An aged host promotes the evolution of avirulent coxsackievirus into a virulent strain

Abstract

The emergence of new, more pathogenic viruses necessitates elucidation of factors that promote viral evolution. Aging, a potential factor, is associated with increased susceptibility to viral infections. We used the enterovirus coxsackievirus B3 (CVB3) to investigate the effects of host age on pathogenicity and viral gene sequence. Old mice infected with a normally amyocarditic strain of CVB3, CVB3/0, had significantly higher mean heart viral titers compared with CVB3/0-infected adult mice. To determine whether a change in the CVB3/0 viral population could contribute to the higher titers observed in the old infected mice, CVB3/0 was passed once through an old or adult host and the changes in pathogenicity and viral genome were examined after subsequent infection of old or adult mice. Adult mice infected with CVB3/0 that was passed through an old host (CVB3/0(Old)) exhibited significantly higher heart viral titers, pathology, and weight loss than adult mice infected with either stock CVB3/0 or CVB3/0 passed through an adult host (CVB3/0(Adult)). Sequence analysis of virus isolated from CVB3/0(Old)-infected mice revealed 13 specific and reproducible nucleotide changes. These changes result in a sequence that matches the virulent CVB3/20 strain and are associated with promoting cardiovirulence. In contrast, we observed only one nucleotide change, low heart viral titers, and no heart and liver pathology in adult mice infected with CVB3/0(Adult). These results demonstrate that the aged host promotes rapid selection of a pathogenic variant of CVB3 from an avirulent strain and introduces a host-virus paradigm for studies of viral infection in the aged.

Fig. 1.

Mice infected with CVB3/0^Old suffered greater morbidity and mortality than mice infected with CVB3/0. Adult and old C57BL/6 mice were infected with CVB3/0 and compared with adult and old mice infected with CVB3/0 isolated from the ground heart supernatant of previously infected old mice (CVB3/0^Old). (A) Survival curves after CVB3/0 infection. Shown are results for adult mice infected with CVB3/0 (□), old mice infected with CVB3/0 (▵), adult mice infected with CVB3/0^Old (■), and old mice infected with CVB3/0^Old (▴). Asterisks indicate significantly greater mortality than adult CVB3/0-infected mice (P < 0.05 by Kaplan–Meier survival plot analysis; n = 7 per group). (B) Heart CVB3 titers (logarithm of TCID₅₀ per gram of heart tissue) are expressed as mean ± SEM as previously described (54). Groups that do not share the same letter are significantly different at P < 0.05 by Kruskal–Wallis one-way ANOVA (n = 7 per group). (C) Heart pathology. (D) Liver pathology. Pathology data shown as the percentage of organ cross-section area affected ± SEM (n = 7 per group). Groups not sharing the same letter are significantly different at P < 0.05.

Fig. 2.

Representative photomicrographs of heart, liver, pancreas, and visceral adipose tissues from adult mice infected with CVB3/0 or CVB3/0^Old compared with those infected with media (control) or CVB3/20 (positive control). Five-microgram paraffin sections of heart, liver, and pancreas were stained with hematoxylin and eosin. Arrowheads indicate lymphocyte infiltration, arrows indicate viral inclusion bodies, and asterisks indicate intact islets of Langerhans. (Scale bars, 100 μm; Inset, 2 μm.)

Fig. 3.

Increased heart viral titer and pathology is specific to passage through an aged host. C57BL/6 mice were infected with CVB3/0 from various origins, stock CVB3/0, CVB3/0 isolated from adult mice (CVB3/0^Adult), and CVB3/0 isolated from old mice (CVB3/0^Old) as described in Materials and Methods. (A) Heart viral titer. Data shown are the mean logarithm of TCID₅₀ per gram of heart ± SEM. (B) Heart pathology. Data shown are mean percentages of heart area affected ± SEM. The asterisk indicates that the value is significantly higher than mice infected with CVB3/0 or CVB3/0^Adult at P < 0.001 and P < 0.05 for heart viral titer and pathology, respectively (Kruskal–Wallis one-way ANOVA; n = 7 per group for CVB3/0 and CVB3/0^Old, and n = 20 per group for CVB3/0^Adult).