Early-life infection depletes preleukemic cells in a mouse model of hyperdiploid B-cell acute lymphoblastic leukemia

Abstract

Epidemiological studies report opposing influences of infection on childhood B-cell acute lymphoblastic leukemia (B-ALL). Although infections in the first year of life appear to exert the largest impact on leukemia risk, the effect of early pathogen exposure on the fetal preleukemia cells (PLC) that lead to B-ALL has yet to be reported. Using cytomegalovirus (CMV) infection as a model early-life infection, we show that virus exposure within 1 week of birth induces profound depletion of transplanted E2A-PBX1 and hyperdiploid B-ALL cells in wild-type recipients and in situ-generated PLC in Eμ-ret mice. The age-dependent depletion of PLC results from an elevated STAT4-mediated cytokine response in neonates, with high levels of interleukin (IL)-12p40-driven interferon (IFN)-γ production inducing PLC death. Similar PLC depletion can be achieved in adult mice by impairing viral clearance. These findings provide mechanistic support for potential inhibitory effects of early-life infection on B-ALL progression and could inform novel therapeutic or preventive strategies.

Graphical abstract

Figure 1.

Depletion of B-ALL–associated cells after timed MCMV infections. (A) Neonatal and adult BALB/c mice injected with 1 to 2 × 10³ Eμ-ret leukemic cells were euthanized at 10 dpi to assess leukemic cell burden in spleens of phosphate-buffered saline (PBS)-injected (control [Ctrl]) or virus-infected (CMV) recipients (Mann-Whitney; n = 14 Ctrl and 16 CMV neonates; n = 14 Ctrl and 11 CMV adults; pooled results for 3 different leukemias). (B) Neonatal and adult BALB/c mice injected with 0.5 to 1 × 10³ Eμ-ret leukemic cells were infected 5 to 6 days later with PBS (Ctrl) or virus (CMV) and then monitored for leukemia-free survival of recipients (log-rank [Mantel-Cox] test; n = 6-10 mice in each group; pooled results from 2 independent experiments). (C) Neonatal and adult C57BL/6 mice injected with 0.5 to 1 × 10³ E2A-PBX1 leukemic cells were euthanized 10 days after subsequent infection to assess leukemia burden in spleens of PBS- or CMV-infected recipients (Mann-Whitney; n = 18 Ctrl and 15 CMV-infected neonates; n = 12 Ctrl and 11 infected adults; pooled results from 4 independent experiments). (D) Splenic PLC burden in Eμ-ret mice 10 days after injection with PBS or CMV at the indicated age (Mann-Whitney test; number of mice in Ctrl/infected groups: week 1 [wk1] = 21/21, wk2 = 11/6, wk3 = 8/5, and wk4 = 12/11). (E) Spleen (n = 11 Ctrl; n = 10 CMV) and BM (n = 12 Ctrl; n = 6 CMV) B-cell Fr cell numbers in neonatal Eμ-ret mice 10 days after injection with PBS or CMV at 5 to 7 days of age (Mann-Whitney test). (F) Leukemia-free survival of neonatal Eμ-ret mice after PBS (Ctrl) or virus (CMV) injection at 5 to 7 days of age (log-rank [Mantel-Cox] test; n = 23 Ctrl; n = 17 CMV; ∗P = .0166). All data presented as mean ± standard error of the mean (SEM). ns, not significant; ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. Ctrl, control.

Figure 2.

Apoptosis of PLC after MCMV infection is STAT4-dependent. (A) Comparison of PLC numbers in spleen (left) and BM (right) of PBS- (blue) or CMV-injected (red) 1-week-old Eμ-ret pups at 6 and 10 dpi (Mann-Whitney t test; n = 7 PBS and 10 CMV neonates; pooled results from 3 independent experiments). (B) Representative images of bioluminescence in 1-week-old Eμ-ret pups infected with luciferase-tagged CMV and uninfected controls at 3, 6, and 10 dpi. (C) Polymerase chain reaction for detecting CMV DNA in 3 different concentrations of virus equal to 200, 20, and 2 pfu as controls, 3 unsorted total splenocytes, and 3 sorted PLC samples from spleens of wild-type Eμ-ret pups 7 days after infection from 3 independent experiments. (D) Comparison of PLC numbers in representative (left panels) or cumulative (right panels) spleens of PBS- or CMV-injected Stat4^–/– (R/Stat4) and Stat6^–/– (R/Stat6) Eμ-ret pups at 10 dpi (Mann-Whitney test; R/Stat4, n = 11 PBS and 10 CMV; R/Stat6, n = 5 PBS and 8 CMV; pooled results from 3 independent experiments). (E) Representative (left panel) or cumulative (right panel) bioluminescence in wild-type (R/wt) and Stat4^–/– (R/Stat4) Eμ-ret pups at 3, 6, and 10 dpi with luciferase-tagged CMV (2-way analysis of variance [ANOVA]; n = 5-10) (for simplicity, nonsignificant comparisons are not labeled). (F) Cumulative results for assessment of caspase 3/7 activity in PLC (CD19⁺, BP1⁺) from spleen and BM of Eμ-ret pups 7 days after PBS (Ctrl) or virus (CMV) injection (Mann-Whitney t test; PBS, n = 6; MCMV, n = 7). Data shown as mean ± SEM. ns, not significant; ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001.

Figure 3.

Depletion of PLC is dependent on IL-12p40. (A) Enzyme-linked immunosorbent assay (ELISA)–based quantification of IL-12p40 (anti-p40 antibody [Ab]), IL-12 (anti-p70 Ab), and IL-23 (anti-p19 Ab) in sera from neonate and adult Eμ-ret mice at 3, 6, and 10 dpi (Mann-Whitney; n = 3-7). (B) Representative flow cytometry plots showing PLC in splenocytes of Eμ-ret pups after CMV infection in the presence of indicated neutralizing antibodies. (C) Cumulative results showing PLC burden in spleens of Eμ-ret pups infected along with p40, p19, and/or p70 neutralizing antibodies (1-way ANOVA; n = 4-11). (D) PLC numbers in spleens of PBS (Ctrl) or virus (CMV)-injected Rag1^–/– (R/Rag, left panel), wild-type (R/wt), and NK-depleted wt (NK-dep, right panel) Eμ-ret pups at 10 dpi (Mann-Whitney; R/Rag1, n = 10 PBS and 10 MCMV; 1-way ANOVA; n = 8 PBS, 8 CMV, and 7 CMV + NK-depleted; pooled results from 3 independent experiments). All data presented as mean ± SEM (∗∗∗P < .001; ∗∗∗∗P < .0001). ns, not significant (nonsignificant comparisons in graphs A and C are not labeled).

Figure 4.

Gene expression analysis indicates pathway utilization changes in PLC after neonatal MCMV infection. Selected gene sets from the Molecular Signatures Database, Molecular Hallmarks (hallmark gene sets) collection, enriched among upregulated and downregulated pathways in PLC from MCMV-infected neonates compared with PBS-injected controls (ranked according to decreasing normalized enrichment score). (A) Top 16 statistically significant upregulated biological pathways. (B) GSEA enrichment plots showing upregulation of gene sets representing IFN-α, IFN-γ, and TNF-α, and inflammatory responses. (C) Top 14 downregulated biological pathways. (D) GSEA enrichment plots showing downregulation of gene sets representing cell cycle and metabolism. ES, enrichment score; FDR, false discovery rate; NES, normalized enrichment score; NOM, nominal; size, number of genes in each set.

Figure 5.

Inflammatory responses following MCMV infection of neonatal and adult Eμ-ret mice. (A) Flow cytometry–based quantification of 15 proinflammatory/inflammatory cytokines and chemokines that were significantly higher in the serum of MCMV-infected pups (1-week-old) compared with adult (4-week-old) Eμ-ret mice at 3, 6, or 10 dpi (2-way ANOVA; comparison between PBS-injected [Ctrl] and infected [CMV] conditions within each group have been shown; n = 3-5). (B) Heat map depicting the inflammatory response. Normalized means (averaged from 3-5 samples per group) calculated for each cytokine by taking smallest mean as 0% and largest mean as 100%. (C) Representative bioluminescence images of Eμ-ret pups at 3, 6, and 10 days after infection with luciferase-tagged MCMV in the presence of the indicated neutralizing antibodies. (D) PLC burden in the spleen of PBS (Ctrl) and virus (MCMV)-infected Eμ-ret pups also injected with the indicated neutralizing antibodies (1-way ANOVA; n = 3-5). (E) ELISA-based quantification of p40 and flow cytometry–based quantification of IFN-γ, and TNF-α in sera from control and MCMV-infected wild-type (R/wt), Stat4^–/– (R/Stat4), IFN-γ–neutralized, and p40–neutralized Eμ-ret mice at 6 dpi (2-way ANOVA; followed by Tukey test; n = 3-4). For the simplicity of graphs, comparisons between MCMV-infected groups are shown. All data are presented as mean ± SEM. ∗P < .05; ∗∗ P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001 (nonsignificant comparisons are not labeled).

Figure 5.

Inflammatory responses following MCMV infection of neonatal and adult Eμ-ret mice. (A) Flow cytometry–based quantification of 15 proinflammatory/inflammatory cytokines and chemokines that were significantly higher in the serum of MCMV-infected pups (1-week-old) compared with adult (4-week-old) Eμ-ret mice at 3, 6, or 10 dpi (2-way ANOVA; comparison between PBS-injected [Ctrl] and infected [CMV] conditions within each group have been shown; n = 3-5). (B) Heat map depicting the inflammatory response. Normalized means (averaged from 3-5 samples per group) calculated for each cytokine by taking smallest mean as 0% and largest mean as 100%. (C) Representative bioluminescence images of Eμ-ret pups at 3, 6, and 10 days after infection with luciferase-tagged MCMV in the presence of the indicated neutralizing antibodies. (D) PLC burden in the spleen of PBS (Ctrl) and virus (MCMV)-infected Eμ-ret pups also injected with the indicated neutralizing antibodies (1-way ANOVA; n = 3-5). (E) ELISA-based quantification of p40 and flow cytometry–based quantification of IFN-γ, and TNF-α in sera from control and MCMV-infected wild-type (R/wt), Stat4^–/– (R/Stat4), IFN-γ–neutralized, and p40–neutralized Eμ-ret mice at 6 dpi (2-way ANOVA; followed by Tukey test; n = 3-4). For the simplicity of graphs, comparisons between MCMV-infected groups are shown. All data are presented as mean ± SEM. ∗P < .05; ∗∗ P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001 (nonsignificant comparisons are not labeled).

Figure 6.

Exogenous IFN-γ replicates BCP Fr–specific depletion of PLC. (A) Change in cell number of flow-sorted PLC cocultured on stromal cells in the absence (Ctrl) or presence (IFN-γ) of 1-ng/mL IFN-γ for 60 hours. Quantification of PLC was achieved by Incucyte (Sartorius) live-cell analysis at 6-hour intervals. (B) PLC burden in the spleens of noninfected wild-type (R/wt) Eμ-ret pups after administration of different concentrations of IFN-γ (1, 2, or 5 μg/dose) every 2 days, from 9 to 16 days of age (1-way ANOVA; n = 3-5; ∗∗P < .005). (C) PLC burden in spleens of noninfected Stat4^–/– (R/Stat4) Eμ-ret mice injected with 5 μg/dose IFN-γ every 2 days, from 9 to 16 days of age (Mann-Whitney test; n = 9-10; ∗∗∗P < .005). (D) Spleen (left panel) and BM (right panel) B-cell Fr burden in PBS (Ctrl) and 5 μg/dose IFN-γ–injected neonatal Eμ-ret mice (Mann-Whitney test; spleen: n = 4 Ctrl and 5 IFN-γ; and BM n = 6 Ctrl and 10 IFN-γ; ∗P < .05). All data are presented as mean ± SEM. Nonsignificant comparisons in all graphs are not labeled.

Figure 7.

Impaired viral clearance generates a PLC-depleting IFN-γ–mediated immune response in adult mice. (A) Representative images and overall plot of bioluminescence in 4-week-old Eμ-ret pups infected with luciferase-tagged CMV with and without NK cell depletion (and uninfected controls) at 3, 6, and 10 dpi. For NK cell depletion, 50-μg asialo GM1 antibody was administered at days –3, –1, +1, +4, and +7 (day 0 was the time of infection) (2-way ANOVA; n = 4-8; ∗∗∗∗P < .0001). (B) ELISA-based quantification of IL-12p40 (anti-p40 Ab) and IFN-γ (anti-IFN-γ Ab) in sera from PBS-injected control and CMV-infected (with and without NK depletion) adult Eμ-ret mice at 10 dpi (1-way ANOVA; n = 6-10; ∗∗P < .005; ∗∗∗∗P < .0001). (C) PLC burden in the spleens of PBS-injected control and CMV-infected (with and without NK depletion) adult Eμ-ret mice at 10 dpi (1-way ANOVA; n = 10-16; ∗P < .05; ∗∗P < .005). (D) Spleen (left panel) and BM (right panel) B-cell Fr burden after daily administration of PBS (Ctrl) or 10 μg/dose IFN-γ to 4-week-old adult Eμ-ret mice for 7 days (Mann-Whitney test; n = 6 Ctrl and 7 IFN-γ; ∗P < .005). All data are presented as mean ± SEM. Nonsignificant comparisons in all graphs are not labeled.