Fig 1.
IFNAR blockade increases MCMV dissemination from a peripheral site.
(a). BALB/c mice were given IFNAR blocking (αIFNAR) or pDC depleting (αpDC) antibodies in PBS, or PBS only (control), then given MCMV-LUC i.f. (106 p.f.u.). We tracked infection by luciferin injection and live imaging of light emission (radiance = photons/sec/cm2 /steradian). Bars show means, other symbols show individuals. Both αIFNAR and αpDC significantly increased luciferase signals in the feet (footpad + PLN) and in the neck (salivary gland) from day 3, with αIFNAR having a significantly greater effect. After day 4, αpDC only affected neck signals. (Student’s two-tailed unpaired t-test; *p<0.05, **p< 0.01, ***p<0.001, ****p<0.0001). The dotted lines show assay sensitivity limits. (b). Mice were treated and infected as in (a), and organs harvested 3 or 6 days later for ex vivo luciferase imaging. Liver and salivary gland signals were not detected at day 3. The Y axis baselines correspond to assay sensitivity limits. Significant signals above the controls are indicated according to the scheme in (a). (c). The organs from (b) were plaque assayed for infectious virus. Bars show means, other symbols show individual organs. Dotted lines show assay sensitivity limits where above the Y axis baseline. Titers significantly above those of controls are indicated. Significant signals above the controls are indicated according to the scheme in (a).
Fig 2.
IFNAR blockade increases early LN infections of C57BL/6 and BALB/c mice.
C57BL/6 (a, c) and BALB/c (b, d) mice were given IFNAR blocking antibody (αIFNAR) or not (control) i.p., then MCMV i.f. (106 PFU). Infectious virus was plaque assayed 3 days (a, b) or 1 day later (c, d). Bars show means, other symbols show individual mice. Dotted lines show assay sensitivity limits. Significant differences were calculated and are indicated according to the scheme of Fig 1.
Fig 3.
IFNAR blockade increases SSM and FRC infections.
C57BL/6 mice were given IFNAR blocking antibody (αIFNAR) or not (control), then MCMV-GR i.f. (106 p.f.u.). PLN taken 1 day later were stained for viral GFP (GFP) and lytic antigens (MCMV), plus the SSM marker CD169 (a), the SSM/DC marker CD11c (b) or the FRC marker ER-TR7 (c). Nuclei were stained with Hoechst 33342 (blue). Arrows show example infected SSM and FRC. Both were more numerous after αIFNAR. In (d) we counted cells across 4–5 fields of view for PLN sections from each of 3 mice per group. Circles show individual means, bars show group means. Significant differences are indicated (Student’s two tailed unpaired t-test; ***, P<0.001; ****, P<0.0001). We quantified infection also for CD68+ (pan-macrophage/DC), CD206+ (mannose receptor, absent from SSM), SiglecH+ (pDC) and NKp46+ (NK) cells.
Fig 4.
IFNAR blockade increases MCMV lytic infection in LN.
C57BL/6 mice were given IFNAR blocking antibody (αIFNAR) or not (control), then MCMV-GR i.f. (106 p.f.u.). PLN taken 1 day later were stained for viral GFP, MCMV IE1 antigen, and LYVE-1 (lymphatic endothelium). (a) shows examples of staining. (b) shows quantitation, counting IE+ cells for 7 fields of view in PLN sections from each of 3 mice per group. Circles show individual means, bars show group means. αIFNAR significantly increased IE+ cell numbers (Student’s two tailed unpaired t-test with Welch’s Correction; ***, P<0.001).
Fig 5.
IFN-I-independent NK cell recruitment aids infection control.
(a). C57BL/6 mice were given IFNAR blocking antibody (αIFNAR) or not (control) then i.f. MCMV-GR (106 p.f.u.). 3 days later PLN were stained for viral GFP, lytic antigens (MCMV), and B cells (B220) or FRC (ER-TR7). Nuclei were stained with DAPI. αIFNAR reduced B220 and DAPI staining, and disorganized ER-TR7 staining, implying loss of LN cellularity and structure. (b). C57BL/6 mice were given αIFNAR, depleted of NK1.1+ cells with mAb PK136 (αNK), given both treatments, or given neither (control). All were then given i.f. MCMV-GR (106 p.f.u.). 1 day later PLN were stained for viral GFP and NKp46+ NK cells. Nuclei were stained with DAPI. Closed arrows show example NKp46+ cells. Open arrows show example GFP+ infected cells. (c). Quantitation of staining as illustrated in (a) and (b), across 7 fields of view for sections from each of 5 mice per group, showing significant loss of LN cellularity in IFNAR treated-mice (B220+, DAPI+) and significant NK cell recruitment (NKp46+) after αIFNAR and effective NK cell depletion by mAb PK136. Bars show group means, other symbols show mean counts of individual mice. (****, P<0.0001; Student’s two-tailed unpaired t-test). (d). Organs of mice treated as in (b) were plaque assayed for infectious virus. Bars show means, other symbols show individual mice. Significant differences are indicated (Student’s two tailed unpaired t-test; *p<0.05, **p<0.01). Dotted lines show assay sensitivity limits.
Fig 6.
NK cells eliminate ER-TR7+ infected cells and cause tissue damage.
(a). C57BL/6 mice were given IFNAR blocking antibody (αIFN), depleted of NK1.1+ cells (αNK), given both treatments, or given neither (control). All were then given i.f. MCMV-GR (106 p.f.u.). 1 day later PLN sections were stained for viral GFP and the FRC marker ER-TR7. Nuclei were stained with DAPI. Arrows show example infected ER-TR7+ cells, which were abundant after αNK. (b). Low magnification PLN images of mice treated as in (a), but at 3 days post-infection, showing reduced DAPI staining and altered ER-TR7 staining (arrows) after αIFN, and restoration of these changes by αNK. (c). Counts of infected cells as shown in a, summing across 6 fields of view from sections of 3 mice per group. Statistical comparison by Fisher's Exact Test is between each treatment group and the control. (d). counts of DAPI+ nuclei as shown in b, for 5 fields of view from sections of 3 mice per group. Bars show means, other symbols show individual fields of view. IFNAR blockade significantly reduced the density of DAPI+ nuclei compared to control PLN but not when NK cells were also depleted. NK depletion alone had a small effect (*p<0.05, ***p<0.001; Student’s two tailed unpaired t test). (e). BALB/c and C57BL/6 mice were given αIFN, depleted of asialoGM1+ cells (αNK), or left untreated (cont). All were then given i.f. MCMV-GR (106 p.f.u.). Organs were plaque assayed for infectious virus 1 day later. Bars show means, other symbols show individuals. Dashed lines or axis baselines show assay sensitivity limits. Significant differences relative to the controls for each mouse strain are indicated (*p<0.05, ***p<0.001; Student’s two tailed unpaired t-test).
Fig 7.
Caspase 3 expression in MCMV-infected PLN.
(a). C57BL/6 mice were depleted of NK1.1+ cells (αNK) or not (control), then given i.f. MCMV-GR (106 p.f.u.). 1 day later PLN sections were stained for viral GFP, the FRC marker ER-TR7 and the apoptotic cell marker caspase 3. Nuclei were stained with DAPI. Arrows show example FRC, caspase 3+ in the control and caspase 3- in the αNK. (b). Staining of an uninfected PLN showed negligible caspase 3 expression. (c). Counting samples of mice treated as in a, across 12 sections from 4 PLN, showed by Fisher's Exact Test a significant reduction in FRC caspase 3 expression after αNK.
Fig 8.
Comparison of SSM depletion and IFNAR blockade effects on PLN infection.
(a). C57BL/6 mice were depleted of SSM with liposomal clodronate (clod), given IFNAR blocking antibody (αIFN), given both treatments, or neither (control), then infected i.f. with MCMV-GR (106 p.f.u.). 3 days later, organs were plaque-assayed for infectious virus. Horizontal bars show means, other symbols show individuals. Dashed lines show assay sensitivity limits where above the y axis baseline. Significant increases above the control group are indicated (**p<0.01, ***p<0.001, ****p< 0.0001; Student’s two tailed unpaired t-test). In PLN, αIFN and clod treatments did not give significantly different titers (p>0.5), and each alone increased titers significantly less than both together (p<0.05). (b). Mice were treated and infected as in (a). Day 3 PLN sections were stained for viral GFP and markers of SSM (CD169), FRC (ER-TR7) or NK cells (NKp46). Nuclei were stained with DAPI. All groups showed loss of CD169 expression. Clod treatment increased FRC infection but caused little of the NK cell infiltration, loss of DAPI staining and ER-TR7 disruption seen with IFNAR blockade. (c). Counts of infected cells as shown in (b), summing across 5 fields of view from sections of 4 mice per group. Statistical comparison is between each treatment group and the control using Fisher's Exact Test. (d). Counts of NKp46+ and DAPI+ cells as shown in (b), comparing mean counts per field of view for 6 fields of view from sections of 3 mice per group. Statistical comparisons were with the control group. ns = not significant (p>0.05).
Fig 9.
Non-myeloid cells produce most of the infectious virus in IFNAR-blocked mice.
(a,d). LysM-cre (a-c) or CD11c-cre mice (d-f) were given IFNAR blocking antibody (αIFNAR) or not (control) then MCMV-GR i.f. (106 p.f.u.). Footpads, PLN and spleens harvested 3 days later were plaque assayed for infectious virus. Squares and circles show individuals, horizontal bars show means. Significant differences are indicated (Student’s two tailed unpaired t-test; *p<0.05, **p<0.01). (b,e). Switched (red) and unswitched (green) infected cells were counted on PLN and spleen sections for 5 fields of view per mouse from each of 4 mice per group. The numbers give the % cell switching and in parentheses the switched and total infected cell counts. The switching rates αIFNAR and control mice were compared for each site by Fisher's Exact Test. We also determined below the % virus switching (infectious) recovered from each mouse. The numbers give mean ± SEM for each group. The αIFNAR and control results were compared (Student’s two tailed unpaired t test). (c,f). Example viral GFP and tdTomato staining is shown for infected PLN of control and αIFNAR mice. Nuclear DAPI staining shows the loss of cellularity in PLN of αIFNAR mice.
Fig 10.
Limited MCMV replication in SSM suggests that they function mainly to limit FRC infection.
(a). LysM-cre mice were given IFNAR blocking antibody (αIFNAR), without or without additional NK cell depleting antibody (αNK). They were then given MCMV-GR i.f. (106 p.f.u.). Footpads, PLN and spleens harvested 3 days later were plaque assayed for infectious virus. Squares and circles show individuals, horizontal bars show means. Significant differences are indicated (Student’s two tailed unpaired t-test; *p<0.05, **p<0.01, ****p<0.0001). (b). Viruses recovered in (a) were typed as green or red fluorescent. Squares and circles show the cre-driven green to red switching rates of individual mice, horizontal bars show means. Significant differences are indicated (Student’s two tailed unpaired t-test with Welch’s correction; **, P<0.01) (c). LysM+/cre (cont) and LysM+/creIFNARflox/flox (IFNARfl) mice were depleted of NK cells then infected i.f. with MCMV-GR (106 p.f.u.). Virus titers and comparisons were determined as in (a). (d). The viruses recovered in (c) were assayed for green to red switching. Significant differences were calculated according to (b) and are indicated (*, p<0.05; **, P<0.01; ****, P<0.0001). (e). The floxed colour switch virus (MCMV-GR) and its switched derivative (MCMV-R) were mixed 1:1 and given i.f. to lysM-cre mice (106 p.f.u.). As a control, mice were given MCMV-GR only. Viruses recovered after 3 days were assayed for green / red switching. The switching of MCMV-GR was <5%. The mixed viruses remained 1:1 (50% switched). Therefore the low switching of MCMV-GR was not due to poor fitness of the switched virus. (f). MCMV-GR was given i.p. to lysM-cre mice (106 p.f.u.). Two days later peritoneal macrophages were recovered by lavage, identified by staining for F4/80, and examined for viral fluorochrome expression. The closed arrow shows an example switched macrophage, with red nuclear fluorescence. The open arrow shows an example unswitched infected cell (F4/80-) with cytoplasmic GFP fluorescence. Some F4/80+ cells showed residual green fluorescence in addition to red fluorescence, but >90% of fluorescent F4/80+ cells were switched.
Fig 11.
IFN-I protects FRC against direct MCMV infection.
(a). C57BL/6 mice were given IFNAR blocking antibody (αIFNAR) or not (control) then infected i.f. with MCMV in which a β-galactosidase (βgal) expression cassette disrupts the essential coding sequence of gL (106 p.f.u.). 1 day later PLN sections were stained for βgal to identify infection and for B cells (B220, upper panels) or for FRC (ER-TR7) and with DAPI (lower panels). Arrows in the lower panels show example infected FRC. (b). βgal+ cells were counted for 6 fields of view from sections of 3 mice per group. Circles show field counts, bars show means. IFNAR blockade significantly increased infection (***p<0.001; Student’s two tailed unpaired t-test with Welch’s Correction). (c). The proportion that were ER-TR7+ was determined for at least 60 βgal+ cells per group (counting more sections from the controls to reach this total). Fisher's Exact Test showed a significantly higher proportion of infected cells was ER-TR7+ after IFNAR blockade.
Fig 12.
Schematic representation of MCMV escape from the LN subcapsular sinus.
Most virions in the afferent lymph are captured by SSM, but some infect FRC. In the absence of IFN-I (a), FRC infection is highly productive, SSM infection rather less so. Virions are released into the LN to infect new FRC and myeloid cells, and also back into the lymph flow, whence they can reach the blood. To counter this (b), IFN-I from SSM capturing virions inhibits infection in both SSM and FRC. pDC also produce IFN-I. They may receive infected cell debris and cytokines via the FRC conduits. Inflammatory mediators also recruit NK cells to kill infected FRC. Thus MCMV propagation is largely suppressed.