When PRRSV replication was assessed in MARC-145 cells at 20–22 h p.i., only a small proportion of cells expressed PRRSV antigen, as determined by FA (<5% in > 10 manual experiments counting fluorescent-positive cells under the microscope; Figure 1 by flow cytometry). However, between 42–72 h p.i. the percent of PRRSV-positive cells increased rapidly as determined by flow cytometric analyses, up to a maximum of about 95% by 96 h p.i. (Figures 1 &2). Low permissiveness of MARC-145 cells to primary (less than about 22 h p.i.) PRRSV infection was not due to insufficient M.O.I., since inoculation of cultures with about 100-times the standard dose resulted in a maximum 5.5% incidence of PRRSV-positive cells at 22 h p.i. by flow cytometry (data not shown). Propidium iodide staining followed by flow cytometry showed that PRRSV inoculated cells underwent major changes in cell cycle parameters (e.g. drop in G1 and G2; increased debris and aggregates) between 72–96 h p.i., consistent with cell damage and the spread of microscopic foci (CPE) typically observed throughout the cultures (data not shown).

FA analyses by confocal fluorescence microscopy, of dozens of independent PRRSV-infected confluent cultures, revealed that the primary (< 22 h p.i.) phase of PRRSV infection was characterized by random targeting of PRRSV-permissive single cells (Figure 3A; see also Figure 5B for another example), representing only a few percent of the total cells, and consistent with the quantitative analyses by flow cytometry.

During the logarithmic phase of viral replication (e.g. 42–48 h p.i.), infection was present mainly in clusters containing multiple PRRSV-positive MARC-145 cells, routinely observed against a PRRSV-negative background of confluent cells (examples indicated by arrows in Figures 3B,C,D; see also Figures 5C,D and 6B). These clusters were never observed in the primary infection analyses, although occasionally infected cell doublets were seen at 20–22 h p.i.. Based on >10 independent analyses, we estimate that about 90% of secondary PRRSV infection of MARC-145 cells line is characterized by formation of infected cell clusters.

Typically, dozens of clusters were seen in each culture, often containing a relatively central, bright-staining cell (e.g. Figure 3B,D), suggesting that cell-to-cell spread originated from a single PRRSV-infected reservoir cell. Absence of cluster formation during primary infection did not appear to be due to insufficient M.O.I. or late-stage development of a soluble cluster-inducer during culture, since using up to about 100-times the standard virus dose (obtained from 45–96 h p.i. cell supernatants) did not induce cluster formation at 20–22 h p.i. in our experimental system (data not shown), and cluster formation was also density-dependent (see below – Figure 8B). Clusters of PRRSV-positive cells were maintained for up to 72 h p.i. as illustrated in Figure 3D, and images at this time were also suggestive of maintenance of central bright-staining cells as reservoirs of virus.

Treatment of cells with 10 μM colchicine simultaneously with PRRSV inoculation, resulted in about 75% inhibition of secondary PRRSV infection (e.g. 70% in control vs. 18% in colchicine-treated culture; counting >1000 total cells at 46 h p.i.) and data representative of numerous experiments are shown in Figure 5E to illustrate the inhibitory effect of colchicine on formation of PRRSV-infected secondary clusters, which are seen spreading throughout the controls (Figure 5C,D). Similarly, exposure of cells to cytochalasin D under several experimental conditions (1 or 2 μM; administered either simultaneously with virus infection or 2 h pre-infection) inhibited PRRSV-positive cells by up to 87% (illustrated in Figure 8A between 13–22 h p.i.; 2 h pre-treatment with 1 μM). These drug effects were confirmed by flow cytometric analyses (about 50–90% inhibition of PRRSV-positive cells; data not shown).

To further evaluate PRRSV transmission in vitro, actin expression was imaged using Alexa Fluor 594-phalloidin (red). Similar patterns of actin expression were observed in control uninfected (Figure 5A) and PRRSV-infected (Figure 5B) cells, and all cells expressed some degree of actin when high levels of gain were applied. Cells often displayed fibrillar actin extensions appearing along the outer membrane as well as less-elongated extensions over the body of the cell. Effects on the actin staining pattern were apparent after exposure of MARC-145 cells to the actin disruptor cytochalasin D (Figure 6F), and also to some extent after exposure to the microtubule inhibitor colchicine (Figure 5E) although intact actin fibrils were still present (see also description of Figure 6A,C,D below). Simultaneous determination of actin expression and PRRSV infection demonstrated that relatively high expression of actin filaments correlated with PRRSV resistance, which was a consistent finding in dozens of control- and colchicine-treated experiments. This observation is illustrated in Figure 6A,C, &6D, where the arrows indicate the location of a PRRSV-positive doublet with low actin expression. Figures 5B and 6E also illustrate the negative correlation between actin and PRRSV-antigen expression. While this was not an absolute correlation (some PRRSV-positive cells did express high levels of actin), the trend was clear from >5 independent experiments. Actin fibrils also appeared to partition PRRSV-positive from PRRSV-negative cells and were often observed surrounding PRRSV-negative cells (Figures 6G and 7A–C; and see below). Cytochalasin D added at 18 h p.i. also suppressed secondary PRRSV-positive cluster formation (Figure 6F; 41 h p.i.; see emerging clusters in the control for this experiment Figure 6E), demonstrating a role for the cytoskeleton in cell-to-cell transmission.

The confluent cultures used in our experiments generally had mean cell-to-cell distances of < 2 μm, and the cells were usually in contact or too close to measure meaningful distances. Due to the wide dynamic range of the fluorescence signals, this is not always apparent in the figures, but some figures, such as 3A, 7A &7B, serve to illustrate the background of cultured cells. By seeding plates with different numbers of cells and then infecting these cultures containing different cell densities, it was possible to measure cell-to-cell distances in hundreds of cells per culture from the captured images, and then correlate the mean with the number of PRRSV-positive clusters (Figure 8B). The results show that formation of PRRSV-positive secondary clusters was density-dependent, with a direct correlation between the cell-to-cell mean distance and the number of clusters (Figure 8B).

In contrast to the ability of PRRSV to spread via cell-to-cell transmission, LDV did not exhibit this property, nor was there any dissociation of actin expression by primary cultured mouse macrophages from LDV permissiveness (probed during the replication peak, Figure 7D &7E). Also in contrast to the response of PRRSV in MARC-145 cells, treatment of mouse macrophages with the same concentrations of colchicine had little effect on LDV replication (+17%, -26%, and -21% LDV-positive cells relative to control in three separate experiments), although LDV infection was completely suppressed by cytochalsin D (no LDV-positive cells detected in three separate experiments).

Combined, these data show that after primary acute PRRSV infection of a small subpopulation of PRRSV-permissive MARC-145 cells, the virus spreads secondarily over the next 2–3 days to surrounding cells by cell-to-cell transmission. These virus mechanisms are both actin- and tubulin-dependent, demonstrating the critical role of the cytoskeleton in the processes of PRRSV infection and spread. In contrast, the related arterivirus LDV displays no cell-to-cell transmission in primary culture mouse macrophages and primary LDV replication is not tubulin-dependent as revealed by absence of colchicine sensitivity.

Pretreatment of MARC-145 cells for about 18 h prior to PRRSV inoculation with AK-2 suppressed primary (Figure 2B) and secondary (Figure 2D) virus infection as assessed by flow cytometry. The pharmacodynamics of AK-2 inhibition of primary PRRSV infection were determined by microscopic FA analyses (Figure 9), demonstrating that the antiviral effect of AK-2 is on the PRRSV-permissive state rather than directly on the virus, since pretreatment was required to fully establish PRRSV resistance. No detectable morphological effects of AK-2 on MARC-145 cells were noted in our studies, and the PI profiles of treated cells were similar to those of control cells demonstrating intact and metabolically viable cells at 22–46 h p.i. (data not shown). AK-2 pretreatment completely inhibited the formation of secondary clusters of PRRSV infection, and only occasionally were single-positive cells observed at 42–46 h p.i. in AK-2-pretreated MARC-145 cells (Figures 3E &4C). However, when added at 20 h p.i., AK-2 only partially inhibited secondary (cluster) PRRSV spread, although there was a shift to single-cell PRRSV infection (Figure 3F). Partial inhibition of secondary PRRSV infection by delayed AK-2 addition was also observed by flow cytometry, since AK-2 added at 18 h p.i. inhibited the 42 h p.i. expression of PRRSV antigen, by about one-half (27% vs. 58% in the drug-vehicle control; Figure 2F). Exposure of MARC-145 cells to AK-2 simultaneously with PRRSV inoculation suppressed all secondary cluster formation at 45 h p.i. (see control in Figure 4E and AK-2-treated in Figure 4F), confirming that, even without pretreatment, AK-2 can suppress the viral permissiveness of target cells for secondary cell-to-cell transmission, while leaving primary single-cell infection relatively intact due to the time required for induction of PRRSV resistance.

A higher-magnification image of AK-2-treated (single-cell infection only) is shown in Figure 5F, which also illustrates the often-observed appearance of actin fibrils surrounding a PRRSV-positive cell, as if to separate it from other uninfected cells as indicated above for Figure 6G. The data also show the absence of a detectable effect of AK-2 on actin expression.

In other studies, AK-2 pretreatment inhibited PRRSV antigen expression at 20–22 h p.i. by about 90% in primary pig macrophages, in each of two experiments. Similarly, pretreatment of primary mouse macrophages with recombinant murine AK-2 inhibited LDV replication by 86% in a single experiment. Thus, the anti-arterivirus effects of AK-2 are expressed over a broad host cell range.

Observations were also made for the anti-PRRSV effect of IFN-γ pretreatment, which was demonstrated in previous studies to inhibit PRRSV replication 14. PRRSV inhibition by IFN-γ was less effective than that of AK-2 under the same experimental conditions, as seen for example in the primary response to pretreatment measured by flow cytometry (Figure 2C,I); the effect of IFN-γ pretreatment also waned by 40 h p.i. (Figure 2J). FA analyses confirmed the reduced efficacy of IFN-γ relative to AK-2, since IFN-γ pretreatment did not completely inhibit formation of secondary PRRSV antigen-positive clusters (Figure 4A,D) and there were higher numbers of PRRSV antigen-positive cells in IFN-γ-pretreated cultures (data not shown). Consistent with these data, flow cytometry demonstrated that IFN-γ added at 18 h p.i. had a reduced effect on secondary PRRSV infection (Figure 2G), relative to AK-2 (Figure 2F) or the combination of AK2 and IFN-γ (Figure 2H); the inhibitory effect of post-treatment with both drugs is also shown in Figure 2H, although little if any synergy was observed. Thus, while IFN-γ mediated significant inhibition of PRRSV infection, AK-2 appears to be a more potent anti-PRRSV agent.

A stable and mycoplasma-free MARC-145 cell line was utilized in these experiments. Cells were cultured in DMEM containing10% fetal bovine serum, and for virus infections the medium was switched to MEM containing 2% horse serum. Cells for virus infections were grown to confluency in either T-25 flasks (seeded with about 5 × 10⁵ cells/culture) or 8-well glass slide chambers (seeded with about 10,000 cells/culture; Lab-Tek II; Nalge Nunc International), and for the cell density studies, serial two-fold dilutions of the cells were used. Cells were inoculated with PRRSV at about 1–2 days after seeding (time to approximate doubling of the population).

Pig cells were collected from 4–8 week old pigs by lung lavage with PBS 212223. Cells were cultured in DMEM containing 10% FBS. After 18–24 h, non-adherent cells were removed by washing. The remaining adherent cells were cultured for an additional 24 h in RPMI containing 2% horse serum, and then inoculated with PRRSV (M.O.I. approximately 0.1 TCID₅₀).

PRRSV isolate SD-23983 was passaged on MARC-145 cells, preparing high-titer (~ 10⁵ TCID₅₀ per ml) virus stocks from culture supernatants at 48–96 h p.i. PRRSV stocks were sequentially filtered through 0.45, 0.22, and 0.10 um filters and confirmed to be mycoplasma-free by testing on PPLO medium. As reported previously (9,14), maximum efficiency of PRRSV infection of MARC-145 cells occurs with low M.O.I as determined by TCID₅₀, probably due to the presence of multiple virions per TCID₅₀. For the present studies, M.O.I. of about 0.01 TCID₅₀ (slide cultures) and 0.001 TCID₅₀ (T-flask cultures) were found to result in near-optimal efficiency of infection, and were thus used for our studies unless otherwise noted in Results.

PRRSV replication was detected using FITC-labeled IgG anti-PRRSV nucleocapsid monoclonal antibody (SDOW17; 24). MARC-145 cells were cultured and inoculated with PRRSV in glass-bottom slide chambers, fixed in 80% acetone, and incubated for 1 h at 37°C with a 1:100 dilution of FITC-conjugated SDOW17 antibody made in PBS containing 5% fetal bovine serum. Then the cells were washed three times with cold PBS prior to examination under a fluorescence microscope, screening about 30–40,000 total cells to obtain the incidence of antigen-positive cells. Confocal fluorescence microscopy was performed using an Olympus BX61 microscope and Fluoview software. Images shown in Figures 3 and 4 display the yellow scale bar captured with the original image, along with a higher-contrast white scale line. For flow cytometry, MARC-145 cells were cultured and PRRSV-inoculated in T-flasks, the cells were suspended in trypsin-versene, pelleted at 1000 rpm, resuspended in DMEM with 2% horse serum, fixed in cold 80% acetone for 10 min, washed twice in PBS, and resuspended in 1 ml PBS containing 60 ul of fetal bovine serum. FITC-conjugated SDOW17 antibody was then added to the cells (2.5 ul/ml), incubation was carried out at 37°C for 60 min, the cells were washed with PBS, examined under a fluorescence microscope, and flow cytometry was performed with a FACSVantage SE (Becton Dickenson) equipped with a 488 Enterprise II coherent laser. Twenty thousand events per sample were analyzed with CellQuest software. Cell cycle analyses were also performed on the same samples, by staining with propidium iodide for 20 minutes at room temperature and analyzing the flow cytometric results with ModFit LT 2.0 software. PRRSV replication in primary pig alveolar macrophages 2223 was assessed by FA under a fluorescent microscope. Cellular expression of actin was determined by incubating acetone-fixed cells with AlexaFluor 594 phalloidin (Invitrogen) according to the manufacturer's instructions, with the modification of simultaneous PRRSV detection as above, such that combined labels were applied for 60 min at 37°C, permitting two-color fluorescence detection by confocal microscopy. Data shown are representative of at least 2 replicate experiments for each type of experiment described in the Results.

Peritoneal macrophages were collected from outbred ICR mice, seeded onto glass coverslips, and inoculated with a standard dose of LDV-P as described previously 25. LDV replication was assessed in cells fixed in acetone at 8 h p.i by IFA assay as described previously 25.

Purified recombinant human interferon-γ (IFN-γ; 100 ug/ml) and actokine-2 (AK-2; 50 ug/ml) were provided by Actokine Therapeutics. AK-2 is a cytokine-based experimental antiviral being developed by Actokine Therapeutics, which consists of recombinant normal human proteins comprising part of the mammalian cell response to virus infection (Wong; unpublished). Soluble stocks of these agents were stored at 4°C in fetal bovine serum, which also served as the drug-vehicle control for the experiments, and were diluted 1:50 or 1:100 in medium to yield concentrations in cell cultures of about 1–2 ug/ml. As noted for individual experiments, cells were exposed to the drug or control treatments prior to PRRSV infection (pretreatment), during the course of PRRSV infection (delayed or post-treatment), or simultaneously with PRRSV infection. Based on previous studies of in vitro efficacy, the microtubule inhibitor colchicine which binds to tubulin (Sigma; 5 or 10 μM; 26–28) or the microfilament disruptor cytochalasin D which depolymerizes actin (Sigma; 1 or 2 μM; 26293031) were added to cell cultures at the times indicated.