Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Lentiviral infections frequently involve the lungs (1), which are the site of typical severe lesions in some sheep late after natural or experimental infection by visna-maedi virus (2). These animals develop interstitial pneumonia with mural and luminal alveolitis consisting mainly of macrophages, neutrophils, and lymphocytes (3, 4), together with lymphocytic nodules and zones of smooth muscle hyperplasia (5). Virus is essentially restricted to alveolar macrophages (6), but only a few of them are actually infected (7). The uninfected majority produce mediators capable of initiating and maintaining the inflammatory reaction and neutrophil chemotaxis (3).

Interleukin (IL)-8, a CXC chemokine that recruits neutrophils, T lymphocytes, and basophils, is produced mainly by macrophages but also by lymphocytes, fibroblasts, and endothelial or smooth muscle cells on appropriate stimulation (8). The expression of IL-8 by macrophages is increased by diverse stimuli, including viral infection, lipopolysaccharide (LPS) from gram-negative bacterial cell walls, or binding of the proinflammatory cytokines IL-1 or tumor necrosis factor  (TNF) to their receptors (9). The mRNA for IL-8 is expressed at high levels in alveolar macrophages from naturally infected animals (10) or experimentally infected lambs, or in normal macrophages infected in vitro with visna-maedi virus (11).

Cells can respond to constituents in the external medium through specific receptors, which may have intrinsic kinase activity on activation by binding of the appropriate ligand, or may cooperatively activate soluble protein tyrosine kinases or G proteins (12, 13). These reactions, located near the internal face of the cytoplasmic membrane, initiate a cascade of sequential activation of cytoplasmic kinases which transmit signals to the nucleus, as elucidated for the mitogen-activated protein kinase pathway (12). Such pathways are probably involved in the events through which a small number of infected macrophages maintain an inflammatory response by a large majority of uninfected cells. We have therefore investigated both the direct effect of visna-maedi virus on alveolar macrophages and possible mechanisms of indirect activation of uninfected cells.

The induction of IL-8 mRNA and the accumulation of IL-8 in the medium of alveolar macrophages from uninfected sheep cultured in vitro with visna-maedi virus were unaffected by heat inactivation of the inoculum, suggesting that viral replication was not required for the initiation of the cellular response. The IL-8 mRNA induction in these cells was considerably reduced in the presence of genistein but not in that of staurosporine. This suggests that a tyrosine kinase-dependent signaling pathway was activated by contact between susceptible cells and virus, even if true infection did not result. The activation of neighboring naive macrophages might be a response to proinflammatory cytokines from the virally stimulated cells. We observed that IL-1 mRNA was induced rapidly after infection of alveolar macrophages, and that TNF mRNA, although significantly expressed in control cells, increased in quantity.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Virus

Visna-maedi virus strain K1514 (gift from Dr. Chappuis, Rhône Mérieux, France) was propagated in ovine skin fibroblasts (IDO5). After inactivation by heating to 56°C for 30 min, the virus produced no cytopathic effect on IDO5 cells over 3 wk culture, and no reverse transcriptase (RT) activity (14) in the supernatants. The amount of endotoxin contamination in virus preparations was checked using a Limulus amoebocyte assay (E-Toxate^®; Sigma, l'Isle d'Abeau, France).

Ovine Alveolar Macrophages

Macroscopically normal sheep lungs (n = 7) were collected from a commercial slaughterhouse over a period of several months. Lungs with macroscopic or histologic evidence of parasitic or bacterial infections, or with blood contamination at slaughter, were excluded. Cells obtained by bronchoalveolar lavage (3) were counted in a hemocytometer and their viability (always > 90%) determined by trypan blue exclusion. Cells were tested for natural visna-maedi virus infection by polymerase chain reaction (PCR) using pol gene primers common to ovine and caprine lentiviruses (15) on cDNA prepared by reverse transcription of total cellular RNA, and positive samples were disregarded. Total cells were seeded at 5 × 10⁶ cells per well in six-well plates (Falcon 3046; Becton Dickinson, Le Pont de Claix, France) in 2.5 ml RPMI 1640 medium supplemented with 5% fetal calf serum (FCS) and antibiotics (11). After incubation at 37°C for 24 h, non-adherent cells were rejected and the adherent cells rinsed with phosphate-buffered saline and incubated for 4 h with fresh RPMI 1640 containing 2.5% FCS. Infectious K1514 virus (multiplicity of infection [m.o.i.]: 0.04), heat-inactivated virus, or fresh medium as a control (250 µl) was then added to each well and incubation continued at 37°C in humidified 5% CO₂-enriched air until supernatant collection. Supernatants from paired wells were pooled after 0, 3, 6, 9, 24, and 48 h for protein determination and the washed cells were stored at 80°C for RNA extraction. For kinase inhibition, cells were incubated for 2 h before treatment with virus or mock infection with either genistein at 50 or 150 µM (Sigma; stock solution at 74 mM in dimethyl sulfoxide [DMSO]) or staurosporine at 10 or 200 nM (Sigma; stock solution at 214 µM in DMSO); RNA was extracted from cells collected 3 h after infection.

Assessment of Cytokine Production

In the absence of specific kits for dosage of ovine IL-8, IL-1, and TNF, we tested human-based kits for cross reactivity. The IL-8 EASIA^TM kit from Medgenix (Fleurus, Belgium) detected ovine IL-8 satisfactorily, but none of the tested human kits detected ovine IL-1 or TNF. The IL-8 kit was recalibrated with dilutions of supernatants from LPS-stimulated ovine macrophages as recently described (10). Results are expressed as picograms (human IL-8 equivalent) per milliliter.

Northern Blot Analysis of IL-8 mRNA Expression

Total RNA was extracted from the frozen cells after thawing in a denaturing solution containing guanidinium iso-thiocyanate as described by Schomczynski and Sacchi (16), and Northern blots were produced as described previously (11). Briefly, approximately equal quantities of total RNA, evaluated by ethidium bromide staining, were migrated through formaldehyde gels with Escherichia coli 23s and 16s RNAs as size markers, then transferred to duplicate nylon membranes (Amersham, Les Ulis, France). One membrane was incubated for 12 h with a ³²P-deoxycytidine triphosphate (Amersham) random-labeled cDNA probe for ovine IL-8 (11), the other with a similarly labeled cDNA probe for ovine glyceraldehyde 3-phosphate dehydrogenase (G3PDH; 17). The washed membranes were autoradiographed (12 h exposure) and the negatives scanned (DeskTopScanning Densitometer^®; PDI/Pharmacia Biotech, St. Quentin-en-Yvelines, France). Band intensity for each hybridization was evaluated using Quantity One^® software (PDI/Pharmacia), and results are expressed as the ratio of IL-8 to G3PDH signal intensities multiplied by 100.

Semiquantitative RT-PCR Evaluation of IL-1 and TNF mRNAs

In the absence of homologous ovine probes for IL-1 or TNF, and with doubts as to the easy use of the human equivalents, we estimated the amounts of the corresponding mRNAs by RT-PCR under conditions allowing semiquantitative evaluation. Total RNA, prepared as above, was dissolved in 20 µl diethyl pyrocarbonate-treated water. First-strand cDNA was obtained by reverse transcription of 2 µg of cellular RNA in a final volume of 40 µl (15). Cytokine cDNAs were amplified from 1:5 dilutions of the reverse-transcribed product using primers selected to cover intron/exon boundaries in ovine IL-1, TNF, and G3PDH sequences (Table 1; synthesized by Eurogentec, Seraing, Belgium). Parallel amplification of 10-µl aliquots of diluted cDNA was performed using each primer pair under mineral oil in a final volume of 50 µl under conditions selected to obtain band intensities in the mid-linear portion of the response curve (annealing temperature 55°C, 25 cycles for G3PDH; annealing temperature 52°C, 27 cycles for IL-1 and TNF). Products were electrophoretically migrated through ethidium bromide gels, and photographs (Polaroid^® 55) were taken under ultraviolet transillumination. Photographic negatives were scanned and band intensity evaluated as above. For each cytokine, results are expressed as the ratio of specific cytokine to G3PDH signal intensities multiplied by 100. 

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Kinetics of IL-8 Protein Release and mRNA Accumulation Induced by Visna-maedi Virus in Ovine Alveolar Macrophages

Following preliminary observations that IL-8 mRNA was present in increased quantities in sheep alveolar macrophages after infection in vitro with visna-maedi virus (11), we confirmed that this increase is reflected by the secretion of IL-8 into the extracellular medium, and established the kinetics of specific mRNA induction. No IL-8, as measured by a human enzyme-linked immunosorbent assay (ELISA), was ever detected in more than trace amounts in uninfected control supernatants, or at time 0 of the infected samples. Measurable quantities appeared rapidly after infection, and the concentration of IL-8 present in the supernatant rose quickly over the first 9 h. Thereafter, IL-8 continued to accumulate more slowly, reaching levels of more than 15 × 10³ pg/ml (human equivalents) at 48 h after infection (Figure 1). The concentration of IL-8 measured in the supernatants of cells from seven individual sheep was very similar, although several months separated their collection and assessment.

View larger version (19K): [in this window] [in a new window]   Figure 1.   IL-8 release from sheep alveolar macrophages. Cells were incubated with either infectious (0.04 m.o.i.) (filled squares), or heat-inactivated visna-maedi virus (open squares), or with medium alone (filled triangles). Supernatants from four different experiments were collected 0, 3, 6, 9, 24, and 48 h after infection and IL-8 release was measured by ELISA. Results are expressed as mean ± SEM in picograms of human IL-8 equivalent per milliliter.

IL-8 mRNA, measured by Northern blotting and densitometry, was generally present at low levels in control cells although it sometimes increased transiently, often following the mock infection. Nevertheless, infected cells always showed clearly increased amounts with an apparently biphasic pattern. A sharp initial increase followed by a drop in expression, occurring between 3 and 9 h after infection, preceded a slower rise, continuing at least until the last observation at 48 h after infection (Figure 2). The relative increase of specific IL-8 mRNA in infected cells ranged from about 30-fold, where control cells expressed low basal levels, to at least 2-fold, when control cell expression was transiently high (Figure 2; 6 h). Maximum mRNA expression in the initial peak occurred at 6 h for five of seven samples, and as early as 3 h for the two others. These increases could not be attributed to bacterial endotoxins because the Limulus assay never detected more than 0.03 units/ml, well below the threshold for IL-8 mRNA induction in sheep macrophages (not shown). Interestingly, the kinetics of IL-8 secretion (Figure 1) or IL-8 mRNA accumulation (Figure 2) were virtually unchanged when the viral inoculum was mildly heat-inactivated, with complete loss of infectivity in susceptible tissue cultures. Similar results were obtained with all six sheep alveolar macrophage samples treated with inactivated virus. On the contrary, removal of structured viral particles by ultracentrifugation at 260,000 × g for 1 h almost abolished the capacity of the inoculum to induce IL-8 mRNA (Figure 3).

View larger version (20K): [in this window] [in a new window]   Figure 2.   IL-8 mRNA induction in sheep alveolar macrophages. Cells were incubated with either infectious (0.04 m.o.i.) (filled squares), or heat-inactivated visna-maedi virus (open squares), or with medium alone (filled triangles). Northern blot analysis was performed on total RNA extracted from cells harvested 0, 3, 6, 9, 24, and 48 h after infection. After quantification of band intensity obtained by Northern blot, results are expressed as IL-8 to G3PDH signal ratio × 100; plot is from one representative experiment.

View larger version (29K): [in this window] [in a new window]   Figure 3.   Loss of IL-8 mRNA induction after ultracentrifugation of viral preparation. Cells were incubated with infectious visna-maedi virus at 0.04 m.o.i. (lane A); the same amount of viral stock solution freed of viral particles by ultracentrifugation (lane B); and medium alone (lane C). Northern blot analysis was performed on total RNA extracted from cells harvested at 48 h. Hybridization with the G3PDH probe demonstrates equal transfer of total RNA.

Effect of Protein Kinase Inhibitors

The similarity of IL-8 induction in ovine alveolar macrophages, by replication-competent or -incompetent visna-maedi virus, led us to investigate possible signaling consequent to activation of cellular receptors by viral components. Such signaling classically involves reversible phosphorylation of cytoplasmic target molecules by tyrosine kinases or by serine-threonine kinases, such as protein kinase C. Specific inhibition of serine-threonine kinases by pretreatment of cells with staurosporine at 10 or 200 nM left induction of IL-8 mRNA by replication-competent or inactivated visna-maedi virus unchanged (data not shown). However, moderate doses of genistein, an inhibitor of tyrosine kinases, greatly reduced IL-8 mRNA levels without completely abrogating the response. Figure 4 shows that similar levels of reduction were obtained by pretreatment of the cells with 50 or 150 µM genistein, whether the viral inoculum was inactivated or not. A sub-optimal dose of 10 µM genistein did not affect IL-8 mRNA levels (not shown).

View larger version (21K): [in this window] [in a new window]   Figure 4.   IL-8 mRNA induction by visna-maedi virus in presence of a tyrosine-kinase inhibitor. Sheep alveolar macrophages were incubated with infectious virus at 0.04 m.o.i. (lanes A, B, and C), inactivated virus (lanes D, E, and F), or medium alone (control). Lanes A and D: without genistein; lanes B and E: with 50 µM genistein; lanes C and F: with 150 µM genistein. Northern blot analysis was performed on total RNA from cells harvested 3 h after infection. Hybridization with G3PDH probe demonstrates equal transfer of total RNA.

Induction of Proinflammatory Cytokine mRNAs

The initial peak of IL-8 mRNA production may be induced by the interaction between viral components and cellular receptors, but the later rise, particularly in the absence of viral replication, requires explanation. Macrophages produce IL-8 as a response to IL-1 or to TNF. Unfortunately, antibodies to the ovine molecules are not available, and we found no cross-reactivity with commercialized human products. The corresponding mRNAs were evaluated by RT-PCR. A clear early induction of IL-1 mRNA was observed in cells treated with replication-competent or inactivated virus, but not in uninfected cells (Figure 5). Control cells frequently contained a rather high basal level of TNF mRNA, but this rose moderately after treatment with the viral inoculum. Two of five samples had low basal levels of TNF mRNA and showed clear transient induction simultaneous with the early peak of IL-8 and of IL-1 (data not shown).

View larger version (20K): [in this window] [in a new window]   Figure 5.   IL-1 mRNA induction in sheep alveolar macrophages. Cells were incubated with either infectious (0.04 m.o.i.) (filled squares), or heat-inactivated visna-maedi virus (open-squares), or with medium alone (filled triangles). RT-PCR analysis was performed on the same RNA samples as in Figure 2. After quantification of band intensity obtained by RT-PCR, results are expressed as IL-1 to G3PDH signal ratio × 100; plot is from one representative experiment.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The specificity of the massive lung lesions occurring in some sheep late after infection with visna-maedi virus implies that this virus plays a determinant role in initiating the inflammatory process involved. Alveolar macrophages infected in vitro or collected from experimentally infected sheep express increased amounts of IL-8 mRNA (11), which is present in the majority of macrophages from lung lavage of naturally infected animals, whose lavage fluid also contains secreted IL-8 (10). The paucity of infected macrophages within the lesions suggests that other factors may amplify the inflammatory process. In the present study, we examined the capacity of ovine alveolar macrophages to secrete IL-8 after contact with infectious or heat-inactivated virus, and the kinetics of expression of the corresponding mRNA.

A rapid and sustained secretion of IL-8 appears to be programmed by a biphasic production of mRNA. IL-8 expression can be induced transiently in macrophages by a wide variety of stimuli, including cell adherence (18) and LPS stimulation (19). We delayed treatment of seeded cells for 24 h to avoid the adherence-related stimulation, and found LPS concentrations in culture supernatants to be constantly below minimum stimulatory levels. The observed expression apparently results from contact with viral components because virus-free supernatant was unable to induce IL-8 expression, whereas inactivated virus was as effective as the infectious mother stock. Our observations that genistein abrogates IL-8 mRNA induction indicate that a tyrosine-kinase signaling pathway is activated on interaction between visna-maedi virus and its host cell, although the nature of the cellular receptor used by visna-maedi virus is controversial (20, 21) and its signaling capacity unknown. Cellular signaling has been reported following interaction of infectious or inactivated human immunodeficiency virus-1 with CD4 (22).

The later and more persistent phase of IL-8 mRNA expression in response to treatment with visna-maedi is presumably driven by a different process. Heat-inactivated virus still induced the later phase of IL-8 mRNA expression, indicating that viral replication is not necessary and contrasting with observations in an epithelial cell line infected with respiratory syncytial virus (26). Addition of IL-1 or TNF to cultured human monocytes induces IL-8 mRNA after about 1 h (27), and antibodies to these two proinflammatory cytokines can abolish the second peak of IL-8 mRNA in stimulated human blood monocytes (28). Unfortunately, no available reagents permitted the measurement or specific blockage of ovine IL-1 or TNF, so we had to rely on mRNA quantitation by RT-PCR to estimate their participation. These mRNAs are rapidly induced and degraded (29), so their instantaneous concentration is a good reflection of gene induction. After contact with virus, the amount of specific mRNA increased 2- to 30-fold, according to the momentary levels measured in the controls, suggesting proinflammatory cytokine involvement. Moreover, tyrosine phosphorylation is necessary for induction of IL-1, IL-6, and TNF in monocytes (30). Since IL-8, IL-1, and TNF all possess NFB targets in their enhancers (31), a common signal could explain their simultaneous transient expression soon after contact with visna-maedi virus.

Experimental and natural infection in sheep result in a long-term increase of the expression of IL-8 mRNA by alveolar macrophages (10, 11) which probably results from several additive mechanisms. Besides the mechanisms proposed above, we have shown that expression of the Tat peptide in target cells directly upregulates IL-8 gene transcription (34). The two-way interaction between lentiviral infection of cells in the lung and their expression of inflammatory and immunomodulatory cytokines probably plays an important part in the development of lentiviral lung disease. The infection of sheep with visna-maedi virus allows a study of these interactions without the complications of immunodepression and consequent opportunistic infections.