and respiratory syndrome virus (isolate ATCC VR-2332) in NorthAmerica and experimental reproduction of the disease in gnotobiotic pigs. Journal of Veterinary Diagnostic Investigations 4, 117–126. Conzelmann, K. K., Visser, N., van Woensel, P. & Thiel, H. J. (1993).

Molecular characterization of porcine reproductive and respiratory syndrome virus, a member of the arterivirus group. Virology 193, 329–339. Cromwijk, W. A. J. (1991). Nieuwe virusziekte bij varkens. Tijdschrift voor Diergeneeskunde 116, 247–248. Duan, X., Nauwynck, H. J. & Pensaert, M. B. (1997). Virus quantification and identification of cellular targets in the lungs and lymphoid tissues of pigs at different time intervals after inoculation with porcine reproductive and respiratory syndrome virus. Veterinary Microbiology 56, 9–19. Halbur, P. G., Paul, P. S., Andrews, J. J., Sanderson, T. P., Ross, R. F., Schwartz, K. J., Frey, M. L., Erickson, B. J., Hill, H. T. & Hoffman, L. J. (1993). Experimental transmission of an apparent viral pneumonia in

conventional and gnotobiotic pigs. Veterinary Record 132, 263–266. Halbur, P. G., Andrews, J. J., Huffman, E. L., Paul, P. S., Meng, X.-J. & Niyo, Y. (1994). Development of a streptavidin–biotin immuno-

peroxidase procedure for the detection of porcine reproductive and FAE

disease of pigs. Veterinary Record 128, 617. Pauly, T., Weiland, E., Hirt, W., Dreyer-Bux, C., Maurer, S., Summerfield, A. & Saalmu$ ller, A. (1996). Differentiation between

MHC-restricted and non-restricted porcine cytolytic T-lymphocytes. Immunology 88, 238–246. Plagemann, P. G. W. & Moening, V. (1992). Lactate dehydrogenaseelevating virus, equine arteritis virus, and simian hemorrhagic fever virus. Advances in Virus Research 41, 99–192. Pol, J. M. A. & Wagenaar, F. (1992). Morphogenesis of Lelystadvirus in porcine alveolar macrophages. American Association of Swine Practitioners Newsletter 4(4), 29. Pol, J. M. A., Van Dijk, J. E., Wensvoort, G. & Terpstra, C. (1993).

Pathological, ultrastructural and immunohistochemical changes caused by Lelystad virus in experimentally induced infections of mystery swine disease [synonym : porcine epidemic abortion and respiratory syndrome (PEARS)]. Irish Veterinary Journal 46, 73–77. Rossow, K. D., Bautista, E. M., Goyal, S. M., Molitor, T. W., Murtaugh, M. P., Morrison, R. B., Benfield, D. A. & Collins, J. E. (1994).

Experimental porcine reproductive and respiratory syndrome virus infection in one-, four-, and 10 week old pigs. Journal of Veterinary Diagnostic Investigations 6, 3–12. Shibata, I., Mori, M., Uruno, K., Samegai, Y. & Okada, M. (1997). In vivo replication of porcine reproductive and respiratory syndrome virus in swine alveolar macrophages and change in the cell population in broncho-alveolar lavage fluid after infection. Journal of Veterinary Medical Science 59, 539–543. Suarez, P., Diaz-Guerra, M., Prieto, C., Esteban, M., Castro, J. M., Nieto, A. & Ortin, J., (1996). Open reading frame 5 of porcine

reproductive and respiratory syndrome virus as a cause of virus induced apoptosis. Journal of Virology 70, 2876–2882.

CD8+ cells in the lungs of pigs with PRRS Tong, S. L., Stueckemann, J. A. & Plagemann, P. G. W. (1977).

Voicu, I. L., Silim, A., Morin, M. & Elazhary, A. S. Y. (1994). Interaction

Autoradiographic method for detection of lactate dehydrogenaseelevating virus infected cells in primary macrophage cultures. Journal of Virology 22, 219–227. Trinchieri, G. (1995). Natural killer cells wear different hats : effector cells of innate resistance and regulatory cells of adaptive immunity and of hematopoiesis. Seminars in Immunology 7, 83–88. Van Reeth, K. & Pensaert, M. (1997). Comparative profile of proinflammatory cytokines following infection with different porcine respiratory viruses. Fourth International Congress of Veterinary Virology (p. 32), Edinburgh, UK, 24–27 August 1997.

of porcine reproductive and respiratory syndrome virus with swine monocytes. Veterinary Record 134, 422–423. Zhou, Y., Barghusen, S., Choi, C., Rossow, K., Collins, J., Laber, J., Molitor, T. & Murtaugh, M. (1992). Effect of SIRS virus infection in

leukocyte populations in the peripheral blood and on cytokine expression in alveolar macrophages of growing pigs. American Association of Swine Practitioners Newsletter 4(4), 28.

Received 1 June 1999 ; Accepted 5 October 1999

FAF