Fig 1.
A Y. pestis Ymt mutant induces PIER in X. cheopis fleas when brown rat blood is used for the infectious blood meal.
A) Incidence of post-infection esophageal reflux (PIER) in groups of 25 to 220 X. cheopis fleas 24 h after feeding on brown rat (Rn), black rat (Rr), mouse (M), or human blood (H) containing 1.5 x 108–1.1 x 109 CFU/ml KIM6+ or KIM6+ymtH188N Y. pestis. Bars show the mean and standard error of 3 independent experiments (n = 164–438 mixed sex fleas). *p < 0.005 by chi-square test. B) Female X. cheopis with PIER 24 h after feeding on black rat blood containing GFP-positive Y. pestis KIM6+; blue arrow indicates where blood and Y. pestis has been refluxed from the proventriculus and/or midgut into the esophagus. C) light and D) fluorescence microscopy images of the digestive tract dissected from this flea showing the presence of partially digested blood components and bacteria in the proventriculus (PV) and esophagus (E). Scale bar = 50 μm.
Fig 2.
Ymt− and Ymt+ Y. pestis colonize female X. cheopis similarly when brown rat blood is used for the infectious blood meal, but not if mouse, human, or black rat blood are used.
Groups of female X. cheopis fleas that fed on mouse (blue), black rat (black), human (orange), or brown rat (red) blood containing 1.5x108–1.1x109 CFU/ml Y. pestis KIM6+, KIM6+ymtH188N, or KIM6+ymtH188N (pYmt) were scored for 1 week for A) the percentage of fleas that remained infected; B) the percentage that developed obstruction of the foregut (partial or complete blockage) that interfered with normal blood-feeding; and C) bacterial burden. Data are cumulative from 3 (KIM6+ and KIM6+ymtH188N groups) or 1 (KIM6+ymtH188N(pYmt) groups) independent experiments. Samples consisted of 7–20 female (A and C) or 25–220 fleas (roughly equal numbers of males and females; B) per experiment. The mean and standard error (A, B) or median (C) are indicated. *p < 0.05 by chi-square (A, B) or by Kruskal-Wallis test with Dunn’s post-test (mouse, human, and brown rat groups) or Mann-Whitney test (black rat group) (C). Dotted lines indicate the limit of detection (40 CFU). KIM6+ymtH188N(pYmt) was not used for black rat blood infections due to the limited availability of this blood.
Fig 3.
The Y. pestis Ymt mutant colonizes male fleas more efficiently than females following infection using mouse blood.
Infection rates for groups of female or male X. cheopis infected using mouse blood (blue symbols) or brown rat blood (red symbols) containing 1x108–5.7x108 CFU/ml GFP-positive KIM6+ or KIM6+ymtH188N Y. pestis were determined 1 day after infection by fluorescence microscopy of dissected flea digestive tracts A); or 0, 1, and 7 days after infection by CFU counts from individual triturated fleas (B, C). For A, each symbol represents the percentage of fleas containing GFP+ bacteria in their digestive tract. n = 4–10 fleas of each sex in 3 independent experiments (Table 1). For B and C, the mean and standard error (B) or median (C) of pooled data from 3 independent experiments for groups of 5–20 fleas infected using mouse blood are shown. *p < 0.05 by chi-square test (B) or two-way ANOVA with Tukey’s post-test (C). D) Examples of the foregut infection in female or male X. cheopis 1 day after ingesting KIM6+ymtH188N Y. pestis suspended in mouse blood (Left) or brown rat blood (Right). Scale bar = 50 μm.
Table 1.
Flea Dissection Summary.
Fig 4.
Survival of the Ymt mutant in the flea correlates with slower RBC digestion.
A) Bacterial titers and infection rates for groups of female X. cheopis infected using reconstituted, plasma-swapped mouse blood (brown rat plasma mixed with mouse RBCs; blue) or brown rat blood (mouse plasma mixed with rat RBCs; red) containing 1.3x108–2.8x108 CFU/ml KIM6+ymtH188N. Data are the pooled results from 3 independent experiments (n = 10); bars represent the median. *p < 0.0001 by Mann-Whitney test. B) Blood meal volumes of individual female or male X. cheopis allowed to feed for 1 h on a neonatal mouse. Mean blood meal volumes are indicated, *p <0.0001 by Student’s t-test. C) The RBC concentration in individual X. cheopis female or male digestive tracts 0.5 or 2 h after ingestion of sterile mouse or rat blood. Bars represent the mean of 3 independent assays using n = 3–6 (0.5 h) or n = 6–10 (2 h) fleas. *p <0.05 by two-way ANOVA with Tukey’s post-test. D) The mean proportion and range of male or female X. cheopis that completely liquified sterile mouse or brown rat blood during the first 8 h of digestion. Data are from groups of 3–6 digestive tracts excised from fleas at each timepoint and condition from 3 independent experiments; n = 9–15. *p <0.05 by Fisher’s exact test compared to rat blood group. A representative image series of these data is shown in S3 Fig.
Fig 5.
Rodent fleas can transmit Ymt mutant Y. pestis for at least 3 weeks when infected using brown rat blood.
Y. pestis transmission dynamics were monitored for 3 to 4 weeks for groups of 150–267 X. cheopis (A) or O. montana fleas infected using 3.4 x 108−1.9 x109 CFU/ml KIM6+ or KIM6+ymtH188N Y. pestis (B) in either mouse (blue) or brown rat (red) blood and subsequently maintained on sterile blood of the same type. Numbers in parentheses indicate the total number of fleas that fed followed by the number of fleas with evidence of foregut obstruction (partially or fully blocked). Roughly equivalent numbers of male and female fleas were used for transmission assays. Infection rate was determined for groups of 10–20 female C) X. cheopis or D) O. montana at various times following infection.
Fig 6.
Model of the adaptive function of Yersinia murine toxin (Ymt) during the evolution of the flea-borne life cycle of Y. pestis.
In this model, ancestral Y. pestis strains lacking the ymt gene (Ymt−; left) could cycle between fleas and certain species of rodent with flea-colonization-permissive host blood, such as brown rats (Rattus norvegicus), but not those with non-permissive blood, such as mice (Mus spp). Following acquisition of ymt on the pMT1 plasmid (Ymt+ strains; right), the progenitor of modern, extant strains of Y. pestis was able to stably colonize fleas that fed on bacteremic hosts with a blood chemistry that is not permissive for Ymt-negative strains. Thus, acquisition of ymt effectively greatly expanded the range of mammalian hosts that could support a flea-mammal transmission cycle. Although male fleas become infected at a moderate rate with Ymt− Y. pestis infected with non-permissive mouse blood (Fig 3), their potential to transmit is likely not sufficient to maintain a stable transmission cycle (Fig 5).
Table 2.
Strain and Plasmid List.