Dissemination of Lymphocytic Choriomeningitis Virus from the Gastric Mucosa Requires G Protein-Coupled Signaling

Abstract

The gastric mucosa is an important portal of entry for lymphocytic choriomeningitis virus (LCMV) infections. Within hours after intragastric (i.g.) inoculation, virus appears in the gastric epithelia, then in the mesenteric lymph nodes and spleen, and then in the liver and brain. By 72 h i.g.-inoculated virus is widely disseminated and equivalent to intravenous (i.v.) infection (S. K. Rai, B. K. Micales, M. S. Wu, D. S. Cheung, T. D. Pugh, G. E. Lyons, and M. S. Salvato. Am. J. Pathol. 151:633–639, 1997). Pretreatment of mice with a G protein inhibitor, pertussis toxin (PTx), delays LCMV dissemination after i.g., but not after i.v., inoculation. Delayed infection was confirmed by plaque assays, by reverse transcription-PCR, and by in situ hybridization. The differential PTx effect on i.v. and i.g. infections indicates that dissemination from the gastric mucosa requires signals transduced through heterotrimeric G protein complexes. PTx has no direct effect on LCMV replication, but it modulates integrin expression in part by blocking chemokine signals. LCMV infection of macrophages up-regulates CD11a, and PTx treatment counteracts this. PTx may prevent early LCMV dissemination by inhibiting the G protein-coupled chemotactic response of macrophages infected during the initial exposure, thus blocking systemic virus spread.

Lymphocytic choriomeningitis virus (LCMV) is the prototype of the Arenaviridae, a family of ambisense, bisegmented RNA viruses. Although laboratory studies of this virus employ primarily parenteral routes of inoculation, virus ingestion via the gastric route is the probable natural route of infection (28). Our previous studies established that virus delivered intravenously (i.v.) appears in several different organs simultaneously, whereas virus delivered intragastrically (i.g.) appears first in the stomach by 12 h, then in the spleen by 24 h, then in the liver by 48 h, and finally, by 72 h, it is as widely disseminated as i.v.-inoculated virus (29). Based on these results we proposed that i.g. infection, unlike i.v. infection, is likely to disseminate in a cell-associated manner through the lymphatics. We used pertussis toxin (PTx), to disrupt mononuclear cell (monocytes and lymphocytes) movement in order to test whether a block to cell trafficking can delay LCMV dissemination. Our studies followed the observation that PTx inhibits reovirus i.g. dissemination (37) and extended this observation by linking the action of PTx to virus-mediated changes in adhesion molecules and describing a model for the events involved in virus dissemination.

PTx, a surface protein of Bordetella pertussis, catalyzes ADP ribosylation of heterotrimeric GTP-binding proteins and disrupts signal transduction (24). Cytokines and chemokines, e.g., interleukin-8, monocyte chemotactic protein (MCP)-1, lymphotactin, and fractalkine, are potent mediators of mononuclear cell migration (14, 20, 33, 38), and they signal through G protein-coupled seven-transmembrane receptors (9, 12). PTx blocks G-coupled signaling and thereby blocks the rapid changes in integrins posttranscriptionally activated by such signaling (4). PTx-mediated leukocytosis can also be attributed to inhibition of G protein-dependent cell migration (16, 22, 23, 36). We propose that LCMV is spread by migrating mononuclear cells after mucosal inoculation and that PTx delays viral dissemination by blocking cell migration. We show that LCMV infection up-regulates cell surface adhesion molecule CD11a, but PTx down-regulates its surface expression. We speculate that PTx is interfering with a mechanism the virus has evolved to influence its tropism.

MATERIALS AND METHODS

Virus inoculation and PTx treatment of mice.

Six-week-old male BALB/c mice were obtained from Harlan Sprague Dawley (Indianapolis, Ind.). LCMV (Armstrong 53b strain) was plaque purified and stored at 10⁷ to 10⁸ PFU/ml. Virus inocula of 10⁶ PFU in 0.1 ml of RPMI 1640 medium were administered to mice either i.v. (tail vein) or by gastric intubation as described previously (28).

Virus was titered by plaque assay on Vero cells (ATCC CCL-81) as described previously (8). Briefly, dilutions of homogenized splenocytes were incubated at 37°C, 5% CO₂ for 1 h with Vero cell monolayers grown in 6-well plates (Costar, Cambridge, Mass.). The plates were then overlaid with 1% agarose in minimal essential medium 199 containing 10% fetal calf serum (FCS) and incubated at 37°C, 5% CO₂ for 5 days. The wells were treated with 25% formaldehyde and stained with 0.1% crystal violet for 30 min. The agarose overlay was removed, and plaques were counted. Each dilution was done in duplicate, and the counts were averaged. Infectious centers were similarly determined by plating murine leukocytes at 10⁵ to 10⁶ cells per 50-mm-diameter plate containing Vero cells at 70% confluence.

Some of the mice received an i.v. injection of 25 ng of PTx (List Biologicals, Campbell, Calif.) per g of body weight in 0.1 ml of phosphate-buffered saline. Mock treatments utilized 0.1 ml of phosphate-buffered saline. Cultured spleen mononuclear cells were treated with 0.1 μg of PTx/ml, which is close to the estimated PTx concentration in vivo and well below its mitogenic concentration (>1.0 μg/ml) (18).

Detection of viral nucleic acid.

Three, 7, or 10 days after infection (or mock infection), the mice were sacrificed by cervical dislocation. Blood was collected from the hearts and kept in heparinized tubes on ice. Spleens, kidneys, and livers were removed for use in virus titration, RNA extraction, flow cytometry, and/or cell culture. The spleens were placed in RPMI 1640 containing 10% FCS and kept on ice. The kidneys and livers were immediately processed for extraction of viral RNA.

Kidneys and livers were homogenized and sonicated in guanidine isothiocyanate solution and then layered on a CsCl cushion for density gradient isolation of total RNA as described previously (5). RNA was extracted from small amounts of tissue or cell cultures by using TRIzol reagent (Life Technologies, Gaithersburg, Md.). Reverse transcription followed by PCR (RT-PCR) was performed with 4 μg of whole-cell RNA and 1 μg of random hexanucleotide primers (Promega, Madison, Wis.) denatured at 65°C for 2 min, annealed on ice, and then added to a solution containing 1 μl of 40-U/μl RNasin (Promega), 1 μl of deoxyribonucleotides at 1.25 mM each, 4 μl of 5× RT buffer (50 mM Tris-HCl [pH 8.3], 50 mM KCl, 10 mM MgCl₂, 10 mM dithiothreitol, 0.5 mM spermidine), and 1 μl of 23-U/μl avian myeloblastosis virus RT (Promega) in a total volume of 20 μl for 1 h at 42°C. RT-PCR template concentrations were used in the range in which they are linearly related to the amount of product. Amplification of the cDNA was performed as described previously (32) with two oligonucleotide primers that anneal to the LCMV glycoprotein (Gp) gene (5′-TCATCGATGAGGTGATCAAC-3′ and 5′-CTTGGTGAACTCTCTAGACT-3′). Another set of oligonucleotide primers that anneal to the dihydrofolate reductase (DHFR) gene (5′-CTCAGGGCTGCGATTTCGCGCCAAACT-3′ and 5′-TATCAGCCTCCGTCAAGACAAATGGTC-3′) served as an internal control (17). The oligonucleotide primers were made on an automated synthesizer (Gene Assembler Plus; Pharmacia LKB, Piscataway, N.J.). RT-PCR products were analyzed by 1% agarose gel electrophoresis.

In situ hybridization to detect virus replication.

At 24 and 48 h after infection, the mice were sacrificed and the livers, stomachs, spleens, kidneys, ilea, and mesenteric lymph nodes were collected. The tissues were fixed in 10% neutral phosphate-buffered formalin and embedded in paraffin, and 5-μm-thick sections were prepared.

In situ hybridization was performed as described previously (15) with digoxigenin-labeled RNA probes generated by SP6 or T7 polymerase transcription from the entire LCMV Gp gene. Briefly, tissue sections were hybridized with 1.5 ng of the riboprobes/ml at 52°C overnight, washed in 2× SSC–50% formamide solution (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and then in 2× SSC, and treated with RNase T1 and RNase A for 30 min at 37°C. The slides were blocked for 1 h with a buffer containing 2% horse serum, 150 mM NaCl, and 100 mM Tris, pH 7.4. After being blocked, the slides were incubated for 1 h with sheep anti-digoxigenin-alkaline phosphatase conjugate (Boehringer Mannheim, Indianapolis, Ind.) at a 1:500 dilution, rinsed in Tris, pH 7.4, and incubated overnight at room temperature with nitroblue tetrazolium-BCIP (5-bromo-4-chloro-3 indolylphosphate) (Vector, Burlingame, Calif.) substrate in the dark. The stained slides were rinsed in water, counterstained with nuclear fast red, dehydrated, and sealed under a coverslip. LCMV Gp sense and antisense probes were used, and uninfected tissues were included as controls.

Flow cytometry to detect cell surface adhesion molecules and lymphocyte subset distribution.

Peripheral blood mononuclear cells (PBMC) were obtained from heparinized blood by Ficoll-Hypaque density gradient centrifugation. Spleen mononuclear cells were obtained by passage through an 80-mesh screen followed by Ficoll-Hypaque density gradient centrifugation. The cells were fixed in 1% paraformaldehyde and stained with fluorescein isothiocyanate-conjugated antibodies against murine CD4, CD8, CD22, γδ T-cell receptor, CD11a, CD18, CD44, CD49d, or CD54 or double stained with phycoerythrin-conjugated antibody against Mac-1 and fluorescein isothiocyanate-conjugated antibodies against CD11a, CD44, or CD54 (Caltag, South San Francisco, Calif.). Labeled isotype controls were included. Flow cytometry was performed on a Becton-Dickinson FACScan, and data was processed with Becton-Dickinson Lysis II and SAS JMP statistical software.

RESULTS

LCMV replication in cultured splenocytes is not inhibited by PTx.

Spleen mononuclear cells (splenocytes) are primarily lymphocytes with approximately 4 to 5% macrophages (adherent Mac-1⁺ cells). The lymphocytes are infectable by LCMV (Armstrong 53b strain) at only 1/100 the efficiency of macrophages (31a); hence, macrophages are the primary infected cell type in this preparation. Splenocytes from four BALB/c mice were divided into four groups: uninfected; infected; infected, with PTx pretreatment; and infected, with PTx posttreatment. The cells were cultured in RPMI 1640 containing 10% FCS, infected (or not) with LCMV Armstrong at a multiplicity of 1 PFU (as determined on Vero cells)/splenocyte, and treated (or not) with PTx (0.1 μg/ml) either 1 day before or 1 day after infection. The cells were cultured for 3 days, RNA was isolated, and RT-PCR was performed as described in Materials and Methods to detect LCMV Gp mRNA. Neither pretreatment nor posttreatment with PTx blocked the LCMV infection in cell culture (Fig. 1).

FIG. 1.

PTx does not affect LCMV mRNA production in splenocyte cell culture. LCMV Gp mRNA was detected by RT-PCR, starting with total RNA from splenocyte cultures. Agarose gel electrophoresis indicates that the expected 1-kb LCMV Gp product is not detected in uninfected splenocytes (lane 1) but is detected in samples from in vitro LCMV-infected splenocytes (lane 2), despite pretreatment (lane 3) or posttreatment (lane 4) with PTx (0.1 μg/ml). Lane N is a negative control for PCR. The 447-bp DHFR gene product was used as an internal control. Lane M is a 1-kb DNA ladder supplied by Gibco BRL (the upper gel depicts only the 1-kb marker fragment, whereas the lower gel depicts 0.5-kb and smaller marker fragments).

Portions of the splenocyte cultures were used for infectious-center assays, with the result that both PTx-treated and untreated cultures had 100 ± 10 infectious centers per 10⁶ cells.

LCMV dissemination after i.g. inoculation but not after i.v. inoculation can be delayed by PTx treatment.

Data on the tissue levels of virus were collected by RT-PCR, by plaque assay, and by in situ hybridization. For the RT-PCR experiments, there were four sets of mice, three time points per set, and six mice per time point. The sets were as follows: set 1, i.v.-infected mice; set 2, PTx-pretreated and i.v.-infected mice; set 3, i.g.-infected mice; and set 4, PTx-pretreated and i.g.-infected mice. The PTx-pretreated mice received PTx 3 days before LCMV inoculation and were sacrificed at 3, 7, or 10 days after infection. The mice inoculated without PTx treatment served as positive controls. Three days after inoculation, LCMV RNA was detected in the livers and the kidneys of all six mice from both the i.v.- and i.g.-infected groups (Fig. 2, lanes 1 and 2, and Table 1).

FIG. 2.

In vivo effects of PTx on virus. LCMV Gp mRNA was detected by RT-PCR of total RNA isolated from kidneys and livers of i.g.- or i.v.-infected mice. Gel electrophoresis of the RT-PCR product of total RNA isolated from the kidneys of i.g.- and i.v.-infected mice 3 days after infection is shown. The expected 1-kb LCMV Gp product is detected in samples from an i.g.-infected mouse (lane 1), an i.v.-infected mouse (lane 2), and a PTx-pretreated, i.v.-infected mouse (lane 4) but not in that from a PTx-pretreated, i.g.-infected mouse (lane 3). Lane 5 is from the kidney of an uninfected mouse. Lane P is a 1-kb LCMV Gp fragment. Lane N is a negative control for PCR. The 447-bp DHFR gene product was used as an internal control. Lane M is a 1-kb DNA ladder supplied by Gibco BRL (the upper gel depicts only the 1-kb marker fragment, whereas the lower gel depicts 0.5-kb and smaller marker fragments).

TABLE 1.

Frequency of detection of LCMV RNA in infected mice with or without PTx pretreatment^a

Treatment	No. of mice with LCMV RNA
	Day 3	Day 7
I.v.	6/6	6/6
I.g.	6/6	6/6
I.v. + PTx	6/6	6/6
I.g. + PTx	1/6	3/6

^aDay 0 is the day of infection. 

PTx inhibited the dissemination of LCMV after i.g. inoculation in five of six mice (Fig. 2, lane 3, and Table 1), but it did not inhibit dissemination after i.v. infection (Fig. 2, lane 4, and Table 1). By day 7, only three of six mice inoculated i.g. had detectable LCMV RNA in the liver or kidneys, while all six of the mice inoculated i.v. had LCMV RNA in those organs (Table 1).

We also determined the effect of PTx treatment on LCMV infection by performing plaque assays in parallel with RT-PCR. Spleens were removed 3 days after infection and analyzed by plaque assay by using Vero cell monolayers grown in 6-well plates. Plaques were counted after 5 days and expressed as the number of plaques per gram of tissue (Table 2). Five of five i.g.-inoculated, PTx-treated animals were RT-PCR negative for viral RNA and contained a mean of 5.4 × 10³ PFU/g of spleen, whereas untreated animals were uniformly positive for viral RNA and had greater than 10⁵ PFU/g (with a mean of 1.9 × 10⁷ PFU/g). The results of the RT-PCR supported those of the large experiment reported above, as well as the results of the plaque assays in the i.g.-infected, PTx-treated group (Table 2).

TABLE 2.

Detection of LCMV infection by plaque assay and RT-PCR in spleens from i.g.- or i.v.-infected mice with or without PTx treatment

Inoculation route	PTxa	Plaques/g	RT-PCRb
Intravenous	−	7.6 × 105	+
	−	1.9 × 105	+
	+	1.3 × 105	+
	+	4.4 × 106	+
Intragastric	−	4.3 × 105	+
	−	5.8 × 105	+
	−	7.0 × 107	+
	−	8.2 × 106	+
	−	1.7 × 107	+
	+	5.9 × 103	−
	+	NDc	−
	+	1.7 × 103	−
	+	1.3 × 104	−
	+	9.0 × 102	−

^a+, treated; −, not treated. 

^b+, RT-PCR positive for viral RNA; −, RT-PCR negative for viral RNA. 

^cND, not determined. 

LCMV was detected in various organs at different time points by in situ hybridization (Table 3). Three days after intravenous injection of 25 ng of PTx/g, each mouse was inoculated i.v. or i.g. with 10⁶ PFU of LCMV. At 24 h after gastric inoculation, viral RNA was detected in the stomach in all mice, whether they were pretreated with PTx or not. Stomach infection appeared within epithelial cells of the gastric mucosa (Fig. 3a). By 24 h virus was detectable in the spleen and liver of one of the mice in the i.g.-infected, PTx-untreated group but not in the other three i.g.-infected mice. By 48 h after gastric infection, viral RNA was detected in the spleen of one mouse and the livers of both mice from the PTx-untreated group, but not in the PTx-treated mice (Fig. 3b and c). In contrast to the i.g.-infected mice, virus appeared in i.v.-infected mice as early as 24 h in the stomach, liver, spleen, ileum, kidneys, and mesenteric lymph nodes, and PTx pretreatment did not affect the infection (Fig. 3 and Table 3).

TABLE 3.

Summary of in situ hybridization for detection of viral RNA with LCMV Gp cRNA probe in various mouse organs at different time points after inoculation^a

Organ	Hybridization
	i.g. infected				PTx treated, i.g. infected				i.v. infected				PTx treated, i.v. infected
	24 h p.i.b		48 h p.i.		24 h p.i.		48 h p.i.		24 h p.i.		48 h p.i.		24 h p.i.		48 h p.i.
	1c	2	1	2	1	2	1	2	1	2	1	2	1	2	1	2
Stomach	+++	+++	+++	+++	++	+++	+++	+++	+++	++	+++	++++	+++	++	+++	++++
Spleen	−	+	−	+++	−	−	−	−	+++	+	++	+++	+++	++	−	+++
Liver	−	+	+	++	−	−	−	−	++	+	+	+	++	+	+	+
Ileum	−	−	−	−	−	−	−	−	++	+	+	ND	+	+	ND	+
Mesenteric LNd	−	NDe	−	ND	ND	ND	−	ND	+	−	ND	+	ND	−	ND	++
Kidney	−	−	−	−	−	−	−	−	−	−	−	+	−	+	−	+

^aSections were observed at ×100 magnification, and three slices were examined for each tissue. −, no hybridization was observed in the section; +, less than 10% of the field was darkened with the hybridization signal; ++, 10 to 25% of the field was darkened with the hybridization signal; +++, 25 to 50% of the field was darkened with the hybridization signal; ++++, >50% of the field was darkened with the hybridization signal. 

^bp.i., postinfection. 

^cMouse number. 

^dLN, lymph node. 

^eND, not done. 

FIG. 3.

In situ hybridization of various mouse tissue sections with an LCMV Gp probe. (a) Stomach sections at 24 h showing infection of the gastric epithelium in the body of the stomach. Magnification, ×100. (b and c) Spleen and liver sections, respectively, at 48 h after infection. Magnification, ×100. IG, i.g. inoculation; IV, i.v. inoculation.

PTx causes leukocytosis.

One week after PTx treatment, the peripheral leukocyte counts of both i.v.- and i.g.-infected mice increased significantly in comparison with those of their PTx-untreated counterparts, with 7.65- and 10.18-fold increases, respectively. However, 10 days after PTx treatment, the peripheral leukocyte counts decreased, with only 1.99- and 2.46-times-higher concentrations of circulating leukocytes in the infected, PTx-treated animals, respectively (Fig. 4).

FIG. 4.

Absolute leukocyte counts in peripheral blood from LCMV-infected mice with or without PTx pretreatment. Leukocytes were counted under a 20×-objective light microscope in the presence of trypan blue to eliminate dead cell counts. Data are represented as leukocytes/milliliter of whole blood at days 7 and 10 after toxin administration. IV, i.v. inoculation; IG, i.g. inoculation.

LCMV infection affects adhesion molecule levels in vivo and in cell culture, and PTx treatment also modulates adhesion molecule expression.

Heparinized blood was obtained from 24 mice, half of which were PTx treated and the other half untreated. By 3 to 7 days after PTx treatment, lower levels of five adhesion molecules (CD11a, CD18, CD44, CD49d, and CD54) were detected on PBMC. However, the levels of these adhesion molecules returned to near normal 10 days after PTx treatment (Fig. 5).

FIG. 5.

Phenotypic analysis of mouse PBMC at different time points (days 0, 3, 7, and 10) after PTx treatment in vivo. All five of the adhesion molecules monitored show decreased expression on PBMC 3 to 7 days after PTx treatment and return to near normal by 10 days.

We also observed the effects of PTx on adhesion molecules in cell culture with or without LCMV infection. Spleen mononuclear cells were isolated from 24 BALB/c mice and divided into four groups, with the cells of six mice in each group. Splenocytes were cultured overnight, infected (or not) with LCMV at a multiplicity of 1 PFU/cell overnight, and then treated (or not) with PTx (0.1 μg/ml) for 3 days. The amount of PTx used in vitro corresponded to the estimated concentration of PTx in vivo and was below its mitogenic concentration (>1 μg/ml) (18). Nonadherent cultured splenocytes were collected, fixed, stained with fluor-conjugated antibodies, and analyzed by flow cytometry. CD11a (LFA-1 α chain) was up-regulated on murine splenocytes after LCMV infection (P < 0.05) (Table 4). PTx treatment significantly down-regulated the expression of CD11a, CD44, and CD54 (ICAM-1) on mouse splenocytes (P < 0.001). Changes were not as significant for CD18 (LFA-1 β chain), CD49d (VLA-4), and lymphocyte subsets CD4 (helper T cells), CD8 (cytotoxic T cells), CD22 (B cells), or γδ T cells (P > 0.1) (Table 4).

TABLE 4.

Changes in adhesion molecule expression on LCMV-infected and -uninfected murine splenocytes after PTx treatment^a

Adhesion molecule or lymphocyte subset	Expression (%)
	Uninfected		Infected
	Untreated	Treated	Untreated	Treated
Adhesion molecules
CD11a	68.8 ± 1.0	62.7 ± 4.2b	72.4 ± 0.9b	66.4 ± 2.7
CD18	45.7 ± 3.1	43.9 ± 3.0	47.9 ± 2.2	50.0 ± 1.5b
CD44	84.8 ± 3.7	73.6 ± 6.6b	81.9 ± 3.6	77.1 ± 3.2b
CD49d	4.4 ± 2.2	4.4 ± 0.7	3.8 ± 1.9	4.4 ± 2.4
CD54	78.9 ± 3.6	71.9 ± 2.0b	77.9 ± 3.2	73.0 ± 2.1b
Lymphocyte subsets
CD4	34.3 ± 3.6	35.2 ± 3.8	40.1 ± 5.5	41.5 ± 6.4b
CD8	29.4 ± 6.5	27.3 ± 4.2	21.5 ± 4.6	29.7 ± 5.1
CD22	7.8 ± 2.1	7.3 ± 0.9	7.8 ± 1.6	6.6 ± 0.8
γδTCRc	6.9 ± 1.5	7.1 ± 1.6	8.0 ± 2.1	7.9 ± 1.5

^aAdhesion molecule expression on splenocytes was examined by flow cytometry. The expression of each lymphocyte subset is also included to show that PTx has no significant effect on the subset distribution of lymphocytes. Each group contained six mice. The results are given as the mean ± standard deviation in percent positive. Differences in expression were analyzed by Student’s t test. 

^bThese values are significantly different from those of the uninfected, PTx-untreated group (P < 0.05). 

^cTCR, T-cell receptor. 

Further study by two-color flow cytometry of splenocyte cultures (which were a mixture of adherent and nonadherent cells) showed that LCMV infection up-regulated CD11a expression on Mac-1⁺ monocyte/macrophages (P < 0.01) (Table 5), and PTx treatment blocked this up-regulation (P < 0.05) (Table 5). Two-color flow cytometry to detect expression of LCMV Gp and CD11a indicated that the up-regulation of CD11a was occurring mostly in infected cells. Approximately 5% of the splenocytes were infected, and 90% of the infected cells were doubly fluorescent for CD11a and Gp (data not shown). PTx also down-regulated the expression of CD44 and CD54 on Mac-1⁺ monocyte/macrophages (P < 0.001) (Table 5).

TABLE 5.

Changes in adhesion molecule expression on cultured Mac-1⁺ adherent splenocytes after LCMV infection and PTx treatment^a

Adhesion molecule	Expression (%)
	Uninfected		Infected
	Untreated	Treated	Untreated	Treated
CD11a	13.0 ± 1.8	9.7 ± 1.0b	17.4 ± 3.2b	9.9 ± 3.1b
CD44	15.2 ± 2.8	9.8 ± 1.7b	15.6 ± 1.4	9.7 ± 1.8b
CD54	19.8 ± 4.5	12.1 ± 2.8b	19.6 ± 2.8	11.6 ± 2.7b

^aAdhesion molecule expression on Mac-1⁺ adherent splenocytes was examined by two-color flow cytometry. Each group contained six mice. The results are given as the mean ± standard deviation in percent positive. Differences in expression were analyzed by Student’s t test. 

^bThese values are significantly different from those of the uninfected, PTx-untreated group (P < 0.05). 

DISCUSSION

Previous studies have shown that the mucosa is the initial site of LCMV infection after i.g. inoculation. Subsequently, the virus spreads to the spleen and liver, then the ileum, and finally, the lungs, kidneys, brain, and esophagus (29). Our results corroborate this order of events and show that treatment with the G protein inhibitor PTx blocks the initial step and delays dissemination from the gastric mucosa to the spleen. Twenty-four hours after gastric inoculation, viral RNA was first detected in the stomachs of all the mice studied. By 48 h, viral RNA was detected in the spleen and liver, as well as the stomach, in PTx-untreated mice but not in the i.g.-inoculated, PTx-treated mice. In contrast, virus appeared at all sites (stomach, liver, kidneys, and brain) as early as 24 h after i.v. inoculation, irrespective of PTx treatment.

Chemokines, such as macrophage-derived chemokine (MDC) and MCP-3, promote the chemotaxis of lymphocytes, monocytes/macrophages, polymorphonuclear leukocytes, NK cells, dendritic cells, and endothelial cells both in vivo and in vitro (9, 13, 31) through a PTx-sensitive G protein signaling pathway (7). In murine cytomegalovirus infections, the loss of a virus-encoded G protein-coupled chemokine receptor abrogates replication in the salivary gland (6), suggesting that this receptor is important for trafficking of infected leukocytes to the target organs. In our studies, PTx was employed at a concentration sufficient to inhibit G protein-coupled signaling but at 2 orders of magnitude below the concentration needed to cause mitogenesis or toxicity in culture (26). Arenaviruses primarily infect reticuloendothelial cells (monocyte/macrophages and dendritic cells) (41), and whereas PTx does not affect their ability to replicate virus (Fig. 1), it directly affects the expression of integrins on their surfaces (Table 5). We showed that PTx does not affect production of interferon-γ, a major inhibitor of viral infection (data not shown), and that it does not cause any significant changes in splenic lymphocyte populations (Table 4). We have also shown that PTx acts before LCMV immune responses come into play (28). Thus, the most likely mechanism for the PTx inhibition of virus dissemination is the inhibition of G protein-coupled chemotactic signals in infected reticuloendothelial cells.

To measure the effectiveness of PTx treatment, we monitored leukocytosis. By 1 week after PTx injection, the peripheral leukocyte count increased significantly in PTx-treated mice and, at the same time, LCMV RNA was detectable in the liver and kidneys in only one of six i.g.-inoculated mice. However, 10 days after PTx injection, the peripheral leukocyte count returned to near normal, and by this time, LCMV RNA was detectable in three of six mice. Thus, we demonstrated a correlation between PTx disruption of normal leukocyte trafficking and PTx disruption of LCMV dissemination after mucosal inoculation.

Inhibition of G protein-coupled signaling blocks the ability of chemokines to activate integrins (4) and leads to a decrease in integrin expression. Accordingly, PTx treatment of mice decreases the levels of five adhesion molecules within 3 to 7 days. The levels return to normal by 10 days after treatment. The decrease in adhesion molecules parallels the peak of leukocytosis, and circulating leukocyte counts return to normal levels as the levels of adhesion molecules return. Leukocytosis coincides with an observable depletion of cells in the spleen and lymph nodes (as shown by histopathology done by C. Yin). Adhesion molecules on leukocytes mediate early and reversible interactions (rolling and margination) along the lumenal surfaces of vascular endothelial cells. Certain members of the immunoglobulin superfamily (VCAM-1 and ICAM-1) regulate later and irreversible steps which lead to firm attachment and subsequent diapedesis (10). Our current understanding of the role of adhesion molecules in leukocyte trafficking allows us to surmise that the down-regulation of adhesion molecules is connected to the depletion of lymphoid tissues.

In order to investigate further the PTx effects on cell surface adhesion molecules, we monitored the effects of PTx in cultured spleen mononuclear cells with or without LCMV infection. Similar to what we found in vivo, PTx significantly down-regulates the expression of CD11a, CD44, and CD54 and has minor effects on CD18 and CD49d. We discovered that CD11a is up-regulated on murine splenocytes after LCMV infection, especially on murine splenic monocytes/macrophages, which are the major cells infected by LCMV. Cultured splenocytes show a small (4%) but significant increase in CD11a-positive cells. This can be reconciled with the large effect on trafficking in vivo by the fact that only 5% of the splenocytes are infected and that CD11a up-regulation occurs preferentially in infected cells.

It is important to note that others have shown that LCMV infection elicits CD11a expression on CD8⁺ lymphocytes (1, 2, 19). Since CD8⁺ lymphocytes are poorly infectable, they account for most of the uninfected CD11a⁺ cells. The marked and long-standing LFA-1 up-regulation on CD8⁺ lymphocytes may explain the rapid infiltration of lymphocytes into infected sites and the accompanying immunopathogenesis that is a hallmark of acute LCMV disease. Our new observation is that LFA-1 α chain (CD11a) expression increases on the surface of monocytes/macrophages, the primary cells infected with LCMV, far more rapidly and extensively than on T cells. We surmise that this increase has some effect on the trafficking of infected cells and ultimately on the dissemination of infection within the host.

Virus-induced changes in adhesion molecule expression have also been observed by many other laboratories. For example, intercellular adhesion molecule-1 (ICAM-1) was induced on human hepatocytes after transfection by hepatitis B virus DNA (40). ICAM-1 expression was also up-regulated on BALB/c mouse brain microvascular endothelial cells by measles virus and herpes simplex type 1 virus and was responsible for increased lymphocyte homing to the central nervous system (3). The expression of LFA-1 and ICAM-1 is also stimulated by human immunodeficiency virus type 1 (HIV-1) infection and suggests a mechanism for extravascular dissemination of HIV-1-infected cells (30). Taken together, these findings suggest that up-regulation of LFA-1 expression may be a critical factor in virus tropism, directing infected cells towards vascular adhesion, and that PTx counteracts this direction by down-regulating LFA-1 and causing more leukocytes to remain in the circulation.

In summary, we used PTx to interfere with the normal leukocyte trafficking that is required for LCMV dissemination. These results encourage further efforts to modify cell trafficking as an approach to manipulating mucosal viral infections, such as LCMV infection. Other viral infections that occur primarily via the mucosal route, e.g., simian immunodeficiency virus infection in rhesus macaques (27), HIV infection in humans (11, 21, 34), reovirus infection (39), and infections by picornaviruses (25), may all be affected by chemotaxis-inhibiting agents.