Abstract
Yersinia pestis, the causative agent of plague, is endemic in certain regions due to a stable transmission cycle between rodents and their associated fleas. In addition, fleas are believed to serve as reservoirs that can occasionally cause enzootic plague cycles and explosive epizootic outbreaks that increase human exposure. However, transmission by fleas is inefficient and associated with a shortened lifespan of the flea and rodent hosts, indicating that there remain significant gaps in our understanding of the vector-animal cycle of Y. pestis. Here, we show that laboratory-reared, infected fleas (Xenopsylla cheopis) can transmit viable Y. pestis from adults to eggs, and the bacteria can be passed through all subsequent life stages of the flea. Thus, our data raise the possibility that transovarial transmission in fleas might contribute to the persistence of Y. pestis in the environment without detectable plague activity in mammals.
Similar content being viewed by others
Introduction
Yersinia pestis is the causative agent of plague, which is classified as a prioritized re-emerging zoonotic disease transmitted by fleas. Worldwide plague foci persist due to enzootic sylvatic cycles consisting of burrowing rodents and their respective flea vectors1,2,3,4,5,6,7. Numerous flea species and primarily rodent hosts are implicated in maintaining the sylvatic plague cycle, with mice, rats, prairie dogs, and ground squirrels commonly found in established plague foci8,9. Fleas can transmit Y. pestis as early as 1 day post-infection in a regurgitative mechanism thought to be triggered by the early occlusion of the esophagus due to bacteria (referred to post-infection esophageal reflux (PIER)), a process that may be enhanced by high flea density10,11,12. Once in the midgut, bacteria respond to environmental cues that activate the expression and production of an exopolysaccharide matrix, which develops into an infectious biofilm13,14. The biofilm can become localized to the proventriculus and, during feeding, occludes ingestion which facilitates regurgitation and deposition of Y. pestis in the dermis of the human or animal host. Whether biofilm-independent or biofilm-dependent regurgitative transmission, as few as one bacterium is sufficient to cause lethal bubonic plague in an animal host15. Though highly efficient, biofilm-induced regurgitative transmission causes flea mortality due to an inability to ingest bloodmeal. Therefore, Y. pestis infection can be lethal to its vertebrate and invertebrate hosts, yet it is highly successful in surviving within the enzootic cycle with more than 30 species of fleas as likely vectors for nearly 300 susceptible animal hosts16.
Early studies of the rodent-flea plague cycle documented proventricular blockage and its potential correlation to transmission in fleas known to vector Y. pestis. Blockage and transmission by blocked fleas was most frequent in the rat flea, Xenopsylla cheopis, with more than 50% occurrence, whereas blockage appeared considerably less frequent in other flea vectors such as Oropsylla montana17. This model has since been refined upon finding early colonization of the foregut prior to biofilm formation in both flea species that appears to represent a transmissible form18,19. Further work established that the occurrence of proventricular blockage in Y. pestis-infected O. montana could be enhanced by using rat blood feeding compared to mouse blood indicating contributions from host bloodmeal to transmission11. Bacterial factors, including the plasmid-encoded Yersinia murine toxin (Ymt), are thought to act on the bloodmeal to protect from cytotoxicity and facilitate transmission20. Recent work, however, has established that Ymt is less important during transmission from Brown Norway rats than from other host species, suggesting there could be mechanistic differences in regurgitative transmission that depend on blood components enriched in specific animal hosts21. These and other observations lead to a model wherein enzootic maintenance and epizootic spread depend heavily on the presence of specific reservoir species, yet there is no evidence that the abundance of susceptible hosts correlates with the cycling of plague in the environment22.
While it is well documented that Y. pestis colonizes the foregut and midgut, there have been no reports of Y. pestis localization to salivary glands, and no indication that bacteria are able to escape the midgut of fleas. Thus, with the prevailing model that Y. pestis is maintained by transmission between adult fleas and susceptible rodent hosts facilitated by terminal infection of each host, the persistence of Y. pestis in an environment where there is little to no plague activity cannot be modeled. In this work, we addressed the historic paradigm that Y. pestis is confined to the midgut and alimentary canal in fleas with the underlying hypothesis that vertical transmission in the flea could account for periods of reduced plague activity in a given endemic area. Using high-resolution imaging with transmission electron microscopy, we reexamined the distribution of Y. pestis following membrane feeding of X. cheopis in male and female fleas and their progeny.
Results
Yersinia pestis escapes to reproductive tissue in the flea vector X. cheopis
To understand Y. pestis interactions in fleas, we used a plasmid system for the constitutive high-level expression of a fluorescent reporter cysZP-tdTomato as a means to localize bacteria following infection of X. cheopis in an artificial membrane feeder23. Previous work suggested that maintenance of Ctenocephalides felis on an artificial membrane feeder using pig blood led to a high degree of successful rearing24. We, therefore, initiated this study by comparing the kinetics of infection in the early stage following infection in different host species blood of X. cheopis in an artificial membrane feeder. To establish the impact of these different blood sources on early colonization, we dissected the gastrointestinal tract of the infected fleas and imaged for Y. pestis tdTomato by confocal microscopy. Bacterial density in the proventriculus, as well as esophagus was determined by quantifying the fluorescent signal, reported as relative fluorescent units (RFU) in the preparation. Higher fluorescence is suggestive of aggregation and/or growth of Y. pestis in the foregut, whereas lower fluorescence suggests either no detectable bacteria in the proventriculus or dispersed growth. Consistent with published data that demonstrated Y. pestis colonization and blockage of the proventriculus of X. cheopis on days 1 and 3 post-infection, respectively, we found that infection of Y. pestis in rat blood resulted in variation in the amount of Y. pestis localized to the proventriculus and esophagus which varied by time post-infection as well as host bloodmeal source and ranged from negative, minimal, mild, moderate or high (Fig. 1A–E)11,21. Confocal 3D imaging illustrated that Y. pestis tdTomato was present in the interior of the proventriculus (Fig. 1F, G). Combining data from all time points showed that 97% of the fleas infected in rat blood harbored detectable Y. pestis fluorescence in the proventriculus, with 22% of the fleas showing high levels of fluorescent bacteria in the foregut and esophagus, where it is presumably transmissible (Fig. 1H). By comparison, infection in mouse blood yielded an overall frequency of proventricular localization of 73%, with 14% in the moderate range and none with high fluorescence. Prairie dog blood was the least efficient, with only a few showing moderate or high fluorescence, 62% with mild proventricular fluorescence, and 18% negative for RFU. In contrast, pig blood was most efficient in yielding detectable Y. pestis fluorescence in the proventriculus, with 37% in the moderate to high-intensity range. These data suggest that pig blood, like rat blood, supports early colonization and aggregation of Y. pestis in the foregut, whereas mouse and prairie dog blood are less efficient.
Groups of 10–50 X. cheopis fleas were infected with Y. pestis ptdTomato in an artificial membrane feeder using rat (blue), mouse (green), prairie dog (Pdog, orange), or pig (red) blood and housed for 14 days with sampling on days 1, 3, 4, 5, 7, 10, and 14. Midguts were dissected, group identities were blinded and processed for confocal fluorescent microscopy. A–E Representative images illustrating quantification: no localization of tomato-fluorescing signal to the proventricular spines (<1 RFU, A), minimal (1–10 RFU, B), mild (10–100 RFU, C), moderate (100–200 RFU, D), or high (>200 RFU, E). A2–E2 Merged brightfield and fluorescent channels shown on top; grayscale conversion used for quantification of RFU shown on bottom row. F, G z-stacks of the proventriculus for two highly fluorescing samples fed on rat and pig blood, respectively, showing fluorescence within the tissue. H Mean proventricular RFU with standard deviation for samples collected from all time points; individual data points for each sample are shown, line within each bar indicates the median. Moderate and high RFU samples are shown above the line, with the percent of fleas exhibiting strong proventricular fluorescence indicated at the top of the graph; lower line indicates minimal and negative signal; the most commonly observed fluorescence was in the mild range. Data were collected in three independent trials with n = 59 (rat), n = 58 (mouse), n = 54 (prairie dog), n = 59 (pig). Data were pooled for statistical evaluation by Kruskal–Wallis followed by Dunn’s multiple comparison’s test, ***P < 0.001 (Mouse:Rat P = 0.000175, Mouse:Pig P = 0.000595), ****P < 0.0001 (Prairie dog:Pig P = 8.2046 e −7, Prairie dog:Rat P = 0.000001).
We next established the behavior of fluorescent Y. pestis in the midgut using confocal microscopy during the first 7 days following infection in rat blood. In individual fleas, RFU values appeared the highest on day 3 post-infection, with bacteria visible in the esophagus in some samples, and median RFU significantly higher than observed on day 1 (Fig. 2A–C, E). On day 7 post-infection, few to no midgut samples showed moderate or high proventricular fluorescence, consistent with an expected drop in efficiency of early phase transmission after day 4 post-infection (Fig. 2D, E)10. In parallel, bacterial load was quantified in pools of 3 midguts per group. High bacterial titer was recovered on day 3 post-infection with a median value of approximately 1 × 108 CFU/3 midguts, whereas the median bacterial titer appeared lowest on day 7, with less than 1 × 106 CFU/3 midgut pool (Fig. 2F). Although there appeared to be little variation within the day 3 group or day 7 group, the day 1 group showed some variation, and the overall P values between all groups suggested a trending difference but was not significant. This suggests a temporal aspect of bacterial aggregation that may not precisely correlate with actual bacterial numbers in the flea. Alternatively, the procedure of pooling three midguts per sample likely reduced the sensitivity of this assay, resulting in a rather inconsistent correlation between bacterial titer and RFU. Overall, the data suggest that bacterial colonization and aggregation in the foregut peaks around day 3 post-infection.
Groups of 10–50 X. cheopis fleas were infected with Y. pestis in rat blood. On the indicated day post-infection, fleas were removed, stored at −80 °C until collection of all samples, then group identities were blinded, fleas were dissected and processed for confocal microscopy. A–D Representative confocal microscope images of dissected midguts from control adult fleas (A) or (B–D) infected, harvested 1(B), 3 (C), or 7 (D) days post-infection (DPI); inset shows HSB image; scale bar indicates 100 µm. Each image is numbered to show the tissues quantified (overall RFU): esophagus (1), proventriculus (2), and midgut (3). E Mean RFU (midgut and proventriculus) over time; red line shows background RFU; box plots represent 25–75% interquartile range, the line indicates the median value, whiskers indicate minimum and maximum values; individual data points are shown on the right, n = 12 (day 1 post-infection (D1)), n = 11 (D3), n = 10 (D7) fleas. F Bacterial load (colony forming unit, CFU) in the dissected midguts determined in pools of three midguts per data point, n = 27 fleas (D1, D3), n = 15 (D7). Data shown were collected in three independent trials. E, F Data were evaluated with the Kruskal–Wallis test followed by Dunn’s multiple comparisons test.
To determine whether Y. pestis may be localized outside of the midgut, we examined the ovarioles and testes of fleas dissected from samples that were collected following Y. pestis infection using any of the four host species bloodmeal (mouse, rat, pig, or prairie dog). Under each of these infection conditions, using confocal microscopy and 3D imaging of cross-sections, we readily observed tdTomato-expressing Y. pestis, appearing to localize inside the ovarioles of female fleas rather than on the outside surface (Fig. 3A–C, an example shown is from tissue harvested from fleas infected with prairie dog blood on day 7 post-infection). Similarly, we found fluorescent bacteria in the spermatheca, with rod-shaped bacteria apparent in the interior of the organ (Fig. 3C–F, an example shown was collected following Y. pestis infection in pig blood on day 5 post-infection). Ovarioles and spermatheca dissected from female fleas that were not infected had no visible red fluorescence (data not shown). In the males, Y. pestis appeared to localize to the periphery of the testes, where there was often bright fluorescence in infected fleas but not in control fleas (Fig. 3G–I, an example shown was dissected from X. cheopis infected with Y. pestis in rat blood on day 5 post-infection). Other tissues that contained Y. pestis that was visible by confocal microscopy included Malpighian tubules, and in one case, we found Y. pestis in Malpighian tubules as well as within biomass in the midgut (Supplemental Fig. S1). Malpighian tubule localization appeared regardless of host blood source used for infection, often days after blood feeding occurred. Furthermore, Y. pestis could be seen in the testes, epididymis, aedeagus, and rectal papillae of a male flea 15 weeks after infection (mouse blood) even when bacteria were not abundant in the midgut (Supplemental Fig. S2). Furthermore, we did not observe bacteria in the salivary glands or in midgut epithelial cells in any of our samples. These results suggest that Y. pestis infection of the reproductive tissue is stable throughout the lifespan of the infected adult flea. In addition, unlike the aggregation of Y. pestis in the foregut and midgut which varied significantly depending on the host blood source, we observed Y. pestis in reproductive tissues whether we used rat, mouse, prairie dog, or pig blood with no obvious differences in frequency between them.
Representative confocal fluorescence microscopy images from individual adult fleas that were infected in an artificial membrane feeder with Y. pestis KIM6 + ptdTomato shown in Figs. 1, 2; host blood source is indicated in parenthesis. A–C Ovarioles dissected on day 7 post-infection (prairie dog blood): A Brightfield merged with the fluorescent channel; B, C 3D rendition of (A) with cross section (C). D–F Spermatheca dissected on day 5 post-infection (pig blood): D Brightfield merged with the fluorescent channel; E, F 3D rendition of (D) with cross section (F) to show bacteria within the tissue. G–I Testes dissected on day 5 post-infection (rat blood): G Brightfield merged with fluorescent channel; H 3D rendition of (G) with cross section (H) to show the bacteria within the tissue. Scale bars indicate 100 µm; yellow boxes show sections used in the 3D renderings. Images were selected from ten control fleas that were not infected; ten female adult fleas (A–C, five of ten were positive), nine female adult fleas (D–F, five of nine were positive, one spermatheca sample was lost during processing), and nine male adult fleas (G–I, five of nine were positive); samples were collected from three independent infections.
Y. pestis is present in the oviposited eggs derived from infected adult fleas
Adult fleas that feed regularly are expected to produce four to six eggs per day that are ~500 µm long and 300 µm wide and off-white in color25,26. Using 500 µm mesh, we sifted the bedding of infected and control fleas to collect eggs every 1–3 days for 3 weeks after infection. Eggs (n = 14) from infected fleas and control (n = 4) fleas fed sterile blood were collected and appeared of the expected size and color. The eggs were extensively washed as described in the Methods to remove any potential contamination from the surface, and then processed for confocal microscopy. Eggs from control fleas exhibited minimal autofluorescence either on the surface or within the egg tissue (Fig. 4A–C). In sharp contrast, a low level of tdTomato fluorescence was present in the eggs produced by Y. pestis-infected fleas (Fig. 4D, G). Using confocal microscopy and 3D cross-sectioning of images, it was clear that the fluorescent signal was inside the egg and not on the surface (Fig. 4E, F, H, I). We found a prominent fluorescent signal in approximately half of the eggs that were tested (examples shown in Fig. 4D–F collected on day 3, and samples in Fig. 4G–I were collected on day 7 post-infection, summarized in Supplemental Table S1). Considering blood intake initiates copulation and subsequent oviposition, the data suggest that egg-derived Y. pestis may be transferred to the egg during oviposition due to the colonization of the reproductive tissue27.
Eggs were collected from adult X. cheopis that were fed uninfected (A–C) or Y. pestis-infected (D–I) rat blood, then processed for confocal microscopy: (A, D, G) Brightfield merged with fluorescent channel; (B, E, H) 3D rendition of (A, D, G) with cross section (C, F, I); yellow box indicates area where 3D cross section was taken; (D–F) representative sample collected on day 3 post-infection; (G–I) representative sample collected on day 7 post-infection. Scale bars indicate 100 µm; images are representative of n = 4 control and n = 14 infected eggs collected from four independent infections.
In order to demonstrate that the tdTomato fluorescent signal that was observed in the reproductive tissues was a result of bacterial presence and not an artifact of the reporter system, we repeated the study using non-recombinant Y. pestis KIM6+ infection of X. cheopis and immunohistochemistry to visualize bacterial localization. On days 1 and 3 post-infection with rat blood, fleas were processed by dissecting midguts and the associated reproductive tissue, then fixing with paraformaldehyde, followed by immunostaining with anti-Yersinia antibody 6B5. By plating, we recovered a median titer of 1 × 107 CFU/3 midguts on days 1 or 3 post-infection, with an apparent decrease in bacterial load on day 7 (Fig. 5A). Therefore, similar growth in the midgut is observed between Y. pestis KIM6+ and the tdTomato-expressing strain suggesting no apparent artifact was caused by the tdTomato plasmid that impacted midgut bacteria. When we conducted immunostaining of eggs, we observed no fluorescence in eggs collected from fleas that fed from sterile blood (Fig. 5B, C). In contrast, eggs from infected adults were stained by anti-Yersinia antibody (Fig. 5D, E). Similarly, when we extracted and stained ovariole tissue from infected adult fleas, we readily identified rod-shaped Y. pestis in the interior of the oocyte (Fig. 5F–H). These bacteria measured approximately 1 µm in length, which is within the expected size range (1–3 µm long × 0.5–0.8 µm diameter)28. Overall, the immunostaining unequivocally confirms that Y. pestis is present in the ovarioles and eggs of the infected adult fleas.
Groups of 25–50 fleas were infected with 1 × 108 CFU/mL Y. pestis KIM6+ in rat blood in an artificial membrane feeder; control fleas were given rat blood with no Y. pestis. A Bacterial load in the midguts on days 1, 3, and 7 days post-infection; bars indicate median, pools of three midguts were combined for each time point (n = 9 fleas for days 1, 3; n = 6 fleas for day 7; no statistical analysis was conducted); the remaining fleas were housed for egg collection. B–E Eggs were collected from control (B, C) or infected (D, E) fleas, sterilized with ethanol, and then fixed and stained with anti-Yersinia monoclonal antibody (MAb 6B5). F–H Representative images from ovarioles that were dissected from fleas on day 3 post-infection, then sterilized with ethanol prior to immunohistochemistry staining with MAb 6B5. B, D, F representative 3D rendition showing brightfield merged with a fluorescent channel, C, E, G fluorescent channel only, H cross section of (F); white and green arrows point to bacteria in the tissues with the green arrows highlighting bacteria that are not associated with fat body-like cells, the yellow box indicates a region where the 3D cross section was taken; scale bars indicate 10 µm. Data were collected in five independent trials (for eggs: n = 5 infected, n = 4 uninfected controls per trial; for ovarioles, n = 5 infected, n = 1 control per trial).
Transmission electron microscopy (TEM) shows Y. pestis localizes to oocytes and seminiferous tubules in adult fleas
We used transmission electron microscopy to image tissues that were dissected from male and female fleas. In the ovarioles of uninfected control female fleas, infrequent small bacteria that were less than 0.5 µm diameter were found, but there were no bacteria in follicle cells of the oocytes (Fig. 6A). In striking contrast, in the Y. pestis-infected female fleas, we observed frequent gram-negative bacteria in vacuoles of follicle cells; the bacteria were ~1 µm in length and rod-shaped, with the inner and outer membranes easily visualized (Fig. 6B, C). The Yersinia-containing vacuoles (YCVs) were sometimes spacious containing one to three bacteria, while other YCVs contained little intravacuolar space and one or two bacteria (Fig. 6D–F). In the male control fleas that were not infected, we found normal seminiferous tubules with a narrow basement membrane, and few to no bacteria in the tissue (Fig. 6G). However, in the tubules of infected male testes, there appeared to be an enlargement of the basement membrane and relatively abundant bacteria within this enlarged compartment (Fig. 6H–L). The bacteria were rod-shaped and ~1 µm in length, and all were contained within the basement membrane enclosure. These data suggest that Y. pestis colonizes the periphery of the seminiferous tubules rather than the spermatocytes, consistent with the fluorescence pattern we observed by confocal microscopy. Overall, the TEM confirms that Y. pestis exit the midgut and localize to the reproductive organs of an adult flea. Furthermore, it appears that oocytes may be infected with primarily intracellular, intravacuolar Y. pestis, whereas the accessory tissue surrounding the spermatocytes is colonized by extracellular Y. pestis in males.
Ovarioles and testes were dissected from groups of 5–10 X. cheopis on day 1 or 3 post-infection with Y. pestis KIM6+ in rat blood (B–F, H–L); control fleas fed sterile rat blood (A, G). Ovarioles from female control (A) or infected (B–F) fleas. Testes from male control (G) or infected (H–L) adult fleas showing numerous double-membraned bacteria within an enlarged basement membrane (white arrows); red arrows indicate section where zoomed-in images were taken. Labels in the images are as follows: FC follicle cell, OC oocyte, BM basement membrane, ST seminiferous tubule, SC spermatocyte. Scale is indicated in each panel; data were collected in two independent trials (samples from 15 fleas in each group were analyzed by TEM).
Yersinia pestis is transmitted transstadially throughout each life stage of infected X. cheopis
To determine whether Y. pestis could be passed between life stages, we also tested larvae collected from sifting through 500 µm mesh following Y. pestis tdTomato infection of adult fleas. Larvae from fleas that were not infected contained minimal autofluorescence (Fig. 7A–C). In contrast, we found five of eight larvae from infected fleas carried tdTomato fluorescence that appeared multi-focal and diffuse in the larvae (Fig. 7D–F). To reduce the risk of contaminating the larvae with fecal material that may contain Y. pestis, additional eggs from Y. pestis-infected adult fleas were collected, placed into sterile sawdust with sterile larvae food where they developed into larvae, pupae and F1 adults. Like those reared directly in the egg chamber, these larvae harbored red fluorescent bacteria that appeared multi-focal and was present in the interior of the tissue (Fig. 7G–I). Pupae reared from infected or control fleas that were not infected exhibited a high degree of autofluorescence (Supplemental Fig. S3A, B). Nevertheless, strong fluorescence could be detected in focal areas of the pupae reared from Y. pestis-infected eggs, consistent with Y. pestis localization (Supplemental Fig. S3C, D). These data suggest that Y. pestis survives metamorphosis and transstadial transmission.
Larvae were collected from X. cheopis adult fleas fed in an artificial membrane feeder from uninfected (A–C) or Y. pestis-infected (D–I) rat blood, then processed for confocal microscopy. Representative images from individual larvae are shown: (A, D, G) Merged brightfield with the fluorescent channel; (B, E, H) Fluorescent channel (C, F, I) 3D cross section. Infected larvae were either collected from the same housing as the infected adults (D–F) or after separating the eggs from infected adults and hatching in fresh, sterile housing (G–I). Images are representative of eight infected larvae and three control larvae collected from three independent infections; scale bars represent 100 µm; yellow boxes outline the area where the 3D cross section was taken.
To determine whether transovarially and transstadially transmitted bacteria were viable, we plated a subset of each life stage on agar plates to enumerate bacterial titer in the eggs, larvae, pupae, and F1 adults. The median number of colonies recovered from eggs (n = 15 tested) was 5 CFU, with approximately half of the eggs testing negative for bacterial growth (Fig. 8A). In the later stages of development, bacteria appeared to replicate, with the median titer recovered in larvae of ~50 CFU, and in pupae, the median titer was over 100 CFU. In the newly emerged adult F1 fleas collected prior to their first blood feeding, midguts were harvested for plating, resulting in a median titer of 173 CFU recovered. We used fluorescence microscopy to confirm Y. pestis in the midgut of the newly emerged F1 adult flea, collected prior to its first feeding. Indeed, tomato fluorescence was readily observed in the midgut, consistent with Y. pestis survival and localization to the digestive tract (Fig. 8B, C). These data suggest that in the F1 adult flea, Y. pestis localizes to the midgut, where it seems probable that it could become transmissible to a mammalian host.
A–C Eggs (n = 15, circles), larvae (n = 8, squares), pupae (n = 4, triangles), and F1 adults (n = 3, diamonds) were collected from Y. pestis-infected X. cheopis fleas and either processed for CFU determination (A) or imaging (B, C, F1 adults only). A Bacterial load: each data point represents a single individual, except in F1 adults (two were pooled then plated and one was individual), samples with no detectable bacteria are not shown. Data were evaluated for statistical significance using the Kruskal–Wallis test followed by Dunn’s multiple comparison’s test, *P < 0.05 (P = 0.0108 between eggs and pupae). B, C Confocal microscopy showing merged brightfield (B) or fluorescent channel (C); scale bar indicates 100 µm. D, E Y. pestis isolated from eggs was used to infect naïve X. cheopis fleas; control fleas were infected with the parent strain KIM6 + ptdTomato. On D3 and D7 post-infection, midguts were extracted and either processed for bacterial load determination in pools of three midguts (D) or confocal microscopy (E); data shown were collected in a blinded fashion in two independent trials (n = 9 fleas for D3, n = 6 fleas for D7, no statistical evaluation was performed). F PCR for six transovarially transmitted isolates (three from eggs, one each from larva, pupa, and F1 adult) for strain-specific Y. pestis plasmids pMT1 (caf) and pPCP1 (pla); positive control DNA was prepared from Y.pestis KIM6+, negative control had no template DNA; 1 kb molecular weight (MW) marker is shown in lane 1. G–I Sequenced genome map of flea egg-isolated aligned with the Y. pestis KIM6+ (pgm + pMT1+, pPCP1+, pCD1-, ptdTomato) parent strain used in the infection: chromosome (G), pMT1 (H), or pPCP1 (I).
We evaluated Y. pestis isolated from the different life stages to query the genome sequence and the potential for the transovarially transmitted Y. pestis to establish an infection in a naïve adult flea. Fleas were challenged with egg-derived Y. pestis in the artificial membrane feeder. On days 3 and 7 post-infection, groups of ten fleas were removed from housing, placed in a membrane feeder where they were provided access to sterile rat blood, and after 1 h, blood was collected to quantify bacterial transmission. Control fleas were infected with the parent KIM6+ strain and processed in parallel. On day 3 or 7 post-infection, both KIM6+ and egg-derived Y. pestis were transmitted to a similar efficiency on day 3 or 7 post-infection (Fig. 8D). By microscopy, we observed localization of Y. pestis biomass to the midgut and proventriculus whether it was egg-derived bacteria or the parent strain (Fig. 8E). To confirm the retention of extrachromosomal plasmids in all of the flea life stages, bacterial stocks were made from bacteria harvested from eggs, larvae, pupae, and F1 adults, and then used for PCR to amplify the Y. pestis plasmid-specific genes caf1 and pla. Following analysis of the PCR products on agarose gel electrophoresis, we observed the expected size products from each life stage (Fig. 8F). The DNA was excised from the gel, and the fragment was sequenced, confirming the expected sequence of Y. pestis caf1 and pla in bacteria harvested from eggs, larvae, pupae, and the F1 adult flea.
To unequivocally identify Y. pestis in progeny eggs from infected adults, we subjected bacteria isolated from eggs to whole genome sequencing. In the first round of sequence data without gap filling, we identified more than 67% of Y. pestis KIM6+ chromosomal genes, 58% of pMT1, and 99% of pPCP1 (Fig. 8G–I). This sequence data provides strong evidence that the bacteria recovered from eggs was intact Y. pestis. Overall, these data suggest that transovarially transmitted Y. pestis could be competent for re-entry into the sylvatic plague cycle, although we have not yet established this experimentally.
Discussion
In the more than 100 years of Y. pestis research, escape from the alimentary tract has never been reported. In this work, we have provided multiple layers of evidence that Y. pestis does indeed escape from the midgut, colonizes reproductive tissue, and is able to undergo transovarial transmission in X. cheopis. These findings provide visual, quantitative, and genetic support that low levels of Y. pestis escape from the midgut throughout the lifespan of an infected flea, entering the reproductive tract where they are able to colonize progeny eggs. Early studies, as well as recent reports, have established a strong likelihood of horizontal transmission through a regurgitative mechanism without colonization of salivary glands, and in agreement with these results, we observed a high degree of foregut colonization without detectable localization of Y. pestis in the salivary glands14,19,29. The amount of colony forming units harvested from flea eggs is small, averaging fewer than 10 CFU. Consequently, it may simply be that previous experiments were not sensitive enough to capture the low level of transovarial transmission. Our tdTomato reporter is expressed on a high copy number plasmid under the control of the constitutively active Y. pestis cysZ promoter and can be detected in a wavelength range where autofluorescence is low. These features likely enabled the visualization of bacteria in the reproductive tissues as well as accessory organs, including the Malpighian tubules, indicative of systemic dissemination of the infection.
Early researchers investigated whether Y. pestis (then Pasteurella pestis) could be maintained independently of the mammalian host. Through feeding Y. pestis-contaminated blood to larvae, it was apparent that Y. pestis survived poorly in the digestive system of larvae, thereby establishing that ingestion of Y. pestis from contaminated fecal material or dead fleas would likely clear the infection30. In the data shown here, we provide strong evidence that vertically transmitted Y. pestis is not cleared by larvae, suggesting the infection may be extra-intestinal at this developmental stage.
Fleas are vectors of other diseases, namely Rickettsiosis and cat scratch fever caused by Rickettsia felis or typhi or Bartonella henselae, respectively. Transovarial transmission of Rickettsia felis and typhi has been characterized and is thought to contribute significantly to the maintenance of these pathogens in the environment31. Rickettsia typhi localizes to the cytoplasm of midgut epithelial cells and was also identified in the ovaries and hypodermal tissue of the testes32. Rickettsia felis, also thought to be transmitted to eggs, was shown to invade the hemocoel and spread systemically with abundant intracellular bacteria observed in salivary glands and oocytes33. In the previous study, R. felis infection of the midgut epithelial cells was readily apparent shortly after infection. However, in contrast, we and others have not found intracellular Y. pestis in midgut epithelial cells. Although we did not detect Y. pestis in the midgut epithelium, and there are no reports of Y. pestis in hemocoel, it is premature to rule out intracellular invasion as a likely mechanism of dissemination. For example, the blood feeding process for infection introduces an opportunity for bacterial interaction with immune cells, whether the fat body cells or those residing in the hemocoel. Since Y. pestis is known to express multiple adherence and invasion factors in flea and mammalian environments such as Ail, OmpA, Ilp, and others, this interaction might be predicted to result in dissemination to reproductive tissue34,35,36. Furthermore, intracellular Y. pestis express virulence factors capable of promoting survival, replication and even killing of their host cells which may collectively contribute to the transovarial transmission process37,38,39,40. In support of this mechanism, there did appear to be fat bodies in the ovariole tissue preparation that stained positive for Y. pestis by immunohistochemistry in addition to the follicle cells and oocytes.
Furthermore, the TEM data suggested that Y. pestis occupies an intracellular niche in the oocytes and follicle cells where multiple bacteria could be observed in spacious vacuoles. This indicates a strong probability that Y. pestis may be able to survive and replicate intracellularly in a vacuolar compartment in non-immune cells of the reproductive tissue. In contrast, during R. felis infection of Ctenocephalides felis, numerous bacteria were observed in oocytes, mainly in tight vacuoles or within the cytoplasm by TEM. Furthermore, in male X. cheopis fleas, we found abundant extracellular Y. pestis, whereas the obligate intracellular lifecycle of Rickettsia precludes its extracellular growth or survival. These observations indicate distinct host–pathogen interactions in male and female fleas. Alternatively, male reproductive tissue may be inoculated by infected females during mating or vice versa. Future studies will focus on experimentally validating the mechanism of midgut escape, the importance of intracellular bacterial survival, and the process by which the infection proceeds during oviposition.
It is intriguing to speculate that transovarial transmission could be a contributor to the enzootic maintenance of plague. We showed that viable Y. pestis could be found in reproductive tissue anywhere from 1 day to 15 weeks post-infection, suggesting that transovarial transmission of Y. pestis may occur during the early stage of infection and persist over the lifespan of the infected flea resulting in maintenance and even spread of infection in the flea population. Furthermore, it appeared that vertical transmission in X. cheopis was not substantially influenced by the host species' blood source. Combined with the known requirement for biofilm in supporting bacterial survival and regurgitative transmission, these observations suggest a model wherein early midgut escape results in stable colonization of reproductive tissues that supports the continual infection of progeny eggs12. However, there are many questions that remain before it will be possible to determine the impact of vertical transmission on the zoonotic plague cycle and the degree to which this mechanism can explain periods of apparent quiescence in plague-endemic areas when the ecological conditions are unable to support the horizontal transmission cycle. From the data collected to date, we estimate transovarial transmission occurs in around 50% of the infected fleas, with transstadial transmission even more efficient (summarized in Supplemental Table 1). However, the efficiency with which Y. pestis develops into a regurgitative transmission state in the F1 adult fleas is unknown. Vertical transmission of pathogens in their arthropod hosts is typically associated with gene loss and attenuation, at least in some instances41. Therefore there may be evolutionary pressure that destabilizes the transovarially transmitted Y. pestis from playing a role in the enzootic plague cycle. Nevertheless, with localization in the midgut of F1 adults, it seems probable that blood feeding could generate the necessary signals and products for bacterial growth, colonization of the proventriculus, and biofilm-mediated transmission. Furthermore, environmental factors such as precipitation and temperature may be predicted to have an impact on the transovarial transmission cycle as they would affect fecundity and feeding, thereby impacting the abundance of infected fleas, which would likely enhance the probability that the vertically transmitted, virulent Y. pestis reaches a mammalian host. By understanding vertical transmission, we may improve our ability to understand plague ecology and the cycling of outbreaks to better predict the risk of human exposure to Y. pestis.
Methods
Animals
Adult male and female C57BL/6J mice (Mus musculus) were originally purchased from the Jackson Laboratory (Maine, USA), then were bred and reared at the University of Missouri under conditions of 12-hour light/12-h dark cycle, 21–23 °C, 30–70% humidity, with mouse chow and water provided ad libitum. Male and female neonatal mice (less than 7 days old) were used for colony maintenance blood feeding; adult male or female mice (>12 weeks old) were used as the source for the artificial membrane feeder as well as a blood source for experiments where indicated42. Other blood sources for infection in the artificial feeder were as follows: rat (Rattus norvegicus, BioChemed Services, Vermont, USA), prairie dog (Cynomys ludovicianus, PMS Recycled Vermin, Texas, USA), and pig (Ossabaw swine, Corvus Biomedical, Indiana, USA). After infection in the indicated host species blood, fleas were provided a bloodmeal every 7 days using the same host blood source through the membrane feeder. All procedures involving animals were approved by the University of Missouri Animal Care and Use Committee.
Flea infections
Fleas (X. cheopis) were originally obtained from the NIH/NIAID Rocky Mountain Laboratory, Montana, USA, and were reared at the University of Missouri in medium comprised of one part food mix (comprised of equal parts powdered bovine blood, dry milk, and ground mouse chow): three parts sawdust under conditions of 22 °C, 60–75% humidity on a 24 h dark cycle; colony feeding was conducted two to three times per week using neonatal mice42. Flea infections were conducted with a laboratory strain of Yersinia pestis KIM6+, which lacks the type III secretion system plasmid pCD1 but carries the other two extrachromosomal plasmids, pMT1 and pPCP143,44,45. The recombinant derivatives of KIM6+ used in this work did not reintroduce the pCD1 plasmid, and therefore are classified as select agent-exempt strains by the US Center for Disease Control and Prevention. All usage of Y. pestis KIM6+ was approved by the University of Missouri Institutional Biosafety Committee.
Prior to use in infection, laboratory-reared, naive X. cheopis were separated from the colony and were not fed for 5–7 days to improve feeding efficiency during infection. Groups of 50 fleas were infected with Y. pestis strain KIM6+ carrying pNE160, which encodes the fluorescent protein tdTomato (Excitation: 554, Emission 581)23. An artificial membrane feeder was constructed using skin from an adult mouse42,46. Blood was inoculated with 5 × 108 to 1 × 109 Y. pestis and maintained at 37 °C, and fleas were allowed to feed for 1 h. When fleas were removed from the feeder, they were observed to determine the intake of the bloodmeal. Fleas that had not fed or were otherwise lost in the processing of samples were removed from the study.
Midgut processing
Fleas were euthanized at the indicated time point by placing in −80 °C, then midguts and other tissues were dissected and placed onto sterile slides, fixed with 4% paraformaldehyde, and washed three times with 1X PBS (2.7 mM KCl, 1.5 mM KH2PO4, 138 mM NaCl, 8 mM Na2HPO4, pH 7–7.3) for a minimum of 30 min each time prior to mounting.
Egg collection
Infected fleas were maintained in modified housing with 300-micron mesh on the lid, 550-micron mesh under the bedding, and a lower chamber that could be easily removed. Eggs and larvae were collected by sifting with care to avoid contact with infected feces and fleas. Sifting for eggs and larvae occurred every 1–3 days post-infection. To aseptically collect the eggs and larvae, a sterilized cotton applicator was moistened with double-distilled, sterile H2O and used to pick up eggs or larvae. These specimens were placed onto a sterile slide, washed in sterile 1X PBS three times for 30 min each, observed between washing to ensure no materials or feces were in contact, then mounted in 35% glycerol for confocal microscopy or transferred to sterile 1X PBS. After washing three additional times in sterile 1X PBS to assure the elimination of surface contamination, a new sterile, moistened cotton applicator was used to transfer the egg to an agar plate. Each egg was punctured with a sterile needle, releasing the contents onto the agar. Similarly, larvae were washed three times with sterile 1X PBS, then homogenized and plated.
Development of Y. pestis-infected eggs or larvae
Eggs and larvae were isolated, washed in sterile 1X PBS as described above, and placed into a sterile flea chamber with mesh on the top and bottom containing sterile sawdust and larval food (one part blend of equal parts sterile powdered bovine blood, ground mouse chow, dry milk to three parts sterile sawdust)42. Development of eggs into larvae, pupae, and F1 adults was relatively efficient, with the majority of the collected eggs reaching the adult stage. Pupae were collected and washed in sterile ddH2O three times to remove loose debris and any possible contaminants. Micro-scissors were used to first gently remove the apical ends of the cocoon to allow for easier viewing under a dissection scope, then the remainder of the cocoon was removed. The pupae were then washed three times in sterile 1X PBS before imaging or plating. Larvae and pupae were mounted in 35% glycerol for microscopy. All egg, larvae, pupae, and F1 adult fleas that were collected and successfully processed in all trials were analyzed by bacterial load determination or confocal microscopy.
Confocal microscopy
Midguts, reproductive tissue, eggs, larvae, and pupae were imaged using a Leica SP8 confocal microscope with settings of 1024 × 1024 pixels, numerical aperture of 0.4, refractive index of 1, pinhole of 1 AU, excitation of 554 nm, and emission of 580 nm. Image J version 1.54 or Fiji version 2.9 software was used to visualize the localization of bacteria and to quantify fluorescent signal47,48. For quantification of proventricular colonization by microscopy, images were converted to 8-bit grayscale, and the fluorescence was quantified as the mean gray value. Mean gray value was used to calculate integrated density (ID) to determine relative fluorescent units (RFU) as a product of area of the sample and mean gray value. Control fleas (n = 10) given blood that was not infected and analyzed 1 day later were used to calculate autofluorescence/ background signal; for graphing data points, background ID was subtracted from the sample RFU values. Proventricular colonization was defined as measurable bacterial tdTomato fluorescence on the proventricular spines.
Immunohistochemistry
Monoclonal anti-Yersinia antibody 6B5 (catalog # EGA401, Kerafast, Massachusetts, USA) has been validated for its binding specificity to a 30 kDa surface antigen of Y. pestis, expressed at 26 and 37 °C, with no cross reactivity to 26 related gram-negative Enterobacteriaceae including Y. enterocolitica, Salmonella enterica, and Escherichia. coli49. Antibodies were biotinylated prior to use in experiments; staining conditions were established on Y. pestis cultured in HIB. For antibody preparation, 2 mg sulfo-NHS-LC biotin was added to monoclonal anti-Yersinia antibody 6B5 and incubated on ice for 2 h or at room temperature for 30 min, then dialyzed in 1X PBS. After collection, eggs and tissues were washed in 1X PBS, followed by two washes in 70% ethanol to eliminate any possible surface contamination, and a final wash in 1X PBS. All tissues and eggs were fixed in 4% paraformaldehyde at room temperature for 30 min or at 4 °C for 2 h, then washed three times in 1X PBS. Samples were permeabilized in 0.5% triton X-100, washed in 1X PBS, and incubated at 4 °C overnight with biotinylated anti-Yersinia antibody (diluted 1:100). Primary antibody-labeled samples were washed three times in 1X PBS, incubated in streptavidin conjugated to Texas Red (Excitation: 596, Emission 615) for 30 min, washed in three times 1X PBS and imaged as described above using the Leica SP8 confocal microscope.
Quantification of bacterial load
For plating, three midguts were homogenized in 30 µL sterile 1X PBS; single larva or pupa samples were homogenized in sterile 1X PBS; and egg samples were lysed directly onto agar plates. Serial dilutions were performed in sterile 1X PBS, and all dilutions were plated in duplicate onto heart infusion agar (HIA) or Yersinia selective agar (YSA). Isolated colonies from eggs, larvae, pupae, and F1 adults were streaked for isolation before storing in bacteriological freezing media (5% monosodium glutamate, 5% bovine serum albumin) at −80 °C.
PCR and genome annotation
Bacteria that were isolated from eggs were grown overnight, and DNA was isolated using a Quick-DNA® Microprep kit (Zymo Research, California, USA). Conventional PCR amplification of caf1 and pla were performed using the primers shown in Supplemental Table S2. All positive PCRs were confirmed by sequencing (shown in FigShare); sequenced nucleotides were aligned in Geneious Prime (GraphPad Software, LLC, Massachusetts, USA) using MAFFT version 750.
Transmission study
Transmission assays were carried out for fleas infected with KIM6 + ptdTomato using rat blood (Biochemed Services, Virginia, USA)51. Briefly, egg-isolated bacteria were used to infect adult fleas in an artificial feeder. On day 3 or 7 post-infection, groups of ten fleas were fed on uninfected rat blood in the artificial membrane feeder for 1 h. Blood and skin were processed to quantify Y. pestis by serial dilution and plating to determine the number of bacteria transmitted per group of 10 X. cheopis.
Transmission electron microscopy
Fleas were either fed an uninfected rat bloodmeal or a Y. pestis-infected rat bloodmeal, and then groups of 5–10 were separated based on sex. On day 1 or 3, fleas were euthanized, and the ovaries, testes, and midguts were dissected and fixed in 2% paraformaldehyde, and 2% glutaraldehyde in 100 mM sodium cacodylate buffer with a pH of 7.35. Each sample was allowed to settle, and the resulting tissue was resuspended in Histogel (Thermo Fisher Scientific, Massachusetts, USA). Tissues were rinsed in 100 mM sodium cacodylate buffer with a pH of 7.35 containing 130 mM sucrose. Secondary fixation was performed using 1% osmium tetroxide (Ted Pella, Inc. California, USA) in cacodylate buffer. Specimens were incubated at 4 °C for 1 h, then rinsed with cacodylate buffer and further with distilled water. En bloc staining was performed using 1% aqueous uranyl acetate and incubated at 4 °C overnight, then rinsed with distilled water. A graded dehydration series was performed using ethanol, transitioned into acetone, and dehydrated tissues were infiltrated with Epon resin and polymerized at 60 °C overnight. Sections were cut to a thickness of 75 nm using an ultramicrotome (Ultracut UCT, Leica Microsystems, Hesse, Germany) and a diamond knife (Diatome, Pennsylvania, USA). Images were acquired with a JEOL JEM 1400 transmission electron microscope (JEOL, Massachusetts, USA) at 80 kV on a Gatan Rio CMOS camera (Gatan, Inc, California, USA).
DNA sequencing and annotation
For sequencing, bacteria were isolated after harvesting from an infected flea egg; and chromosomal DNA was prepared using the Zymo DNA/RNA Miniprep (Zymo Research Corporation, California, USA) prior to processing for whole genome sequencing using the NovaSeq platform. Sequencing data was annotated using Geneious Prime version 2024.0.1 (Geneious, Massachusetts, USA) and Proksee52.
Statistical analysis
For quantifying Y. pestis in the midgut by microscopy or CFU determination, group sizes were selected based on pre-established criteria formulated by power analysis to detect a 50% change with alpha <0.05, 80% confidence. All experiments were conducted with a minimum of two independent trials, and data from all trials were pooled for statistical evaluation. Kruskal–Wallis followed by Dunn’s multiple comparison’s test was used for evaluation of CFU and RFU data; descriptive statistics were evaluated using SPSS version 26 and graphed using OriginPro version 10.1 (OriginLab Corporation, Massachusetts, USA).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
Source data are provided with this paper. The sequences of PCR-amplified products, chromatographs and alignments, image data sets, and corresponding metadata that support the findings of this study are available in Figshare (https://doi.org/10.6084/m9.figshare.25674594). The Y. pestis genome sequencing data from the original infecting strain KIM6+ (pgm+, pCD1-, pPCP1+, pMT1+, pNE160-tdTomato) and the bacteria isolated from eggs derived from infected X. cheopis adults have been deposited in the NCBI database under accession codes SAMN42388871 and SAMN42388872.
References
Abedi, A. et al. Ecologic features of plague outbreak areas, Democratic Republic of the Congo, 2004–2014. Emerg. Infect. Dis. 24, 210–220 (2018).
Andrianaivoarimanana, V. et al. Understanding the persistence of plague foci in Madagascar. PLos Negl. Trop. Dis. 7, e2382 (2013).
Biggins, D. & Eads, D. Prairie dogs, persistent plague, flocking fleas, and pernicious positive feedback. Front. Vet. Sci. 6, 1–12 (2019).
Davis, D. Plague in Africa from 1935–1949: a survey of wild rodents in African territories. Bull. World Health Organ. 9, 665–700 (1953).
Davis, D. Plague in South Africa: a study of the epizootic cycle in gerbils (Tatera brantsi) in the northern orange free state. J. Hyg. 51, 427–449 (1953).
Du, H., Wang, Y., Zhuang, D. & Jiang, X. Temporal and spatial distribution characteristics in the natural plague foci of Chinese Mongolian gerbils based on spatial autocorrelation. Infect. Dis. Poverty 6, 1–10 (2017).
Vogler, A. et al. Temporal phylogeography of Yersinia pestis in Madagascar: insights into the long-term maintenance of plague. PLoS Neg. Trop. Dis. 11, e0005887 (2017).
Eisen, R., Eisen, L. & Gage, K. Studies of vector competency and efficiency of North American fleas for Yersinia pestis: state of the field and future research needs. J. Med. Entomol. 46, 737–744 (2009).
Gage, K. & Kosoy, M. Natural history of plague: perspectives from more than a century of research. Ann. Rev. Entomol. 50, 505–528 (2005).
Eisen, R. et al. Early phase transmission of Yersinia pestis by unblocked fleas as a mechanism explaining rapidly spreading plague epizootics. Proc. Natl Acad. Sci. USA 103, 15380–15385 (2006).
Bland, D., Jarrett, C., Bosio, C. & Hinnebusch, B. Infectious blood source alters early foregut infection and regurgitative transmission of Yersinia pestis by rodent fleas. PLoS Pathog. 14, e1006859 (2018).
Hinnebusch, B., Jarrett, C. & Bland, D. Molecular and genetic mechanisms that mediate transmission of Yersinia pestis by fleas. Biomolecules 11, 210 (2021).
Bobrov, A. G., Kirillina, O. & Perry, R. D. Regulation of biofilm formation in Yersinia pestis. Adv. Exp. Med. Biol. 603, 201–210 (2007).
Jarrett, C. et al. Transmission of Yersinia pestis from an infectious biofilm in the flea vector. J. Infect. Dis. 190, 783–792 (2004).
Hinnebusch, B. The evolution of flea-borne transmission in Yersinia pestis. Curr Iss. Mol. Biol. 7, 197–212 (2005).
Mahmoudi, A. et al. Plague reservoir species throughout the world. Integr. Zool. 16, 798–809 (2021).
Burroughs, A. L. Sylvatic plague studies: the vector efficiency of nine species of fleas compared with Xenopsylla cheopis. J. Hyg. 45, 371–396 (1947).
Hinnebusch, B., Bland, D., Bosio, C. & Jarrett, C. Comparative ability of Oropsylla montana and Xenopsylla cheopis fleas to transmit Yersinia pestis by two different mechanisms. PLoS Neg. Trop. Dis. 11, e0005276 (2017).
Dewitte, A. et al. A refined model of how Yersinia pestis produces a transmissible infection in its flea vector. PLoS Pathog. 16, e1008440 (2020).
Hinnebusch, B. et al. Role of Yersinia murine toxin in survival of Yersinia pestis in the midgut of the flea vector. Science 296, 733–735 (2002).
Bland, D., Miarinjara, A., Bosio, C., Calarco, J. & Hinnebusch, B. Acquisition of yersinia murine toxin enabled Yersinia pestis to expand the range of mammalian hosts that sustain flea-borne plague. PLoS Pathog. 17, e1009995 (2021).
Mitchell, C. et al. A role for early-phase transmission in the enzootic maintenance of plague. PLoS Pathog. 18, e1010996 (2022).
Bland, D., Eisele, N., Keleher, L., Anderson, P. & Anderson, D. Novel genetic tools for diaminopimelic acid selection in virulence studies of Yersinia pestis. PLoS ONE 6, e17352 (2011).
Kernif, T. et al. Responses of artificially reared cat fleas Ctenocephalides felis felis (Bouche, 1835) to different mammalian bloods. Med. Vet. Entomol. 29, 171–177 (2015).
Krasnov, B. Functional and Evolutionary Ecology of Fleas: A Model for Ecological Parasitology (Cambridge Univ. Press, 2008).
Chen, J. & Wang, D. Comparative morphology of rodent flea eggs in China. Med. Vet. Entomol. 7, 384–386 (1993).
Prasad, R. Host dependency among haematophagous insects: a case study on flea-host association. Proc. Anim. Sci. 96, 349–360 (1987).
Fetherston, J. & Perry, R. Yersinia pestis - etiologic agent of plague. Clin. Microbiol. Rev. 10, 35–66 (1997).
Bacot, A. & Martin, C. LXVII. Observations on the mechanism of the transmission of plague by fleas. J. Hyg. 13, 423–439 (1914).
Molyneux, D. Investigations into the possibility of intervector transmission of Pasteurella pestis. Ann. Trop. Med. Parasit. 63, 403–407 (1969).
Farhang-Azad, A., Traub, R. & Baqar, S. Transovarial transmission of murine typhus rickettsiae in Xenopsylla cheopis fleas. Science 227, 543–545 (1985).
Adams, J., Schmidtmann, E. & Azad, A. Infection of colonized cat fleas, Ctenocephalides felis (Bouche), with a Rickettsia-like microorganism. Am. J. Trop. Med. Hyg. 43, 400–409 (1990).
Thepparit, C., Hirunkanokpun, S., Popov, V., Foil, L. & Macaluso, K. Dissemination of bloodmeal acquired Rickettsia felis in cat fleas, Ctenocephalides felis. Parasit. Vectors 6, 149 (2013).
Bartra, S. et al. Resistance of Yersinia pestis to complement-dependent killing is mediated by the Ail outer membrane protein. Infect. Immun. 76, 612–622 (2008).
Bartra, S. et al. The outer membrane protein A (OmpA) of Yersinia pestis promotes intracellular survival and virulence in mice. Micro. Pathog. 52, 41–46 (2012).
Seo, K. S. et al. Role of a new intimin/invasin-like protein in Yersinia pestis virulence. Infect. Immun. 80, 3559–3569 (2012).
Grabenstein, J., Marceau, M., Pujol, C., Simonet, M. & Bliska, J. The response regulator PhoP of Yersinia pseudotubeculosis is important for replication in macrophages and for virulence. Infect. Immun. 72, 4973–4984 (2004).
Pujol, C., Grabenstein, J., Perry, R. & Bliska, J. Replication of Yersinia pestis in interferon g-activated macrophages requires ripA, a gene encoded in the pigmentation locus. Proc. Natl Acad. Sci. USA 102, 12909–12914 (2005).
Ke, Y., Chen, Z. & Yang, R. Yersinia pestis: mechanisms of entry into and resistance to the host cell. Front. Cell Infect. Microbiol. 3, 106 (2013).
Vadyvaloo, V. et al. Role of the PhoP-PhoQ gene regulatory system in adaptation of Yersinia pestis to environmental stress in the flea digestive tract. Microbiology 161, 1198–1210 (2015).
McCutcheon, J. & Moran, N. Extreme genome reduction in symbiotic bacteria. Nat. Rev. Microbiol. 10, 13–26 (2011).
Bland, D., Brown, L., Jarrett, C., Hinnebusch, B. & Macaluso, K. Methods in flea research. https://www.beiresources.org/Portals/2/VectorResources/Methods%20in%20Flea%20Research.pdf (2017).
Fetherston, J., Kirillina, O., Bobrov, A., Paulley, J. & Perry, R. The yersiniabactin transport system is critical for the pathogenesis of bubonic and pneumonic plague. Infect. Immun. 78, 2045–2052 (2010).
Devignat, R. Varieties of Pasteurella pestis; new hypothesis. Bull. World Health Organ. 4, 247–263 (1951).
Brubaker, R. R. Interconversion of purine mononucleotides in Pasteurella pestis. Infect. Immun. 1, 446–454 (1970).
Earl, S. et al. Resistance to innate immunity contributes to colonization of the insect gut by Yersinia pestis. PLoS ONE 10, e0133318 (2015).
Schneider, C., Rasband, W. & Eliceiri, K. NIH image to ImageJ: 25 years of image analysis. Nat. Methods 9, 671–675 (2012).
Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012).
Zhao, T., Zhao, P. & Doyle, M. Detection and isolation of Yersinia pestis without fraction 1 antigen by monoclonal antibody in foods and water. J. Food Prot. 75, 1555–1561 (2012).
Katoh, K., Misawa, K., Kuma, K. & Miyata, T. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30, 3059–3066 (2002).
Lorange, E., Race, B., Sebbane, F. & Hinnebusch, B. Poor vector competence of fleas and the evolution of hypervirulence in Yersinia pestis. J. Infect. Dis. 191, 1907–1912 (2005).
Grant, J. et al. Proksee: In-depth characterization and visualization of bacterial genomes. Nucleic Acids Res. 51, W484–W492 (2023).
Acknowledgements
We wish to give special thanks to Dr. Alexander Jurkevich (University of Missouri) for expert technical guidance with confocal imaging and to DeAna Grant (University of Missouri) for conducting the transmission electron microscopy processing and imaging. In addition, we are grateful to Lynda Watson (PMS Recycled Vermin, Texas, USA) for the gift of prairie dog blood, to Dr. Darla Tharp (University of Missouri) for the gift of pig blood, and to Dr. David Bland and Dr. Joe Hinnebusch (NIH/NIAID Rocky Mountain Laboratory, Montana, USA) for the gift of X. cheopis fleas for rearing. We are also grateful to Dr. Paul Anderson and Jenna Canfield at the University of Missouri for helpful discussion and critical comments on the manuscript. Sequencing was performed by the University of Missouri Genomics Technology Core; microscopy was conducted at the University of Missouri Advanced Light Microscopy Core; and transmission electron microscopy was conducted at the Electron Microscopy Core at the University of Missouri. This work was supported by PHS NIH/NIAID 1R21AI178547 (DMA) and the University of Missouri System Tier 2 Strategic Investment Program (DMA and BTB).
Author information
Authors and Affiliations
Contributions
C.D.P. designed and performed experiments, interpreted data, and co-wrote the manuscript; B.T.B. and Q.S. contributed to data interpretation and editing of the manuscript; D.M.A. contributed to experimental design and interpretation, and co-wrote the manuscript. All authors approved the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Joseph Hinnebusch and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Pauling, C.D., Beerntsen, B.T., Song, Q. et al. Transovarial transmission of Yersinia pestis in its flea vector Xenopsylla cheopis. Nat Commun 15, 7266 (2024). https://doi.org/10.1038/s41467-024-51668-0
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-024-51668-0
