Introduction

Yersinia pestis, the causative agent of plague, relies on flea vectors to spread between mammalian hosts [1]. After a blood meal from a septicemic host, the bacteria rapidly encounter oxidative stress and other antibacterial pressures in the flea gut [2–6]. Within hours of feeding, a transient bactericidal soft mass forms in the proventriculus (foregut) and entraps the bacteria [5]. Despite this early barrier, Y. pestis secretes extracellular polysaccharides that consolidate this mass into a dense biofilm that can block the proventriculus and facilitate transmission [2,7]. While biofilm-mediated foregut blockage is well studied [6], the molecular events that enable Y. pestis to survive the critical early hours of flea colonization remain poorly understood.

Traditionally, blood has been regarded as little more than a passive vehicle and nutrient source for both the flea and the pathogen [6,7]. Yet, beyond its nutritional role, blood contains molecules with the potential to shape the outcome of infection. Recent work has hinted that the origin and composition of the blood meal can influence colonization dynamics [8,9], but whether host blood components actively promote the early survival of Y. pestis in the flea (defined here as the first hours post-feeding (up to 7 h), corresponding to a critical early phase in the kinetics of flea infection) has remained an open question. Here, we show that albumin contributes to early Y. pestis survival in the flea, consistent with a role in counteracting constitutive oxidative defenses during the early post-feeding phase.

Results and discussion

We first asked whether cellular elements are required during early colonization. To this end, fleas were fed on whole blood, plasma, or serum containing Y. pestis (Fig 1A), and bacterial survival and recovery were monitored. Infection rates and bacterial loads were indistinguishable at ≤1, 24, and 48 h post-feeding (hpf) (Fig 1B and 1C). Under these conditions, cellular components were not required for early colonization, and plasma alone was sufficient to support initial survival.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. A.Experimental design for panels B-E, K-L.

Infection prevalence (B,D,K) and bacterial loads (c.f.u. per flea; C,E,L) measured at ≤1, 24 and 48 h post-feeding (hpf) following feeds on whole blood, serum, plasma, or HPLM supplemented or not with bovine serum albumin (BSA), protease-treated BSA (BSA Dig), synthetic peptide (Peptide) or lipid fraction. n = 2 biological replicates for B and C (5-6 fleas per replicate) and n = 3 for D, E, K and L (15 fleas per replicate). F. Experimental design for panels G-I. G. Hydrogen peroxide levels were measured in whole dissected guts from fleas pre-treated with 20 mM NAC, with samples collected before infection and at < 1 h and 48 h post-feeding (hpf). n = 6 biological replicates, each a pool of six guts. H,I. Infection prevalence (H) and bacterial loads (I) in fleas at ≤1, 24 and 48 hpf on HPLM. n = 3 biological replicates with 6 fleas (H) and 15 fleas (I) per replicate. J. Bacterial survival in HPLM with or without BSA measured by bioluminescence as the ratio of relative light units before (N₀) and after (N) a 10 min exposure to 20 mM H₂O₂ (N₀/N). Data are mean ± s.d. n = 16 biological replicates. Data are presented as mean ± 95% confidence interval. Statistical analyses were performed using two-way ANOVA. Statistical significance is indicated as follows: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

https://doi.org/10.1371/journal.ppat.1014022.g001

To identify which plasma constituents support early colonization, we fed fleas on Human Plasma-Like Medium (HPLM) [10], which lacks proteins and lipids, and assessed colonization by Y. pestis. Survival fell within ≤1 h post-feeding and reached zero by 48 hpf (Fig 1d and 1e). These results raise the possibility that HPLM either fails to support bacterial growth, exerts a detrimental effect on the bacteria, or compromises survival during the pre-ingestion phase in the feeder (glass artificial feeding apparatus used to deliver the infectious blood meal). However, HPLM supported in vitro growth of Y. pestis at both 21°C and 37°C, sustaining bacterial proliferation despite modest differences in growth kinetics relative to standard laboratory media (S1a-c Fig). In line with this, bacterial counts remaining in the glass artificial feeder after 1 hour were comparable across meal (S1d Fig), ruling out significant bacterial death prior to ingestion.

We therefore tested whether plasma constituents could restore early survival by supplementing HPLM with either bovine serum albumin (BSA) or a defined lipid mixture. Only BSA rescued early bacterial viability, partially rescuing infection rates and bacterial loads (Fig 1D and 1E), whereas the lipid mixture had no effect. Together, these data identify plasma proteins as key blood-borne constituents contributing to early survival in the flea.

Because oxidative stress is a constitutive defense in the flea gut [4,11], we tested whether plasma proteins promote early Y. pestis survival by mitigating this stress. Pre-feeding fleas with the thiol antioxidant N-acetylcysteine (NAC) lowered whole-gut H₂O₂ levels (Fig 1F and 1G) and increased Y. pestis survival when fleas were later fed on HPLM without BSA (Fig 1H and 1I). Importantly, NAC was absent from the infectious meal and is bactericidal to Y. pestis at the concentrations used (S1 Table), indicating a host-mediated effect.

Consistent with a redox-based mechanism, BSA supplementation attenuated the loss of Y. pestis survival induced by H₂O₂ in an in vitro oxidative stress assay (Fig 1J). Although we did not directly quantify H₂O₂ levels in the presence of BSA in vivo, the similar effects of NAC pre-treatment and BSA supplementation on early bacterial survival (Fig 1D,1E,1H, and 1I), together with the attenuation of H₂O₂-induced survival loss observed in vitro (Fig 1J), support a role for plasma proteins in mitigating oxidative stress in the flea gut.

To further explore the molecular basis of this effect, we tested whether bacterial survival requires intact BSA or could be recapitulated by free thiol-containing amino acids. Survival was preserved in fleas fed on HPLM supplemented with intact BSA or protease-treated BSA, but not in the presence of a free cysteine-containing peptide (Fig 1K-1L). These observations suggest that the observed effect does not rely solely on free cysteine residues or amino acid composition, but instead depends on protein or large-fragment context rather than free amino acids alone. The precise molecular mechanisms underlying this effect remain to be determined.

Albumin is the most abundant plasma protein, representing approximately 60% of total plasma protein content. In this study, we used commercially available BSA as a representative plasma protein to investigate the contribution of plasma components to early Y. pestis survival in the flea. A limitation of this approach is that albumin was not purified directly from plasma and that other plasma constituents, such as globulins or fibrinogen, which together represent a substantial fraction of the plasma proteome, were not individually tested. As a result, this experimental setup may not fully recapitulate the complexity of the plasma protein milieu encountered by Y. pestis in vivo, and additional plasma components may also contribute to the observed effect on early bacterial survival.

Collectively, these results underscore that in vector biology, the host’s own blood can shape the earliest steps of infection, a principle likely to extend to many arthropod-borne pathogens.

Materials and methods

Supporting information

Acknowledgments

The authors would like to thank Dr Benjamin Baetz, Camille Dupont and Blandine Charlet for their assistance, and members of Sebbane’s lab for discussion and suggestions.