Identification of a Locus in Mice that Regulates the Collateral Damage and Lethality of Virus Infection

SUMMARY

Arenaviruses can cause severe hemorrhagic disease in humans, which can progress to organ failure and death. The underlying mechanisms causing lethality and person-to-person variation in outcome remain incompletely explained. Herein, we characterize a mouse model that recapitulates many features of pathogenesis observed in humans with arenavirus-induced hemorrhagic disease, including thrombocytopenia, severe vascular leakage, lung edema, and lethality. The susceptibility of congenic B6.PL mice to lymphocytic choriomeningitis virus (LCMV) infection is associated with increased antiviral T cell responses in B6.PL mice compared with C57BL/6 mice and is T cell dependent. Pathogenesis imparted by the causative locus is inherited in a semi-dominant manner in F1 crosses. The locus includes PL-derived sequence variants in both poorly annotated genes and genes known to contribute to immune responses. This model can be used to further interrogate how limited genetic differences in the host can remarkably alter the disease course of viral infection.

Graphical Abstract

In Brief

Arenaviruses can cause devastating illness, including severe systemic hemorrhaging and death in humans. Misumi et al. characterize a mouse model that recapitulates many features of pathogenesis observed in humans and identify a genetic locus that regulates T cell responses, platelet frequencies, and pathogenesis during arenavirus infection.

INTRODUCTION

New World and Old World arenaviruses can cause severe hemorrhagic disease in humans. Lassa fever virus is endemic in western Africa, infecting roughly 300,000 people annually and killing 5,000–10,000 (Smith et al., 2014; Sogoba et al., 2012). New World arenaviruses, such as Sabia, Junin, and Machupo, cause smaller outbreaks with higher frequencies of hemorrhagic disease and lethality, approaching a case fatality rate of 10%–50% (Sarute and Ross, 2017; Smith et al., 2014). Severe arenavirus infections are characterized by thrombocytopenia, disseminated intravascular coagulation, and, in the case of New World viruses, petechia and tissue hemorrhage.

The outcome of infection for any particular individual is difficult to predict and is likely influenced by infection dose, pre-existing immunologic memory, and underlying genetic variation among patients. Although sequences in virus variants can be linked to pathogenesis, it has been far more difficult to identify polymorphic loci that infiuence disease severity in the human population. Such studies are simplified in mouse models in which genetics, viral strain, and viral load can be manipulated. Lymphocytic choriomeningitis virus (LCMV) is a rodent-borne arenavirus that causes a non-cytopathic infection in mice. LCMV-Clone13 rapidly disseminates in mice, overwhelming early immune responses to establish chronic infection, typically associated with T cell exhaustion. Adult C57BL/6J (B6) or C57BL/6NJ mice that are given LCMV-Clone13 show limited pathogenesis during early stages of infection, including modest weight loss and reduction in temperature, and the mice recover 1–2 weeks after infection. In contrast, PL/J (PL) mice perish following LCMV-Clone13 (Oldstone et al., 1993), but the genetic basis for these divergent outcomes is unknown.

Herein, we describe a mouse model that reproduces several cardinal features of viral hemorrhagic disease following LCMV-Clone13 infection and use forward genetics to reveal a susceptibility locus. B6.PL-Thy1^a/CyJ (B6.PL) is a mouse congenic strain with >98% of its genome derived from the C57BL/6 inbred strain, including at the major histocompatibility complex (MHC) locus (Morgan et al., 2015). Upon infection with LCMV-Clone13, B6.PL mice generate robust antiviral T cell responses that are greater than those generated in B6 mice, resulting in lower virus burdens. However, infected B6.PL mice show a precipitous loss of platelets, compromised vascular integrity, and pulmonary edema, and they abruptly perish early during infection. Using high-resolution genotyping, forward genetics, and sequencing, we identify a locus associated with the lethal phenotype. PL sequences at this locus contain deleterious alterations in genes with known immune-related functions, which may explain the increased T cell response during infection. Finally, we describe a line of congenic mice that are >99.5% identical to B6 mice but with a narrowed susceptibility locus that can be used to interrogate the underlying mechanisms of viral pathogenesis during arenavirus infection.

RESULTS

LCMV-Clone13 Induces Severe Weight Loss, Vascular Leakage, and Pulmonary Edema in B6.PL Mice

PL/J mice develop lethal disease following LCMV-Clone13 infection (Figure S1A; Oldstone et al., 1993). To better understand the genetic basis for the lethal outcome and whether it was dependent on H-2^u MHC haplotype, we examined immune responses to LCMV-Clone13 in B6.PL mice. B6.PL mice were originally generated in the 1970s by crossing PL/J mice to the B6 background and are frequently used in immunology because they express the Thy1.1 congenic allele and are H-2^b.

B6.PL mice infected with LCMV-Clone13 (2 3 10⁶ PFU) developed an acute pathogenesis that resulted in 100% lethality, preceded by significant weight loss and hypothermia (Figures 1A–1C). Physical signs of illness (e.g., rapid breathing, shivering, bristled fur, and hunched back) suddenly deteriorated at day 6 post-infection, followed by death or mandatory euthanasia by day 8. Lethality occurred with similar kinetics in male and female B6.PL mice (Figures S1B and S1C). In contrast to B6.PL mice, B6 mice showed less weight loss and limited hypothermia at day 6 and began recovering at day 7 with normal appearance and activity by days 9–10 (Figures 1A–1C; data not shown). At a medium dose of infection, B6.PL mice showed significantly greater weight loss than B6 mice (Figure S2), though there was no difference between B6 and B6.PL in survival or temperature at low or medium doses of infection.

Figure 1. LCMV-Induced Lethality in B6.PL Mice.

B6, B6.PL, and F1 (B6 × B6.PL) mice were given LCMV-Clone13 (2 × 10⁶ plaque-forming units [PFUs], intravenous [i.v.]) and analyzed for physical signs of pathogenesis.

(A) Percentage survival following infection. ***p < 0.0001, Mantel-Cox test; n = 20–37 mice per group.

(B) Body weight loss (percentage initial body weight). Significant difference between B6, B6.PL, and F1 mice at day 7 is indicated. ***p < 0.0001, Kruskal-Wallis test; n = 20–37 mice per group.

(C) Body temperature decline. Significant difference between the three groups at day 7 is indicated. ***p < 0.0001, Kruskal-Wallis test; n = 20–37 mice per group.

(D) H&E staining of lungs at days 0, 6, 8, and 9 post-infection (40× magnification). Yellow scale bar, 100 μm. Representative images from one of two independent experiments; 2–3 mice per group per experiment.

(E) Evans blue dye extravasation from the lungs of uninfected mice or at days 7 or 9 post-infection. *p < 0.05, Mann-Whitney test. Data from two independent experiments with 3–9 B6 and B6.PL mice or two experiments with 2–6 B6 and F1 mice.

(F) H&E staining of spleens at day 7 (43 magnification). White scale bar, 1,000 mm.

(G) IFN-β and IFN-α levels in sera of LCMV- infected B6, F1, and B6.PL mice at 19 h post-infection. IFN-β: 7–10 mice per group from three experiments, **p < 0.01 (Kruskal-Wallis with Dunn’s multiple-comparison test); IFN-α: n = 4 from one experiment, *p < 0.05 (Mann-Whitney test).

(H) CBC quantitation of blood platelets at day 7. ****p < 0.0001, Student’s t test.

(I) Percentage of CD42d+ platelets among blood cells as measured by flow cytometry. Data from at least two experiments at each time point with 3–7 mice per group. **p < 0.01 at day 7, Mann-Whitney test.

See also Figures S1–S3 and Table S1.

To determine whether susceptibility to high-dose LCMV was inherited as a recessive or dominant trait, we crossed B6.PL mice to B6 mice to generate F1 mice. Infected F1 mice showed an intermediate phenotype, with a delay in weight loss and an intermediate reduction in temperature, and 57% of F1 mice perished by day 10 (Figures 1A–1C). F1 mice surviving past day 10 subsequently recovered (not shown). Thus, there is an acute stage of pathogenesis with an underlying genetic mechanism that is independent of MHC haplotype, and LCMV susceptibility is inherited as a semi-dominant trait.

LCMV was present in the lung, liver, and kidneys of B6, B6.PL, and F1 mice (Figure S1D). Compared with B6 mice, B6.PL mice showed 10-fold lower titer in the lung at day 6 and significantly lower levels in the lung (42-fold lower), liver (106-fold lower), and kidney (68-fold lower) at day 7, before perishing. Virus levels in F1 mice were modestly reduced in the lungs at day 8 (Figure S1D). These data indicate that the B6.PL allele does not prevent LCMV replication or spread, but B6.PL mice show a trend toward enhanced immune control.

Lung injury is a major contributor to morbidity following viral infection in mouse models and humans with hemorrhagic virus infection. Because B6.PL mice showed labored breathing and lower viral titer in the lung at day 6, we considered that the B6.PL allele negatively affected the respiratory system after infection. Before infection, the lung architecture of B6.PL mice appeared healthy and without infiltrating cells or fluid (Figure 1D). Between days 7 and 8 after infection, severe edema with infiltrating mononuclear cells appeared in the lungs of B6.PL mice but not B6 mice (Figures 1D and S1E). A similar but delayed pattern held for F1 mice at day 9 (Figure 1D). To assess vascular leakage into the lung, Evans blue dye was injected intravenously 30 min before tissue isolation. Evans blue was largely excluded from the lungs of uninfected mice (Figure 1E). However, the lungs of infected B6.PL mice at day 7 and F1 mice at day 9 showed a significant uptake of Evans blue compared with B6 mice (Figure 1E). There was also a trend toward increased protein in the BAL exudate fluid of infected F1 mice compared with B6 mice (Figure S1F), suggesting the B6.PL allele increases vascular leakage into the lung and impairs respiratory function.

Other differences in pathogenesis were noted, including a trend toward higher serum alanine aminotransferase (ALT) levels in B6.PL and F1 mice compared with B6 mice at days 6–7 (Figure S1G), suggesting increased hepatic injury during infection. Infected B6.PL and F1 mice had spleens with poorly defined structure (Figures 1F and S1H), whereas B6 mice showed well-separated white pulp and red pulp and enlarged follicles, suggesting that the B6.PL allele promotes splenic necrosis. Interestingly, B6.PL mice had significantly lower levels of IFN-β and IFN-α at 19 h of infection (Figure 1H), implying that pathogenesis occurs even when early type 1 interferon levels are reduced compared with B6 mice.

The loss of vascular integrity, prominent lung edema, and splenic necrosis prompted us to quantify platelet levels in B6.PL mice. Platelets maintain hemostasis during viral infections (Assinger, 2014), and thrombocytopenia is a typical feature of severe arenavirus infections in humans, non-human primates, and mice (Baccala et al., 2014; Schnell et al., 2012; Zapata et al., 2014). Platelets maintain the endothelial barrier in the lung and other tissues and protect splenocytes from necrosis (Iannacone et al., 2008; Loria et al., 2013). We observed fewer platelets and red blood cells in the blood of infected B6.PL mice (Figure S3A). The ratio of blood cell to total blood volume was lower for infected B6.PL mice than B6 mice (Figure S3B), consistent with a reduction in the absolute number of platelets and red blood cells in the blood at day 7. CBC analyses and flow cytometry of blood showed significantly smaller numbers of platelets in the blood at day 7 (Figures 1H and S3C; Table S1), reaching levels associated with hemorrhagic disease (13−93 × 10³/μL), though not sufficiently low (<20 × 10³/μL) to cause petechiae (Iannacone et al., 2008). Infection resulted in a significant reduction in the percentage of CD42d+ or CD42a+ (GPIX+) platelets in blood for both groups of mice at day 4; however, B6.PL mice continued to lose platelets, whereas platelet frequencies rebounded in B6 mice (Figures 1H and S3C and data not shown). Platelets increased in size after infection, but there was no difference in size between the B6.PL and B6 platelets. Overall, infected B6.PL mice showed an exaggerated reduction in platelet frequency, possibly explaining their increases in vascular leakage, lung edema, splenic necrosis, and death.

Hematopoietic Cells with B6.PL Allele Cause Lethal Outcome

We assessed whether pathogenesis was linked to B6.PL sequence in bone marrow-derived cells or non-hematopoietic cells (e.g., mesenchymal). Four cohorts of bone marrow chimera mice were established. In one set, B6 mice were irradiated and reconstituted with bone marrow from B6.PL or B6 mice (Figure 2A). After verifying successful reconstitution, the mice were infected and analyzed for survival, body temperature, and body weight. B6 mice containing B6-derived bone marrow showed 100% survival to day 10 (Figure 2A, black line). The mice showed transient hypothermia that reversed after day 6 and weight loss that appeared to abate at day 10. In contrast, B6 mice with B6.PL bone marrow showed 91% lethality (Figure 2A, red line). These mice showed significantly worsened hypothermia compared with B6 mice with B6 bone marrow, and there was continued weight loss among the mice that survived to day 10. These data indicate that the B6.PL allele in hematopoietic cells is sufficient to cause severe pathogenesis.

Figure 2. B6.PL Allele in Hematopoietic Cells Is Sufficient to Cause Pathogenesis after Infection.

B6 or B6.PL recipients were irradiated and reconstituted with the indicated bone marrow cells. At day 60 post-bone marrow transfer, blood from recipients showed chimerism, with 80%–90% of T cells derived from donor bone marrow and 10%–20% from host-derived bone marrow. Mice were infected at day 90 post-transfer with LCMV-Clone13 and analyzed for physical signs of pathogenesis.

(A) B6 recipients reconstituted with B6 or B6.PL bone marrow. Data combined from two independent experiments with 11 recipients per group.

(B) B6.PL recipients reconstituted with B6 or B6.PL bone marrow. One experiment with 7–10 recipients per group.

Survival: Mantel-Cox test; temperature and weight loss at day 7: Student’s t test; *p < 0.05, **p < 0.01, and ***p < 0.001.

Because the B6.PL allele in non-hematopoietic cells (e.g., mesenchymal or stromal cells) could affect cell tropism, interferon levels, or other aspects of physiology, we established a separate set of bone marrow chimera mice in which B6.PL mice were irradiated and repopulated with bone marrow from B6 or B6.PL mice (Figure 2B). B6.PL mice with B6 bone marrow (black line) showed improved survival (75%) compared with B6.PL mice containing B6.PL bone marrow (red line), which showed 0% survival to day 10 (Figure 2B). The B6.PL mice with B6 bone marrow began to recover temperature after day 8, whereas B6.PL mice with B6.PL bone marrow showed continued loss of temperature. Although both groups showed weight loss across time, the B6.PL mice with B6 bone marrow maintained more weight than the B6.PL mice with B6.PL bone marrow, though the effect was modest. Altogether, these data indicate that the B6.PL allele in hematopoietic cells confers a risk for severe pathogenesis after infection.

B6.PL Mice Generate Stronger and Quicker T Cell Responses Compared with B6 Mice

Immune control of disseminated LCMV infection depends on virus-specific CD8+ T cell responses that kill infected cells and express antiviral cytokines. CD4+ T cells play an essential role in supporting CD8+ T cell responses across time and promote anti-body responses that limit viral burdens. B6.PL mice showed lower viral burdens in the lung (Figure S1D), suggesting that B6.PL mice make elevated antiviral T cell responses that contribute to pathogenesis. B6, B6.PL, and F1 mice had similar numbers of splenocytes and T cells before infection (Figure S4A). At day 6 post-infection, a time preceding the catastrophic decline in health, B6.PL mice showed elevated frequencies of CD8+CD44^hi T cells in the spleen and lung, which corresponded to a marginal increase of T cell number in the spleen, but not lung (Figure S4B). Virus-specific tetramer+ CD8+ T cells were higher in frequency in the lungs and spleens of B6.PL mice compared with B6 mice, which corresponded to an increase in the number of DbGP₃₃+ CD8+ T cells in the lungs and a >2-fold increase in GP₃₃-specific and NP₃₉₆-specific CD8+ T cells in the spleens of B6.PL mice (Figures 3A and 3B). B6.PL mice generated more CD8+ T cells expressing IFNg in the spleen than B6 mice (Figure 3B, bottom) and had 2-fold to 3-fold more CD8+ T cells able to co-express IFNg and TNF (data not shown). Moreover, there were more CD4+CD44^hi T cells in B6.PL lungs and spleens at day 6 compared with B6 mice (Figure S4C). B6.PL mice had 2-fold to 5-fold more tetramer+ and cytokine-competent (IFNγ+ or IFNγ+TNF+) CD4+ T cells in the lungs or spleens compared with B6 mice (Figure 3C and data not shown). B6.PL splenic GP₆₁-specific CD4+ T cells produced more IFNγ at a per cell level (geometric mean fluorescence intensity [gMFI] 119 versus 75; p = 0.02, Student’s t test).

Figure 3. T Cells Are Required for Lethality.

(A–E) B6 and B6.PL mice were given LCMV-Clone13. T cell responses in the spleen and lung were quantified at day 6 post-infection (A–C). In other cohorts, survival was assessed in mice depleted of T cells before infection (D and E).

(A) GP₃₃-specific or NP₃₉₆-specific CD8+ T cell responses in the lung were quantified by tetramer staining or intracellular staining (ICCS). The top dot plots show an example of CD8+ T cells co-stained for DbGP₃₃ tetramer and CD44+ and the bottom plots show CD8+ T cell co-expression of IFNγ and TNF after GP_33–41 peptide stimulation. Numbers within the dot plots indicate the percentage of cells in each region. The graphs show the overall number of tetramer+ T cells (top) or IFNg+ T cells (bottom) per lung. Six mice per group; *p < 0.05 and **p < 0.01, Student’s t test.

(B) The total number of CD8+CD44^hiDbGP₃₃- tetramer+ or CD8+CD44^hiDbNP₃₉₆-tetramer+ T cells per spleen (top graphs) and number of specific CD8+ T cells that expressed IFNg (bottom graphs). Eleven mice per group; **p < 0.01 and ***p < 0.001, Mann-Whitney or Student’s t test.

(C) GP₆₁-specific CD4+ T cell responses in the lung were quantified by tetramer staining (top) or ICCS (bottom). The dot plots are gated on CD4+ cells. Graphs show the total number of CD4+CD44^hiGP₆₁-tetramer+ cells per lung or spleen (top). Bottom graphs show the number of GP₆₁-specific CD4+ T cells expressing IFNg per lung or spleen. Symbols represent individual mice, and bars depict mean ± SEM. Six to 11 mice per group; **p < 0.01 and ***p < 0.001, Mann-Whitney or Student’s t test.

(D) Cohorts of B6.PL mice were depleted of CD8+ T cells or given isotype control followed by LCMV-Clone13. Survival was measured until day 14.

Data pooled from two independent experiments with 6 mice per group. ***p < 0.001, Mantel-Cox analysis.

(E) Cohorts of B6.PL mice were depleted of CD4+ T cells or given isotype control antibody, followed by LCMV-Clone13 infection and monitored for survival to day 18. Data pooled from three independent experiments with 9 mice per group. **p < 0.01, Mantel-Cox analysis.

See also Figure S4.

Intravascular staining can distinguish T cells in capillaries from those in the parenchyma of tissues (Anderson et al., 2014). At day 6 of infection, cohorts of mice were given intravenous anti-body against CD8β, followed by lung and blood harvest moments later and ex vivo staining with anti-CD8α and tetramers (Figure S4D). The overall frequency of CD8+ T cells and tetramer+ CD8+ T cells was higher in the lungs and blood of B6.PL mice compared with B6. The intravascular antibody stained the vast majority of CD8+ T cells in the lungs and blood of both groups of mice, suggesting that infection weakens the endothelial barrier in the lung and allows circulating antibody to pass freely into lung parenchyma, consistent with the increased permeability of the lungs during infection (Figure 1E). The frequencies of CD8+ T cells and tetramer+ subsets were higher in lung compared with blood, suggesting that T cells are retained in the lung and possibly distinct from those in circulation. Similar findings were observed for CD4+ T cells and the GP₆₁-tetramer+ subset (Figure S4E): higher frequencies of CD4+ T cells were found in the lungs than blood, and B6.PL mice showed higher frequencies of CD4+ T cells or the tetramer+ subset in both compartments compared with B6.

CD8+ T cell depletion fully protected B6.PL mice, resulting in 100% survival to day 10, whereas isotype control-treated mice showed 0% survival (Figure 3D). Cohorts of CD4+ T cell-depleted mice showed partial protection (Figure 3E): 25% survived to day 14 when CD4+ T cells were depleted, compared with 0% given isotype control antibody. These data indicate that CD8+ T cells are essential for lethality, with CD4+ T cells also contributing to a lesser extent.

PL Sequence at Chromosome 9 Contributes to Lethal Outcome of Infection

To better understand the genetic basis of pathogenesis, we genotyped tail DNA from a male mouse using the Mouse Universal Genotyping Array (MegaMUGA) (Morgan et al., 2015). Genome-wide, B6 and B6.PL mice differ at 1.3% of SNPs, with a dense cluster of PL/J genotype on chromosome (Chr) 9 and a cluster of non-B6, non-PL/J genotypes on Chr11 that appears to be related to C57BL/10 (Figure S5A) (Morgan et al., 2015). C57BL/10 mice do not show lethality after LCMV infection (Leist et al., 1989), suggesting that the severe pathogenesis in B6.PL mice likely associates with the PL allele at Chr9:30–67 Mb.

Forward genetics was used to map the locus responsible for the lethal phenotype after infection. The initial goal was to evaluate whether the PL sequence at Chr9 or the non-B6, non-PL/J sequence at Chr11 was responsible for pathogenesis. F1 mice were backcrossed to B6 mice to generate N2 mice (Figure 4A) with B6/B6 (homozygous) or B6/PL (heterozygous) genotypes at Chr9 and Chr11. PCR primers were designed to distinguish B6 and PL sequence at the proximal (Siae) and distal regions of the Chr9 interval (Thsd4), and at one locus at Chr11 (Osm). N2 mice were infected and analyzed for weight loss, temperature, and survival and assigned a disease score according to the criteria described in Table S2. Forty-two N2 mice showing distinct phenotypes by day 10 were selected for PCR genotyping at Chr9 and Chr11 (Table S3). Among the 42 N2 mice, those with B6/PL heterozygous genotype at Chr9 (Siae) showed significantly higher (p < 0.05) disease score compared with those with B6/B6 homozygous sequence at Chr9 (Siae) (Figure 4B). Likewise, N2 mice with B6/PL genotype at Chr9 (Thsd4) showed higher disease score than mice with B6 homozygous genotype at Chr9 (Thsd4). Thus, inheritance of the PL sequence on Chr9 was associated with worsened outcome. In contrast, there was no significant association between genotype at Chr11(Osm) and disease score (Figure 4B). These results strongly implicate a genetic variant(s) near Chr9:38–60 Mb that is responsible for severe pathogenesis.

Figure 4. A PL Sequence on Chromosome 9 Associated with Pathogenesis.

Forward genetics and SNP genotyping were used to implicate chromosome 9 with pathogenesis.

(A) An illustration of the crosses to generate N2 mice and SNP locations used to discriminate B6 or B6.PL sequence at Chr9 or Chr11.

(B) N2 mice were infected and disease severity was scored on day 10 post-infection. The graph shows disease severity for mice that are B6/B6 or B6/B6.PL throughout the Chr9 or Chr11 loci. Data are presented as mean + SEM. Chr9: n = 20 for B6/B6 and 11 for B6/B6.PL; Chr11: n = 20 for B6/B6 and 20 for B6/B6.PL. *p < 0.05 or not significant, Mann-Whitney test.

(C) The scatterplot shows the disease score for 11 individual N2 mice representing B6/B6 or B6/B6.PL genotypes at Chr9. Each point represents an individual mouse with the mean ± SEM indicated. See also Tables S2 and S3 and Figure S5.

Focusing on Chr9, we noted that some N2 mice were recombinant between Siae and Thsd4 (Table S3). Six of nine of these recombinant mice were heterozygous at the proximal end of the interval (Siae) and had a lethal outcome of infection, even though they were B6/B6 at Thsd4 (Figure 4C). Two mice that were B6/B6 at Siae and B6/B6.PL at Thsd4 also succumbed to infection (Figure 4C). To better resolve the position of the causative locus, we identified additional recombinant N2 progeny and selected these for further backcrosses to B6 mice. These recombinant N2 mice were Siae^B6/PL, Thsd4^B6/B6 or Siae^B6/B6, Thsd4^B6/PL and were further genotyped at additional positions along Chr9 at 41, 44, 48, 56, 58, 60, and 66 Mb (data not shown). Mice were backcrossed to B6 mice to generate heterozygous N3 progeny that were subsequently intercrossed to generate N3(F1) (G4) mice. G4 siblings showing similar genotype at Chr9 were crossed to each other to establish colonies of G5 mice that were B6 congenic with PL/PL sequence along different intervals of Chr9 and were B6/B6 elsewhere (including Chr11), as determined by high-resolution MUGA genotyping (Figures 5A and S5A).

Figure 5. A Narrow Interval on Chromosome 9 Is Responsible for Lethal Pathogenesis.

PCR genotyping revealed a subset of N2 mice with evidence of meiotic crossover within chromosome 9. These were backcrossed to B6 and then intercrossed to generate “G5” mice with homozygous B6.PL sequence at Chr9:30–57 or at Chr9:55–66.

(A) GigaMUGA genotyping of B6, B6.PL, and several G5 lines of mice. Black represents homozygous B6 genotype; maroon, homozygous PL/PL; pink, B6/B6.PL. G5 mice were homozygous B6/B6 throughout the rest of the genome. The percentage of mice that perish following infection is indicated to the right.

(B) B6, B6.PL, G5–328, and G5–348 mice were given LCMV and analyzed for survival, weight loss, and temperature loss. A significant difference in survival for G5–328 compared with other groups is indicated (***p < 0.0001, Mantel-Cox test). A significant difference in weight loss between G5–328 and G5–348 at day 8 is indicated (**p < 0.005, Mann-Whitney test). A significant difference in temperature between G5–328 and G5–348 at day 8 is indicated (***p = 0.0001, unpaired

(legend continued on next page)

Student’s t test). Data are from two experiments: B6, n = 4 mice; B6.PL, n = 4 mice; G5–328, n = 14 mice; G5–348, n = 10 mice. Dashed lines indicate period when there was a loss of mice in the G5–328 group.

(C) Lung sections from day 8-infected mice stained with H&E. Magnification 403; yellow bar, 100 μm.

(D) Percentage of platelets in blood at days 8 and 9 after infection. n = 3 or 4 per group at each time point; *p < 0.05, Mann Whitney test.

(E) Splenic T cell responses at days 8 and 9. Top graphs depict the number of tetramer+ T cells per spleen. Middle row shows the gMFI of IFNg expression among peptide-responsive cells. Bottom row shows the gMFI of TNF expression after peptide stimulation. n = 3 or 4 per group at each time point; *p < 0.05 and **p < 0.01, Mann-Whitney test.

(F) Levels of virus in the serum or lungs of G5–328 and G5–348 mice at day 9. *p < 0.05, Mann-Whitney test.

(G) Whole-genome sequencing was performed on a B6.PL mouse and analyzed at Chr9:30–50 Mb. Each symbol represents a sequence variation away from the consensus B6 sequence. Variant Effect Predictor and SIFT analyses identified several sequence variants predicted to result in significant changes to associated proteins (red). Gene Ontology (GO) analysis revealed several of these genes are immune related (stars).

See also Table S4 and Figure S5.

Upon infection, G5 mice with PL genotype along Chr9:30–50 Mb interval (e.g., G5–328 and G5–307) showed poor survival compared with B6 mice or compared with G5–348 mice or G5–330 mice that were PL genotype only at Chr9:55–67 (Figure 5B). Compared with B6.PL mice, the G5–328 and G5–307 mice exhibited an intermediate phenotype in terms of the proportion of mice that perished (~60%–70%) and in the delay in disease course (about days 9–10) (Figures 5B and S5B). G5–328 mice showed significantly more weight loss at day 8 compared with B6 and G5–348 mice (Figures 5B), and among mice surviving beyond day 8, G5–328 mice continued to weigh less than the rest. Both G5–328 and G5–307 mice had significantly reduced body temperature at day 8 compared with B6 and G5–348 or G5–330 mice (Figures 5B and S5B), and G5–328 and G5–307 mice surviving beyond day 8 were slow to recover temperature. Thus, poor survival, extreme weight loss, and hypothermia were associated with cohorts of G5 mice containing PL sequence along the Chr9:30–57 region, whereas G5–330 and G5–348 mice containing PL sequence at Chr9:55–67 Mb did not show evidence of severe pathogenesis and more closely resembled B6 mice.

At day 8, lung edema was associated with the Chr9:30–57 interval, but not the Chr9:55–67 interval (Figure 5C). The G5–328 mice showed a trend toward fewer platelets at day 8 that achieved significance by day 9 (Figure 5D). The G5–328 mice showed same or increased virus-specific T cell numbers or expression of cytokine compared with G5–348 mice at day 8 or 9 (Figure 5E). Moreover, G5–307 mice had significantly more IFNg in blood at day 5 than B6 mice (Figure S5C). Levels of virus in blood were not different between groups, but G5–328 mice showed reduced levels of virus in the lung compared with G5–348 (Figure 5F), consistent with an increased T cell response in G5–328 mice. Cumulatively, these data show the PL allele at Chr9:30–57 is associated with improved T cell responses, reduced platelet frequencies, and worsened pathogenesis after infection.

Tail DNA from a male B6.PL mouse was subjected to whole genome sequencing and analyzed for genetic variations along the Chr9:30–50 Mb region. We found 223 putative non-synonymous variants within reported exons or changes to predicted splice sites (Table S4), including 18 not reported on the National Center for Biotechnology Information (NCBI) dbSNP database (https://www.ncbi.nlm.nih.gov/snp). Variant Effect Predictor (https://useast.ensembl.org/Mus_musculus/ Info/Index) (McLaren et al., 2016) predicted a total of 43 variants would have deleterious effects on protein expression or function (Table S4). SIFT analysis (https://useast.ensembl.org/Mus_musculus/Info/Index) predicted several of these variants would alter expression, protein structure, or function (Table S4). We used the Gene Ontology Browser and Phenotypes, Alleles & Disease Models Search at (http://www.informatics.jax.org) to identify genes within this interval that have immune-related functions. Of particular interest were changes to two immune-related genes that might result in enhanced immune responses (Figure 5G; Table S4). We found a missense mutation in Siae, a gene that encodes sialic acid acetyl esterase and is involved in recruiting inhibitory SIGLEC receptors to limit immune cell signaling (Mahajan and Pillai, 2016). We identified a 10 nt deletion in Crtam that reduces surface expression of CRTAM (MHC class 1-restricted T cell-associated molecule; data not shown), a T cell and NK or NKT cell-associated activation molecule that is involved in T cell retention in peripheral tissues (Cortez et al., 2014). In total, our findings reveal genetic variants within Chr9 of B6.PL that are associated with lethal pathogenesis during arenavirus infection.

DISCUSSION

Human arenaviruses cause periodic outbreaks associated with severe hemorrhagic disease. An improved understanding of how severe illness unfolds could improve efforts to develop modern treatments or predict who may need additional attention during an outbreak. Herein, we describe B6 congenic mice that reproduce many features of viral hemorrhagic disease in humans, including loss of platelets, breakdown of vascular integrity, lung edema, and lethality. We identified a critical locus on Chr9 that is responsible for lethality when present in hematopoietic cells, and we demonstrate that T cells are essential to pathogenesis.

Other models of LCMV-induced pathogenesis have implicated lung injury, platelet loss, cytokine storm, and IFNg- induced nitric oxide and shock (Baccala et al., 2014; Christiaansen et al., 2017; Oldstone et al., 1993, 2018; Puglielli et al., 1999; Remy et al., 2017; Schnell et al., 2012). Similar to our findings with B6.PL mice, NZB, NZO, and FVB strains of mice show thrombocytopenia, vascular permeability, soaring inflammatory cytokine expression, and lethality following LCMV-Clone13 (Baccala et al., 2014; Oldstone et al., 1993, 2018; Puglielli et al., 1999; Schnell et al., 2012). However, some differences were noted. For example, infected FVB mice show petechial lesions and hepatocellular necrosis (Schnell et al., 2012), whereas B6.PL mice displayed no apparent petechial lesions at necropsy and minimal liver enzyme release (Figure S1G and data not shown). FVB mice show limited lung edema and fewer infiltrating leukocytes (Schnell et al., 2012), whereas these were prominent features of B6.PL (Figure 1D) and NZB mice (Baccala et al., 2014; Puglielli et al., 1999). It is likely that strain-specific mutations influence these phenotypic differences.

T cells play a major role in immune-mediated pathogenesis to LCMV-Clone13, and dysregulated CD8+ T cells responses can result in CTL-mediated destruction of vascular endothelial cells (Frebel et al., 2012). B6.PL mice generate a vigorous T cell response early on (Figures 3 and S4). Likewise, susceptible NZB mice generate stronger CD8+ T cell responses than resistant BALB/c mice (Baccala et al., 2014); susceptible B6 mice expressing H-2^bxd (D2B6F1 mice) generate stronger CD8+ T cell responses than resistant BALB/c mice expressing H-2^bxd (BCF1mice) (Christiaansen et al., 2017). The contribution of CD4+ T cells to pathogenesis also varies with each stain, perhaps because of differences in the overall size or activity of the CD4+ T cell response on the different genetic backgrounds. For example, CD4+ T cell depletion protects FVB mice (Schnell et al., 2012) yet fails to fully protect NZB (Baccala et al., 2014; Puglielli et al., 1999) or B6.PL mice (Figure 3E).

B6.PL mice showed a significant loss of platelet frequencies. Platelet loss during LCMV is linked to IFN-I signaling, which can inhibit megakaryocyte maturation or induce apoptosis (Binder et al., 1997; Iannacone et al., 2008), and there may be a role for T cells, if infected megakaryocytes (Binder et al., 1997) are destroyed by CTL. T cells may disrupt the lung, a major site of platelet biogenesis (Lefrançais et al., 2017), during infection. IFNg can stimulate macrophages to phagocytose platelets (Zoller et al., 2011), and we find that B6.PL mice had T cells with increased expression of IFNg (Figure 3), their blood accumulated significantly more IFNg (Figure S5C), B6.PL macro-phages were more activated (Figure S3F), and we observed more B6.PL macrophages with ingested apoptotic bodies or cells (Figure S3G). Thus, it is plausible that increased T cell activity may lead to innate mechanisms that reduce platelet frequencies.

We assessed the genetic basis of pathogenesis. F1 mice resembled B6.PL mice except that lethality was delayed 2 days and 60% of F1 mice succumbed rather than 100% for B6.PL mice (Figure 1), consistent with an allelic variant that confers a semi-dominant phenotype. Likewise, N2 mice with heterozygous B6/PL sequence throughout the Chr9 interval resembled F1 mice in the proportion of mice that perish and in the delay in death, suggesting a semi-dominant alteration at Chr9. Despite having two copies of the B6.PL sequence at the proximal Chr9 locus, G5–328 and G5–307 mice failed to show accelerated disease, suggesting there may be a PL modifier at Chr9:55–67 that worsens outcome when PL is present at the proximal locus.

B6.PL sequence data were analyzed at the Chr9 locus to identify mutations that would alter amino acid sequence or exon use and possibly improve T cell responses. We identified several interesting mutations in genes relevant to the immune system that are predicted to alter protein function. For example, the B6.PL missense mutation at Siae may affect the activity of the resulting enzyme, possibly preventing the recruitment of inhibitory SIGLEC receptors on immune cells. The deletion at Crtam results in a frameshift; we observed that B6.PL T cells fail to express CRTAM on the cell surface on the basis of flow cytometry surface staining following stimulation (data not shown). The absence of surface CRTAM may affect T cell interaction with CD8+ DCs and epithelial cells, alter the cytotoxic or cytokine functions of T cells, or affect T cell retention in infected peripheral tissues. FVB mice share the B6.PL mutation at Crtam (Table S4), suggesting that overlapping phenotypes in these mice may be caused by the same variants, despite genetic divergence elsewhere in the genome. The biological significance of other alterations at long non-coding RNAs (Gm3898, Gm17540), and at Ubash3a, a negative regulator of TCR signaling, awaits further investigation. Finally, it may be that B6.PL includes T cell promoting genetic elements that are absent in B6.

STAR★METHODS

CONTACT FOR REAGENT AND RESOURCE SHARING

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Jason Whitmire (jwhitmir@email.unc.edu).

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Mice

C57BL/6J (B6), C57BL/6NJ, B6.PL-Thy1^a/CyJ (B6.PL), and PL/J mice were purchased from Jackson Laboratories (Bar Harbor, Maine) during the period 2011– 2016. The mice were bred and housed in a facility managed by the Division of Comparative Medicine at UNC-CH in accordance with the policies and guidelines of the Institutional Animal Care and Use Committee (IACUC). Adult (8–10 weeks old) male and female mice were used for experiments. Mice received an intravenous injection of 2×10⁶ plaque-forming units (PFU) of LCMV-Clone13, unless indicated otherwise. All experiments were approved by the UNC-CH IACUC.

Virus

Viral stocks of plaque-purified LCMV-Clone13 were prepared from infected BHK-21 monolayers that tested negative for mycoplasma. Virus titer in blood and tissues was performed by plaque assay on Vero cell monolayers.

Cell isolation and purification

Single-cell suspensions were prepared from spleen, bone marrow, or lung. Spleens were physically disrupted over a 70 μm nylon cell strainer (Corning, NY). Bone marrow was collected by flushing the femurs of mice with RPMI followed by filtration through a 70 μm cell strainer. Erythrocytes were removed from spleen and bone marrow suspensions using ACK lysing buffer (Life Technologies-BRL, NY). Lungs were cut into small pieces and digested at 37°C for 30 min with 1mg/ml collagenase (Calbiochem) and 10mg/ml DNase I (Sigma-Aldrich) in RPMI before filtering through a 70 μm cell strainer. The cells were exposed to ACK to eliminate red blood cells. All single-cell preparations were rinsed and re-suspended in 10% RPMI media.

Cell culture

Vero cells and BHK cells were propagated in DMEM supplemented with 5% heat inactivated FBS, penicillin, streptomycin, and fungizone.

METHOD DETAILS

Bone marrow chimeras

Bone marrow chimeras were generated by intravenously reconstituting lethally irradiated (2 × 600 RADs) mice with 4 × 10⁶ cells bone marrow cells. Following irradiation and reconstitution, mice were given drinking water with sulfamethoxazole (6 mg/ml) for two consecutive weeks. Chimerism was assessed at 8 weeks post-transfer by quantifying donor versus host T cell frequencies in the blood. Mice were rested for an additional month (day 90 post-transfer) before infection.

T cell depletion

To deplete CD8+ T cells, mice were given (ip) 250 μg anti-CD8 (clone 2.43) at day — 1 and day 3 post-infection, or were given a non-depleting, isotype control antibody. CD4+ T cells were depleted using ip injections of 250 μg anti-CD4 (clone, GK1.5) at day — 1 and day 3 post-infection, or the mice were given a non-depleting, isotype control antibody.

Vascular permeability

Evans Blue dye was delivered to mice intravenously. The dye binds albumin and is maintained in circulation when endothelial cells are intact but diffuses into tissues when the integrity of the endothelium is compromised. 30 minutes after injection, tissues were collected and the level of Evans Blue retained by the tissues was quantified by a colorimetric assay using an ELISA reader with a standard curve.

Histology

Tissues were harvested at necropsy and fixed in 10% neutral phosphate-buffered formalin for 48 hours and stored in 70% ethanol until processed for histology. Formalin-fixed paraffin-embedded tissues were sectioned at 4 or 5 μm thickness for histology. Sections were stained with hematoxylin and eosin (H&E) and examined for histological changes by light microscopy (an Olympus BX61 microscope) at UNC Microscopy Services Laboratory. All histological processes were conducted by Animal Histopathology Core Lab, Lineberger Comprehensive Cancer Center, UNC-CH.

Flow cytometry

Single cell suspensions of spleen, lung, or bone marrow cells were stained directly ex vivo with fluorochrome-conjugated anti-CD4, anti-CD8, anti-CD44, and Streptavidin-APC-conjugated tetramers (DbGP_33–41, DbNP_396–404, and I-AbGP₆₇). The cell-surface staining reaction was done in the presence of unlabeled antibodies against Fc-receptors to block fluorochrome-conjugated antibodies from binding to FcR+ cells. The intracellular staining (ICCS) assay was performed by culturing splenocytes with or without LCMV peptide in the presence of brefeldin A. After 5 h of incubation, cells were stained for surface markers, washed, fixed with formaldehyde, then permeabilized and exposed to mAbs specific for IFN-γ and TNF. Antibody stained cells were detected by a FACSCalibur cytometer (BD Biosciences) and the data were analyzed with FlowJo software.

Intravascular staining

This method distinguishes T cells located in the parenchyma of tissues from those in capillaries (Anderson et al., 2014). Mice were intravenously injected with FITC-conjugated anti-CD4 (clone RM4–5) or anti-CD8β to label T cells in circulation. Three minutes later, blood and lung were harvested and single cell suspensions were re-stained with PE-conjugated anti-CD4 (clone GK1.5) or anti-CD8α and APC-conjugated tetramers to identify virus-specific T cells.

Hematocrit test

Blood samples were collected into heparinized micro-hematocrit tubes, and hematocrit values were measured after centrifugation.

Complete Blood Count (CBC)

Peripheral blood analysis performed on a Procyte Dx Hematology Analyzer (IDEXX Laboratories, Inc.) at the Animal Histopathology & Laboratory Medicine Core at UNC.

Platelet analyses

Blood (50 μL) was drawn from the tail artery and diluted into an equal volume of 30 IU/mL of low molecular weight heparin. Platelets in blood were identified by flow cytometry using forward and side scatter to exclude debris, along with fluorescent antibodies to CD42d, CD41, and CD62P. Platelet size was determined by Forward Scatter Height.

IFN-β and IFN-α ELISA

Blood was collected at 19 hours post-infection. IFN-b was quantified by ELISA, using plates coated with capture antibody (anti-IFN-β; Santa Cruz Biotechnology) and bound IFN-β detected with rabbit polyclonal anti-IFN-β (PBL). IFN-α levels were determined using VeriKine mouse IFN-α ELISA kit, following the manufacturer’s protocol (PBL Assay Science).

RT-PCR

Total RNA was extracted from lung using RNeasy mini kit (QIAGEN, Valencia, CA). Quantitative PCR was performed using iTaq universal SYBR Green one-step kit (Bio-Rad) with a 7900 HT Real-Time PCR System (Applied Biosystems). Samples were analyzed in duplicate and normalized to β-actin expression. PCR was performed using IFNg_FW and IFNg_RV primers to amplify the IFNγ region; and TNF_FW and TNF_RV primers to amplify the TNF region; IL6_FW and IL6_RV primers to amplify the IL-6 region; and β-actin_FW and β-actin_RV primers to amplify the β-actin region.

Mouse genotyping arrays

MegaMUGA and GigaMUGA arrays (https://csbio.unc.edu/CCstatus/index.py?run=Genotype) were used for high resolution SNP-based genotyping of tail DNA (Morgan et al., 2015). These arrays contain markers that are uniformly spaced throughout the mouse genome and can distinguish numerous lines of inbred mice.

High-throughput DNA sequencing, read alignment, and bioinformatics

The deep sequencing and analysis was performed using tail DNA from one adult male B6.PL mouse at the UNC High Throughput Genomic Sequencing Facility. We used KAPA Hyper (KAPA Biosystems) kit for WGS library preparation. Size selection of constructed libraries was performed using Pippin prep (SAGE Science). DNA libraries were used to obtain Paired-End 125 bp (V4 chemistry) reads from HiSeq 4000 sequencer (Illumina). DNA library was sequenced on 1 lane with an average coverage of 14X for whole genome. The raw FastQ files were filtered to keep only read pairs with an average base quality score of 20 or more and were then aligned to the reference mouse genome (B6) using CLC Genomic Workbench package and default parameters (not uniq mapped reads were removed). Duplicate reads were removed from the resulting bam file. The bam files were processed according to the variant caller from CLC Genomic Workbench package with default parameters for diploid genomes.

Variant analysis

B6.PL sequence was compared to consensus C57BL/6J (GRCm38.p5) sequence. All non-synonymous variants and splice variants were selected within the Chr9:30–50Mb region. Variant Effect Predictor (VEP) (http://useast.ensembl.org/Tools/VEP) was used to identify the likely effect of each variant on exon splicing, gene expression, or protein sequence, as well as effects on putative intergenic regulatory regions. The effects of the variants were evaluated using the canonical transcript for each gene. Sorting Intolerant From Tolerant (SIFT) analysis software (https://useast.ensembl.org/Mus_musculus/Info/Index) predicted whether changes to coding sequence in B6.PL would impact protein structure or function. Highly significant changes were identified by either a SIFT score of < 0.05 and/or a VEP “Putative Impact Score” of high (McLaren et al., 2016). Immune-related genes were defined by the Gene Ontology Browser at the JAX database (http://www.informatics.jax.org), using the “Phenotypes, Alleles & Disease Models Search.”

PCR-genotyping

Tail DNA was isolated using a DNeasy isolation kit (QIAGEN) according to the manufacturer’s protocol. Primers were designed for specific SNP markers using Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). Oligonucleotide primers were custom-synthesized through UNC Genomics Core Facilities and Resources. PCR amplification of DNA fragments (185 – 788 bp) was performed with the specific primers and Taq DNA polymerase (Invitrogen). The PCR conditions were: 95° C, 3 minutes; [95° C for 30 s; 55–65° C for 30 s; 72° C for 90 s] × 35 cycles; 72° C for 3 minutes. PCR products were cleaned with ExoSAP-IT (Affymetrix) prior to sequencing by SimpleSeq Webless (Eurofins Genomics). All sequence profiles were visually checked using 4 peaks software (https://nucleobytes.com).

QUANTIFICATION AND STATISTICAL ANALYSIS

All data were subjected to D’Agostino-Pearson normality test. When dataset was normally distributed and passed the normality test, parametric tests were conducted using unpaired two-tailed Student’s t test for two groups or one-way analysis of variance (ANOVA) with Bonferroni multiple comparison test for more than two groups. When data were not normally distributed, non-parametric tests were used (Mann-Whitney U test for two groups, Kruskal-Wallis with Dunn’s multiple comparison test for more than two groups). Statistical analyses and graphing were done with Prism software (GraphPad Software Inc.). P values considered significant are indicated in figure legends as * = p < 0.05; ** = p < 0.01; *** = p < 0.001; ns (not significant) = p > 0.05. All data are presented as mean ± SEM values.

DATA AND SOFTWARE AVAILABILITY

The accession numbers for the deep sequencing data for B6.PL are deposited at NCBI under record number: sra/SRP128734 or SRX3549454.

KEY RESOURCES TABLE

REAGENT or RESOURCE		SOURCE	IDENTIFIER
Antibodies
TruStain fcX (anti-mouse CD16/32)		Biolegend	Cat#101320; RRID: AB_1574975
Anti-mouse CD4, APC, clone RM4–5		Biolegend	Cat#100516; RRID: AB_312719
Anti-mouse CD4, FITC, clone RM4–5		Biolegend	Cat#100509; RRID: AB_312712
Anti-mouse CD4, PE, clone GK1.5		Biolegend	Cat#100407; RRID: AB_312692
Anti-mouse CD4, PE, clone RM4–5		Biolegend	Cat#100511; RRID: AB_312714
Anti-mouse CD8, APC, clone 53–6.7		Biolegend	Cat#100712; RRID: AB_312751
Anti-mouse CD8, PE, clone 53–6.7		Biolegend	Cat#100707; RRID: AB_312746
Anti-mouse CD8β,FITC,clone YTS156.7.7		Biolegend	Cat#126605; RRID: AB_961293
Anti-mouse CD44, FITC, clone IM7		Biolegend	Cat#103006; RRID: AB_312957
Anti-mouse IFNβ, clone 7F-D3		Santa Cruz Biotech	Cat#sc57201;RRID: AB_2122911
Anti-mouse IFNpβ, Rabbit polyclonal		PBL	Cat#32400–1; RRID: AB_387872
Anti-mouse IFN-γ, FITC, clone XMG1.2		Biolegend	Cat#505806; RRID: AB_315400
Anti-mouse TNF, APC, clone MP6-XT22		Biolegend	Cat#506308; RRID: AB_315429
Anti-mouse/human CD11b, APC, clone M1/70		Biolegend	Cat#101211; RRID: AB_312794
Anti-mouse I-A/I-E, FITC, clone M5/114.15.2		Biolegend	Cat#107605; RRID: ABJ313320
Anti-mouse F4/80, PE, clone BM8		Biolegend	Cat#123109; RRID: AB_893498
Anti-mouse Ly-6G,PerCP/Cy5.5,clone1A8		Biolegend	Cat#127615; RRID: ABJ877272
Anti-mouse CD42d, APC, clone 1C2		Biolegend	Cat#148505; RRID: AB_2564601
Anti-mouse CD41, FITC, clone MWReg30		Biolegend	Cat#133903; RRID: AB_1626237
Anti-mouse CD62P, PE, clone RMP-1		Biolegend	Cat#148305; RRID: AB_2565274
InVivoMAb anti-mouse CD8α, clone 2.43		BioXcell	Cat#BE0061; RRID: AB_1125541
InVivoMAb anti-mouse CD4, clone GK1.5		BioXcell	Cat#BE0003–1; RRID: AB_1107636
InVivoMAb rat lgG2b isotype control		BioXcell	Cat#BE0090; RRID: AB_1107780
Bacterial and Virus Strains
LCMV (Clone 13)		Whitmire laboratory	N/A, generated in house
Chemicals, Peptides, and Recombinant Proteins
Brefeldin A Solution (1,000X)		Biolegend	Cat#420601
Fixation Buffer		Biolegend	Cat#420801
Intracellular Permeabilization Wash Buffer		Biolegend	Cat#421002
Biotinylated DbGP33–41 monomer		NIH Tetramer core	N/A
Biotinylated DbNP396–404 monomer		NIH Tetramer core	N/A
APC-conjugated l-AbGP67 tetramer		NIH Tetramer core	N/A
Streptavidin-Allophycocyanin		Biolegend	Cat#405207
Heparin		Sanofi-Aventis	Lovenox
RPMI		Lonza	Cat#12–167F
FBS		GIBCO	Cat#26140–079
ACK lysing buffer		GIBCO	Cat#A1049201
Collagenase, Type IV		GIBCO	Cat#17104–019
Percoll		GE Healthcare Percoll Centrifugation Media	Cat#17089102
DNase I		Sigma-Aldrich	Cat#D4527
ExoSAP-IT		Affymetrix	Cat#78200
Continued
REAGENT or RESOURCE	SOURCE			IDENTIFIER
Critical Commercial Assays
RNeasy Mini Kit	QIAGEN			Cat#74106
MaxDiscovery ALT Color Endpoint Assay	Bioo Scientific			Cat#3460–08
Mouse TNF-alpha DuoSet ELISA	R&D systems			Cat#DY410
VeriKine Mouse Interferon Alpha ELISA	PBL Assay Science			Cat#42120–1
iTaq universal SYBR Green one-step kit	Bio-Rad			Cat#1725150
KAPA Hyper WGS library preparation kit	KAPA Biosystems			Cat#KK8500
DNeasy isolation kit	QIAGEN			Cat#69504
Deposited Data
B6.PL deep sequencing	https://www.ncbi.nlm.nih.gov/ sra/SRP128734			N/A
Experimental Models: Cell lines
Vero-E6	Michael Buchmeier			The Scripps Research Institute
BHK-21	American Type Culture Collection			Cat#CCL-10
Experimental Models: Organisms/Strains
Mouse: C57BL/6J	Jackson Laboratory			Cat#000664
Mouse: C57BL/6NJ	Jackson Laboratory			Cat#005304
Mouse: B6.PL-Thy1a/CyJ	Jackson Laboratory			Cat#000406
Mouse: PL/J	Jackson Laboratory			Cat#000680
Oligonucleotides
See Table S5 for oligonucleotide information				N/A
Software and Algorithms
Flo-Jo Software (version 9.8.3)	FlowJo			https://www.flowjo.com/
GraphPad Prism 7	GraphPad Software			https://www.graphpad.com
CLC Genomic Workbench	QIAGEN			https://www.qiagenbioinformatics.com/
4 peaks software	Nucleobytes			https://nucleobytes.com/4peaks/
Mouse phylogeny viewer	csbio.unc.edu			http://msub.csbio.unc.edu/ttviewer
Varient Effect Predictor	elEnsembl.org			http://useast.ensembl.org/ToolsA/EP
SIFT (Sorting Intolerant From Tolerant)	elEnsembl.org			https://useast.ensembl.org/Mus_musculus/ Tools/VEP
Gene Ontology Browser	JAX			http://www.informatics.jax.org
Other
SimpleSeq Webless service	Eurofins Genomics			https://www.eurofinsgenomics.com/en/ home.aspx

Highlights.

• B6.PL mice develop lethal hemorrhagic disease during disseminated LCMV infection

• The PL allele enhances T cell responses, thrombocytopenia, and lung edema

• The PL allele at a chromosome 9 locus includes alterations in immune-related genes

• Limited genetic differences in the host can remarkably alter outcome to infection