ABSTRACT
Transmission of Yersinia pestis by fleas depends on the formation of condensed bacterial aggregates embedded within a gel-like matrix that localizes to the proventricular valve in the flea foregut and interferes with normal blood feeding. This is essentially a bacterial biofilm phenomenon, which at its end stage requires the production of a Y. pestis exopolysaccharide that bridges the bacteria together in a cohesive, dense biofilm that completely blocks the proventriculus. However, bacterial aggregates are evident within an hour after a flea ingests Y. pestis, and the bacterial exopolysaccharide is not required for this process. In this study, we characterized the biochemical composition of the initial aggregates and demonstrated that the yersinia murine toxin (Ymt), a Y. pestis phospholipase D, greatly enhances rapid aggregation following infected mouse blood meals. The matrix of the bacterial aggregates is complex, containing large amounts of protein and lipid (particularly cholesterol) derived from the flea’s blood meal. A similar incidence of proventricular aggregation occurred after fleas ingested whole blood or serum containing Y. pestis, and intact, viable bacteria were not required. The initial aggregation of Y. pestis in the flea gut is likely due to a spontaneous physical process termed depletion aggregation that occurs commonly in environments with high concentrations of polymers or other macromolecules and particles such as bacteria. The initial aggregation sets up subsequent binding aggregation mediated by the bacterially produced exopolysaccharide and mature biofilm that results in proventricular blockage and efficient flea-borne transmission.
IMPORTANCE
Yersinia pestis, the bacterial agent of plague, is maintained in nature in mammal-flea-mammal transmission cycles. After a flea feeds on a mammal with septicemic plague, the bacteria rapidly coalesce in the flea’s digestive tract to form dense aggregates enveloped in a viscous matrix that often localizes to the foregut. This represents the initial stage of biofilm development that potentiates transmission of Y. pestis when the flea later bites a new host. The rapid aggregation likely occurs via a depletion-aggregation mechanism, a non-canonical first step of bacterial biofilm development. We found that the biofilm matrix is largely composed of host blood proteins and lipids, particularly cholesterol, and that the enzymatic activity of a Y. pestis phospholipase D (Ymt) enhances the initial aggregation. Y. pestis transmitted by flea bite is likely associated with this host-derived matrix, which may initially shield the bacteria from recognition by the host's intradermal innate immune response.
KEYWORDS: Y. pestis, plague, biofilms, arthropod vectors
INTRODUCTION
Fleas transmit Yersinia pestis by regurgitation from the foregut during blood feeding, and the development of a transmissible infection in the flea occurs in stages. Within an hour after a flea acquires Y. pestis in an infectious blood meal, the bacteria aggregate into large clumps in the flea digestive tract that are embedded in a viscous gel-like material (1, 2). This first establishment stage enables the persistence of the infection and avoidance of rapid elimination by defecation, which is particularly important since Y. pestis does not directly associate with or adhere to flea digestive tract tissues or structures.
Within a few hours, the newly formed bacterial aggregates can localize to the proventriculus, the valve in the flea foregut between the midgut and esophagus (1, 2). This second stage is hypothesized to account for early-phase regurgitative transmission because in some fleas the aggregates can fill the proventriculus to an extent that transiently interferes with normal blood feeding (1, 3). At this stage, however, the bacterial aggregates are not firmly entrenched in the proventriculus and after initially impeding blood feeding, potentiating some transmission, they are dislodged and swept back into the midgut, allowing normal blood feeding to resume (1, 2). Early-phase transmission occurs primarily during the first blood meal after the infectious blood meal and is inefficient, with few bacteria transmitted, and requires several newly infected fleas feeding simultaneously for successful transmission (4–6).
The third stage of a transmissible infection occurs after Y. pestis produces an extracellular polysaccharide (poly-β-1,6-N-acetyl-D-glucosamine), encoded by the Hms operon. This exopolysaccharide is not required for the initial aggregation and proventricular localization (1, 2) or normal early-phase transmission efficiency (3, 7), but is required for biofilm formation in vitro and for the fixation of the bacterial aggregates to the proventriculus (8–11). This more cohesive, mature biofilm infection develops within one to a few weeks after infection, and in some fleas consolidates to fill the proventriculus and prevent blood flow into the midgut during feeding attempts, again leading to regurgitation of blood mixed with bacteria into the bite site (9, 12). These completely blocked fleas are the most efficient transmitters, as they can transmit large numbers of bacteria during continuous, persistent feeding attempts before they succumb to starvation within a few days (6). Thus, the two transmission modes essentially correspond to different stages of Y. pestis biofilm development in the flea, characterized by aggregated bacteria within a matrix of extracellular polymeric substances (EPS).
An early hypothesis of Y. pestis aggregation in the flea gut proposed that the bacteria are embedded in a fibrin clot formed by a secondary, low-temperature coagulase activity of the Y. pestis plasminogen activator Pla, a cell-surface protease that is an important virulence factor in mammalian hosts because it acts to degrade fibrin clots in vivo (13, 14). This hypothesis was disproven when it was found that Pla-negative Y. pestis can block fleas normally (15). Furthermore, typical aggregation occurs when fleas are infected using defibrinated blood sources, the biofilm aggregates are not degraded by proteases, including plasmin, and the Pla coagulase activity is not seen with rodent blood (9, 15).
The EPS associated with biofilm formed by bacteria in natural settings is usually composed of a complex mixture of macromolecules (16). A universal component is a bacterial extracellular polysaccharide induced at the onset of biofilm development. However, the biofilm matrix also often incorporates proteins, lipids, extracellular DNA (eDNA), and other components derived from the bacteria or adsorbed from the environment. The EPS associated with Y. pestis in the flea has not been characterized. The aggregates have been described as gel-like, enveloped in a viscous, amorphous matrix that is brown-colored, presumably due to the accretion of hemin byproducts from the flea blood meal (9, 12). The goal of this study was to biochemically characterize the EPS of the Y. pestis aggregates that form after infection and to identify factors required for this initial stage of biofilm formation. This information is pertinent to the pathogenesis of bubonic plague because Y. pestis transmitted from the flea is presumably associated with EPS, and it is in this context that the host first encounters the bacteria in the dermis.
RESULTS
Within a few hours after a flea ingests highly bacteremic blood, many Y. pestis accumulate into discrete, dense aggregates in the midgut that often localize to the proventricular valve (1, 2) (Fig. 1). These aggregates represent the initial step of biofilm formation in the flea, which does not require the Hms operon-encoded bacterial extracellular polysaccharide that is the hallmark of the mature Y. pestis biofilm (1). The bacteria in these initial aggregates are, however, embedded within a viscous, amorphous matrix of unknown composition (1, 9). To better understand this initial stage of biofilm formation, we first determined whether host blood components were involved (Fig. 2). Y. pestis produced comparable aggregative infections at 24 h in fleas that fed on bacteremic mouse blood or serum. By contrast, fleas that fed on a PBS suspension of washed mouse red blood cells or artificial plasma (buffered salts plus albumin) containing Y. pestis did not—the incidence of proventricular aggregates was lower, and bacterial numbers in many of these fleas had decreased by 24 h. This reduction is probably due to the rapid elimination of many unaggregated bacteria in flea feces because Y. pestis survives well in washed red blood cell solutions and artificial plasma in vitro. These results indicate that the presence of one or more components of host blood serum significantly enhances the prominent initial stage of Y. pestis biofilm formation in the flea.
Fig 1.
Y. pestis aggregates in the flea digestive tract and localizes to the proventriculus soon after an infected blood meal. (A) X. cheopis flea after feeding, with different parts of the digestive tract indicated. The white box outlines the area imaged in the other panels. Light microscopy of dissected digestive tracts from an uninfected flea (B), and fleas 1 h (D) or 24 h (F) after feeding on mouse blood containing Y. pestis KIM6+ (pAcGFP1) showing a brown-colored aggregate associated with the proventriculus in the infected fleas. Images are representative of fleas scored as PV+ in this study. Fluorescence microscopy of the same dissection preps (C, E, G) verifies that the aggregates contain large numbers of the green-fluorescent (GFP+) bacteria. The proventricular spines autofluoresce yellow. Scale bar = 0.1 mm.
Fig 2.
Early infection status of X. cheopis fleas that fed on mouse blood, mouse serum, a suspension of washed mouse red blood cells (RBC), or artificial plasma containing Y. pestis KIM6+ (pAcGFP1). (A) Percentage of fleas still infected 24 h after feeding on the different bacteremic sources. (B) Average bacterial load in fleas immediately after feeding (0 h) and 24 h after infection. (C) Percentage of fleas with a colonized proventriculus (PV+) 24 h after infection. (D) Light and (E) fluorescence microscopy images from a typical PV+ flea dissected 24 h after infection using mouse serum, showing the dense biofilm-like aggregate of Y. pestis that occupies and extends from the proventriculus. For panels A–C, the means and SEM of three independent experiments are indicated; in A and C, the small black circles indicate the means of the three individual experiments. *, a significant difference (P < 0.01) from infections using mouse blood or serum; **, 24 h/0 h CFU ratio significantly different (P < 0.01) from infections using mouse blood (one-way ANOVA and Tukey’s multiple comparisons test).
Immunostaining and microscopic imaging
Immunostaining and confocal microscopy were applied to Y. pestis aggregates dissected from the posterior end of the proventriculus of fleas 24 h after infection using whole blood. In separate experiments, aggregates were immunostained using polyclonal rabbit serum raised against Y. pestis KIM6+, mouse red blood cells, or mouse serum as the primary antibody. Other aggregates were examined by DIC and transmission electron microscopy (TEM). Peripheral regions of the aggregates stained brightly in indirect immunofluorescence assays using antibodies to both mouse red blood cells and serum proteins (Fig. 3A and B). Patchy amorphous material was also seen in assays using anti-Y. pestis antibody and by DAPI staining, indicating that protein from dead or degraded bacteria and extracellular DNA (eDNA) from these bacteria or degraded mouse white blood cells were also present (Fig. 3A through C). The TEM images revealed that Y. pestis was unevenly distributed within the aggregates, with regions of variable bacterial density (Fig. 3D through F). Several rod-shaped and rectangular crystals typical of cholesterol (17) were also evident in DIC and TEM images (Fig. 3D through F).
Fig 3.
Imaging different components of the Y. pestis biofilm aggregates from the flea. Confocal microscopy images of aggregates dissected from fleas 24 h after infection with Y. pestis KIM6+ (pAc1GFP) stained with anti-Y. pestis (A), anti-mouse red blood cell (B), or anti-mouse serum (C) polyclonal antibodies and with DAPI. The secondary antibody used for immunostaining fluoresces red, the GFP+ bacteria are green, and eDNA stains blue with DAPI. Individual aggregate imaged by differential interference contrast microscopy (D) or by transmission electron microscopy (TEM) (E, F) shows the amorphous nature of the aggregate matrix that incorporates refractile cholesterol crystals (blue arrows). Aggregates A–C and E were from fleas infected using mouse blood, and D and F from fleas infected using mouse serum. PS, a cross-section of proventricular spines. Scale bars = 5 µm (A–C); 50 µm (D); and 2 µm (E, F). Images are representative of at least 10 individual samples examined.
Proteomics profile of the aggregates
The confocal imaging indicated that blood proteins were major components of the initial biofilm matrix, so we analyzed the protein profile of aggregates collected from fleas 24 h after infection. Mass spectrometry analysis was performed on these samples as well as on control samples of sterile mouse blood, Y. pestis in vitro biofilm grown at 21°C, and the midgut contents of fleas collected 24 h after feeding on sterile mouse blood.
Around 1,700 proteins were identified in the biofilm aggregates from fleas infected using whole blood; 54% of these were Y. pestis proteins; 23% were mouse proteins, and 23% were flea proteins. Only trace amounts of most of them were detected (<0.1% of the total protein content). Cumulatively, mouse proteins accounted for 72% of the total protein content, Y. pestis proteins for 20%, and only 8% were flea proteins. The 25 most abundant proteins accounted for 76% of all the protein in the sample; and of these 10 were bacterial and 10 from mouse blood (Table 1). Five of the ten mouse proteins derive from hemoglobin or red blood cell membranes, and these five alone accounted for 64% of the total protein content in the sample. Only 5 of the top 25 were identified as flea protein.
TABLE 1.
The most abundant proteins detected in Y. pestis biofilm dissected from fleas 24 h after infection using mouse blood and their rank in relevant control samplesa
| Protein name | Source | Abundance rank and (percentage of total protein) in: | ||||
|---|---|---|---|---|---|---|
| Flea biofilm | Mouse blood | Flea gut contents | Y. pestis in vitro 21°C biofilm | Y. pestis in vitro 37°C | ||
| Beta-globins 1 and 2 | Mouse | 1 (44) | 1 (61) | 8 (5.5) | − | − |
| Alpha-globin | Mouse | 2 (16) | 2 (27) | 12 (1.9) | − | − |
| Anion transport protein | Mouse | 3 (3) | 13 (0.2) | 91 (0.01) | − | − |
| Ail protein | Y. pestis | 4 (1.7) | − | − | 2 (3.4) | 4 (1.7) |
| Glutamine transaminase K | X. cheopis | 5 (1.4) | − | ND | − | − |
| Mitochondrial carrier protein | X. cheopis | 6 (1.4) | − | ND | − | − |
| Outer membrane protein F | Y. pestis | 7 (1.0) | − | − | 4 (2.5) | 7 (1.0) |
| Outer membrane protein A | Y. pestis | 8 (0.9) | − | − | 1 (5.2) | 8 (0.9) |
| Aquaporin-1 | Mouse | 9 (0.8) | 48 (0.02) | 120 (0.03) | − | − |
| Outer membrane protein C | Y. pestis | 10 (0.7) | − | − | 16 (0.9) | 10 (0.7) |
| Yersinia murine toxin (Ymt) | Y. pestis | 11 (0.6) | − | − | 7 (1.3) | 11 (0.6) |
| Albumin | Mouse | 12 (0.4) | 3 (4.7) | 3 (9.4) | − | − |
| Alpha-1-antitrypsin | Mouse | 13 (0.4) | 7 (0.3) | 1 (16.8) | − | − |
| Elongation factor Tu | Y. pestis | 14 (0.4) | − | − | 6 (2.4) | 14 (0.4) |
| Actin | Mouse | 15 (0.3) | 15 (0.1) | 199 (0.01) | − | − |
| 60 kDa chaperonin GroEL | Y. pestis | 16 (0.3) | − | − | 24 (0.6) | 16 (0.3) |
| Pregnancy zone protein | Mouse | 17 (0.3) | 11 (0.2) | 2 (13.9) | − | − |
| Serotransferrin | Mouse | 18 (0.3) | 6 (0.4) | 4 (5.7) | − | − |
| Plasminogen activator Pla | Y. pestis | 19 (0.3) | − | − | 13 (1.0) | 19 (0.3) |
| Transcription activator Mbf2 | X. cheopis | 20 (0.3) | − | 11 (2.0) | − | − |
| Creatine kinase | X. cheopis | 21 (0.3) | − | 39 (0.2) | − | − |
| Secreted protein precursor | X. cheopis | 22 (0.3) | − | 45 (0.02) | − | − |
| Glyceraldehyde-3-phosphate dehydrogenase | Y. pestis | 23 (0.3) | − | − | 11 (1.0) | 23 (0.3) |
| ATP synthase beta subunit | Y. pestis | 24 (0.3) | − | − | 32 (0.5) | 24 (0.3) |
| Erythrocyte protein 4.1 | Mouse | 25 (0.3) | 45 (0.02) | ND | − | − |
−, not applicable; ND, not detected. Complete descriptions of the proteins and their accession numbers are available in the online proteomics data set.
Approximately 1,900 proteins were identified from biofilm aggregates dissected from fleas infected using mouse serum, again most of them present in only trace amounts. In this sample type, 62% were Y. pestis proteins, 28% were flea proteins, and 10% were mouse proteins. Quantitatively, Y. pestis proteins constituted 54% of the total protein content, flea proteins 34%, and mouse proteins 12%. The top 25 most abundant proteins accounted for 45% of the total content (Table 2). Of these, 13 were Y. pestis proteins (22% of the total), seven were flea proteins (18%), and five were mouse serum proteins (6%). As expected, hemoglobin and blood cell proteins were not abundant, indicating that although these predominate in the biologically relevant context of an infectious whole blood meal, they are not essential components of the initial biofilm matrix in the flea gut.
TABLE 2.
The most abundant proteins identified in Y. pestis biofilm dissected from fleas 24 h after infection using mouse serum and their rank in relevant control samplesb
| Protein name | Source | Abundance rank and (percentage of total protein) in: | ||||
|---|---|---|---|---|---|---|
| Flea biofilm | Mouse blood | Flea gut contents | Y. pestis in vitro 21°C biofilm | Y. pestis in vitro 37°C | ||
| Outer membrane protein C | Y. pestis | 1 (7.4) | − | − | 16 (0.9) | 4 (2.7) |
| Trypsin-like serine proteasesa | X. cheopis | 2 (6.9) | − | 7 (4.6) | − | − |
| Mitochondrial carrier protein | X. cheopis | 3 (4.7) | − | ND | − | − |
| pH six antigen | Y. pestis | 4 (2.6) | − | − | 1557 (<0.01) | 82 (0.2) |
| Outer membrane protein A | Y. pestis | 5 (2.5) | − | − | 1 (5.2) | 6 (2.2) |
| Alpha-1-antitrypsin | Mouse | 6 (1.9) | 7 (0.3) | 1 (16.8) | − | − |
| Secreted protein precursor | X. cheopis | 7 (1.7) | − | 109 (0.03) | − | − |
| Juvenile hormone-binding protein | X. cheopis | 8 (1.7) | − | 45 (0.14) | − | − |
| Actins 1 and 2 | Mouse | 9 (1.6) | 15 (0.1) | 195 (0.01) | − | − |
| Yersinia murine toxin (Ymt) | Y. pestis | 10 (1.5) | − | − | 7 (1.3) | 22 (0.7) |
| Ail protein | Y. pestis | 11 (1.5) | − | − | 2 (3.4) | 1 (10.0) |
| Secreted protein precursor | X. cheopis | 12 (1.2) | − | 42 (0.16) | − | − |
| Palmitoyl protein thioesterase | X. cheopis | 13 (1.1) | − | ND | − | − |
| Glyceraldehyde-3-phosphate dehydrogenase | Y. pestis | 14 (1.0) | − | − | 11 (1.1) | 2 (3.4) |
| Elongation factor Tu-A | Y. pestis | 15 (1.0) | − | − | 6 (1.5) | 3 (2.7) |
| Outer membrane protein F | Y. pestis | 16 (0.9) | − | − | 4 (2.5) | 322 (0.04) |
| Outer membrane lipoprotein Lpp | Y. pestis | 17 (0.8) | − | − | 19 (0.8) | 47 (0.4) |
| Pregnancy zone protein | Mouse | 18 (0.8) | 11 (0.2) | 2 (13.9) | − | − |
| Secreted protein precursor | X. cheopis | 19 (0.7) | − | 82 (0.06) | − | − |
| Putative exported protein | Y. pestis | 20 (0.7) | − | − | 178 (0.1) | 89 (0.2) |
| Albumin | Mouse | 21 (0.6) | 3 (4.7) | 3 (9.4) | − | − |
| Serine protease inhibitor A3K | Mouse | 22 (0.6) | 9 (0.3) | 5 (5.3) | − | − |
| 60 kDa chaperonin GroEL | Y. pestis | 23 (0.6) | − | − | 24 (0.6) | 11 (1.3) |
| Putative lipoprotein | Y. pestis | 24 (0.5) | − | − | 149 (0.1) | 105 (0.2) |
| Alkyl hydroperoxide reductase | Y. pestis | 25 (0.5) | − | − | 18 (0.8) | 14 (1.0) |
Cumulative amount of six different trypsin-like serine proteases. A complete description of the proteins and their accession numbers are available in the online proteomics data set.
−, not applicable; ND, not detected.
Mouse blood proteins in the initial EPS associated with the biofilm aggregates in the flea are listed separately in Table 3. Fourteen of these were common to aggregates collected from fleas infected using either whole blood or serum, and most of these biofilm-associated proteins were also abundant in mouse blood. Thus, the majority of mouse proteins in the initial EPS of the biofilm appear simply to be abundant proteins in the blood, and so are incorporated into biofilm within the flea gut.
TABLE 3.
The 25 most abundant mouse proteins in Y. pestis biofilm aggregates dissected from fleas 24 h after infections using mouse blood (Biofilm-blood) or mouse serum (Biofilm-serum) and in the midgut contents of fleas 24 h after feeding on sterile mouse blood (Flea gut blood meal)a
| Biofim-blood | Biofilm-serum | Flea gut blood meal |
|---|---|---|
| Beta-globin | Alpha-1-antitrypsin | Alpha-1-antitrypsin |
| Alpha-globin | Actin | Pregnancy zone protein |
| Anion exchange protein | Pregnancy zone protein | Albumin |
| Aquaporin-1 | Albumin | Serotransferrin |
| Albumin | Serine protease inhibitor | Serine protease inhibitor |
| Alpha-1 antitrypsin | C-reactive protein | Complement C3 |
| Actin | Phosphatidylinositol-glycan-specific phospholipase D | Carboxylesterase |
| Pregnancy zone protein | Phospholipid transfer protein | Beta-globin |
| Serotransferrin | Murinoglobulin | Alpha-globin |
| Erythrocyte protein 4.1 | Serotransferrin | Murinoglobulin |
| Platelet glycoprotein 4 | Histone | Antithrombin-III |
| Erythrocyte band 7 | Carboxylesterase | Hemopexin |
| Fibrinogen | Transthyretin | Transthyretin |
| Serine protease inhibitor | LPS-binding protein | Alpha-1B-glycoprotein |
| CD147 (Basigin) | Beta-globin | Fibrinogen |
| CD47 antigen | Alpha-1B-glycoprotein | Ig gamma-1 chain |
| Blood group Rh(D) protein | Aquaporin-1 | Alpha-2-antiplasmin |
| Ammonium transporter Rh type A | Ammonium transporter Rh type A | Retinol-binding protein |
| Murinoglobulin | Complement C4 | Carbonic anhydrase |
| Carbonic anhydrase | Apolipoprotein B-100 [ | Adiponectin |
| Phospholipid transfer protein | Annexin A2 | H-2 class 1 histocompatibility antigen |
| Histone | Ig mu chain C | Ig mu chain C |
| Monocarboxylate transporter | Platelet glycoprotein 4 | Plasma protease C1 inhibitor |
| Carboxylesterase | Alpha-globin | Glutathione peroxidase 3 |
| Serine incorporator | Hemopexin | Complement factor H |
Proteins common to all three samples are highlighted in gray.
The bacterial proteomic profiles of aggregated Y. pestis in the flea gut 24 h after infection using either whole blood or serum included outer membrane proteins OmpA, OmpC, OmpF, and Ail as well as the Yersinia murine toxin Ymt, a phospholipase D (PLD) and the fimbrial pH 6 antigen (Table S1). Metabolic proteins were also prominent, indicating that the bacteria were actively multiplying in the flea. Most of these proteins were also prominent in Y. pestis harvested from 24 h in vitro biofilm growth at 21°C (the flea gut temperature) or from 37°C planktonic cultures (representing the physiological state of the bacteria added to the blood or serum that the fleas ingested). The pH 6 antigen was more abundant in the in vivo samples, probably reflecting the lower pH of the flea gut environment (18).
Lipidomics profile of the aggregates
A previous study suggested that the extracellular matrix of Y. pestis biofilm in the flea contains lipids (9). Untargeted lipidomics analysis of biofilm from fleas infected using mouse serum was performed to determine the relative abundance of different lipid components (Table 4). For comparison, the lipid profiles of sterile mouse serum, uninfected flea gut contents, and Y. pestis grown at 21°C in vitro were also determined. The most striking finding was the high concentration of cholesterol. It constituted more than a third of total matrix lipids, more than twice its relative concentration in mouse serum. This finding is consistent with the abundant cholesterol crystals within biofilm aggregates that were observed in fleas (Fig. 3D through F). Sphingolipids were also present in higher concentrations in in vivo biofilms than in any other contributing source.
TABLE 4.
Lipid composition (% of total lipid) of Y. pestis biofilm dissected from fleas 24 h after infection using mouse serum compared to control samples
| Lipid class | Flea biofilm (serum) | Mouse serum | Flea gut contents | Y. pestis in vitro 21°C |
|---|---|---|---|---|
| Phospholipids (total) | 33.2 | 34.6 | 78.9 | 99.1 |
| Phosphatidylcholine | 7.9 | 28.3 | 4.7 | 2.0 |
| Phosphatidylethanolamine | 14.3 | 6.1 | 48.5 | 63.2 |
| Phosphatidylglycerol | 2.3 | 0 | 4.6 | 13 |
| Phosphatidylinositol | 3.3 | 2.2 | 0 | 0 |
| Phosphatidylserine | 0.4 | 0 | 11.7 | 3.9 |
| Phosphatidic acids | 0 | 0 | 0.1 | 0.1 |
| Cardiolipin | 4.9 | 0 | 9.3 | 16.8 |
| Glycerolipids (total) | 25.1 | 47.9 | 11.8 | 0 |
| Diglycerides | 2.5 | 0.2 | 3.2 | 0 |
| Triglycerides | 22.6 | 47.7 | 8.6 | 0 |
| Sterol lipids (total) | 37.1 | 15.8 | 3.3 | 0 |
| Cholesterol | 33.8 | 2.0 | 2.6 | 0 |
| Cholesterol ester | 3.2 | 15.1 | 3.3 | 0 |
| Sphingolipids (total) | 3.7 | 0.1 | 0.2 | 0 |
| Sphingomyelin | 1.7 | 0 | 0 | 0 |
| Ceramide | 2.9 | 0 | 0 | 0 |
| Fatty acids (total) | 0.6 | 1.6 | 5.7 | 0.9 |
eDNA is not required for aggregation
To determine whether eDNA is important in the Y. pestis biofilm EPS, 0 or 1,000 U/mL of DNase I was added to blood meals containing Y. pestis used to infect fleas. DNase did not affect biofilm formation in the flea at 24 h post-infection. The percentage of fleas with a proventricular biofilm-like aggregate was the same whether or not DNase was present in the infectious bloodmeal (6 of 10 fleas positive from each group), and the overall mass and appearance of the aggregates were similar. DNase added to the bloodmeal was still active after 24 h in the flea, as midgut contents collected from treated fleas were able to degrade DNA ex vivo. Furthermore, some intrinsic DNase activity was detectable in flea midgut contents 24 h after an untreated blood meal. Thus, although eDNA can be a component of the initial extracellular matrix of Y. pestis biofilm in the flea (Fig. 3A through C), it does not appear to be essential.
Ymt enhances Y. pestis aggregation in the flea
The rapid aggregation of Y. pestis in the flea digestive tract depended on bacterial activity, because proventricular clots were not observed in fleas fed sterile mouse blood or serum. To determine whether live bacteria were required, we next fed fleas with whole blood or serum containing filtered lysates of Y. pestis (Fig. 4). Ymt was required for normal infection rate and incidence of proventricular aggregation (Fig. 4A through C). Surprisingly, filtered material from either French-pressed lysates or heat-killed Y. pestis was sufficient to generate proventricular aggregates in fleas, which appeared identical to those generated by live bacteria (Fig. 4D through F). This indicated that one or more soluble factors of Y. pestis, and not metabolically active bacteria, were required to generate aggregates that localized to the proventriculus.
Fig 4.
Infection status of fleas after feeding on mouse blood containing filtered lysates, live cells, or combinations of live cells plus filtered lysates of Y. pestis wild-type (WT), Ymt mutant (ymt– and ymt– EV), complemented Ymt mutant (ymt– pCH16), or wild-type and Ymt + E. coli (pCH16). Percentages of fleas with proventricular aggregates 24 h after infection (A); the infection rates 24 h after infection (B); and the bacterial load distributions of fleas immediately after infection (T0; black bars) and 24 h after infection (T24; gray bars) for the infections indicated (C) are shown. The mean, SEM, and range of three independent infections are indicated, except for infections with the live Y. pestis Ymt– complemented mutant and Ymt– empty vector (EV) control, for which only one experiment was performed. Typical proventricular aggregates of fleas 24 h after infection with wild-type Y. pestis filtered lysate in mouse blood (D) or mouse serum (E) or with Ymt+ E. coli live cells (F) are shown. *, significant difference from Y. pestis wild type; **, significant difference from cognate Ymt+ sample (one-way ANOVA and Tukey’s multiple comparisons tests; or t-test for comparison of paired sample types).
It was shown previously that the Y. pestis PLD enzyme Ymt is required for the survival of Gram-negative bacteria in fleas infected using mouse blood. A Y. pestis mutant encoding an inactive Ymt protein, and E. coli, do not survive in the flea digestive tract, but they survive normally when complemented with pCH16, a plasmid containing a wild-type copy of the Y. pestis ymt gene (19). More recently, Ymt has also been implicated in the rapid proventicular localization of live Y. pestis (2, 20). To further examine the role of Ymt in rapid aggregation, we fed fleas mouse blood containing live cells or filtered lysates of a Ymt Y. pestis mutant, the complemented mutant, and a Ymt + E. coli (pCH16) strain. Only ~5% of fleas infected with the Y. pestis Ymt mutant showed evidence of proventricular aggregates and only ~30% of fleas were still infected at 24 h (Fig. 4C) compared to the 90%–100% rates of fleas infected with wild-type bacteria (Fig. 2C). By contrast, fleas infected with the complemented Y. pestis mutant or Ymt+ E. coli strains had a similar incidence of proventricular aggregates as wild-type Y. pestis. These same results could be recapitulated when fleas were fed sterile filtered lysates of the strains (Fig. 4A). Furthermore, fleas infected with a mixture of Ymt mutant live cells and filtered lysate of the parent wild-type strain showed wild-type levels of proventricular aggregation at 24 h. This did not occur in fleas infected with Ymt mutant live cells and filtered mutant lysate (Fig. 4A). In sum, these results indicate that Ymt PLD activity, host blood serum, and one or more bacterial components are sufficient to rapidly produce normal levels of a colloidal, gel-like aggregate in the flea digestive tract that becomes ensnared in the proventriculus.
Ymt-dependent survival of Y. pestis in the flea gut is much more pronounced when fleas feed on infected mouse blood than with infected rat blood (20). We therefore also examined the role of Ymt on the aggregation phenotype in fleas that fed on rat blood containing live bacteria or their corresponding filtered lysates (Fig. 5). Unlike the case with mouse blood infections, the Ymt mutant bacteria formed proventricular aggregates at an intermediate level (~60% incidence). This correlated with the results that, as previously reported (20), the mutant’s infection rate and mutant bacterial load per flea were much higher in fleas infected with rat blood than when mouse blood was used (Fig. 4 and 5). By contrast, however, filtered lysates of the Ymt mutant in rat blood generated little proventricular aggregation, similar to what was seen for mouse blood (Fig. 5A and 4A).
Fig 5.
Infection status of fleas after feeding on rat blood containing filtered lysates or live cells of Y. pestis wild-type (WT), Ymt mutant (ymt–), or complemented Ymt mutant (ymt– pCH16). Percentages of fleas with proventricular aggregates 24 h after infection (A) and the overall infection rates 24 h after infection (B) are shown. The mean, SEM, and range of three independent infections are indicated, except for infections with the Y. pestis Ymt– complemented mutant, for which only one experiment was performed. *, significant difference from Y. pestis wild-type filtrate (one-way ANOVA and Tukey’s multiple comparisons test).
DISCUSSION
Transmission of Y. pestis by fleas depends on the presence of bacterial aggregates in the proventriculus that partially or completely interfere with normal blood-feeding hydrodynamics, which results in the regurgitation of contaminated biomaterial into the bite site during feeding attempts. This transmission mechanism is essentially based on the ability of Y. pestis to form a biofilm in the flea. Biofilm formation is a common phenomenon of many bacteria in which they grow as a dense microcolony embedded within a complex extracellular matrix of EPS, often attached to a surface. In this study, we examined the initial stage of Y. pestis biofilm formation in the flea, in particular the protein and lipid composition of the EPS, and identified a bacterial factor that is important for its production and for the rapid aggregation of Y. pestis after it is ingested by fleas.
The predominant conceptual model of biofilm development is that planktonic bacteria first adhere to a surface, begin to produce extracellular matrix components, and multiply to form a dense surface-attached community that is embedded within the matrix that also incorporates environmental components (16, 21). It has recently become clear, however, that not all bacterial biofilm development follows this model (22). In some cases, bacteria autoaggregate without attachment to a surface and then produce the EPS, resulting in suspended biofilm aggregates. In other cases, there is neither initial surface attachment nor cell-cell-dependent autoaggregation. Aggregate formation occurs passively when physicochemical forces cause the bacteria to concentrate by a process called depletion aggregation, also known as depletion attraction or “macromolecular crowding” (23–27). This process is akin to phase separation and occurs spontaneously in environments containing high concentrations of non-adsorbing polymers or other macromolecules and bacteria, in which the exclusion of the macromolecules from the spaces between neighboring bacteria forces them together (23–27). Host-derived polymers present at tissue infection sites can induce depletion aggregation in vitro, suggesting that it would also occur in vivo (26, 28).
Our results and the rapidity with which aggregation occurs suggest a model in which Y. pestis biofilm formation in the flea begins with a depletion aggregation phenomenon. Most bacteria within the EPS in the flea gut are not in contact with each other and no direct adhesion to the proventricular surface is seen, indicating that the initial steps of biofilm formation are not mediated by autoaggregation via bacterial cell-surface adhesins. Foregut infections likely result from the viscous aggregates first simply becoming enmeshed among the proventricular spines. Thus, rather than a consequence of a biofilm development pathway per se, the initial aggregation step in the flea gut likely is the result of a spontaneous colloidal-like transition. The Y. pestis Hms-dependent exopolysaccharide produced during the subsequent stage of biofilm maturation may function as a bridging autoagglutinin, however. The bacteria are much more densely and cohesively packed together later after infection, as seen in blocked fleas (15).
The EPS of the Y. pestis biofilm in the flea is complex, incorporating substantial amounts of protein derived from the digestion of the blood meal. In the normal context, this includes erythrocyte and thrombocyte proteins. These do not appear to be essential, however, because aggregation occurred after fleas fed on infected serum. The EPS also contains large amounts of lipids, with a significant sequestration of cholesterol. The importance of the Ymt PLD in inducing aggregation suggests that phospholipids are critical. Notably, Ymt was one of the most abundant bacterial proteins detected in the aggregates, and intact bacteria are not even necessary, as aggregated material associated with the proventriculus is seen following blood meals containing filtered lysates of Ymt+ Y. pestis. The Y. pestis outer membrane protein Ail, a known adhesin, was also abundant and may bind blood proteins in the matrix. The detection of large amounts of cholesterol and other blood lipids in the aggregates, their viscous and semi-fluid nature, and the fact that they are not disrupted by proteases and stain with certain lipid-soluble dyes (9) all indicate that blood lipids are important constituents of the EPS. Except cholesterol, specific blood components were not noticeably enriched in the matrix. The EPS largely mirrored abundant protein and lipid components present in the flea’s blood meal. Insect hemocytes and coagulation factors can aggregate and encapsulate bacteria that invade the hemocoel, an aspect of insect innate immunity (29). However, hemocytes are not present in the flea digestive tract, and no insect immunity proteins were detected in the EPS; thus, it is highly unlikely that the rapid aggregation of Y. pestis in the midgut results from a flea immune response.
The original role described for Ymt in the flea was that it protected Y. pestis from a toxic byproduct of blood digestion (19). More recently, the importance of Ymt was shown to vary depending on the host blood source, with Ymt-negative Y. pestis surviving much better in the flea gut following an infected rat blood meal than an infected mouse blood meal (20). Ymt has also been implicated as being important for the initial proventricular aggregation (2, 20). Our results confirm and extend those findings. Ymt appears to have two roles in the flea—it protects Y. pestis from a toxic byproduct of blood meal digestion and it is essential for the initial aggregation that leads to transmission. The two attributes may be linked, as proposed previously that the rapid aggregation within a lipid-rich EPS may be responsible for survival in the flea gut by shielding the bacteria from exposure to a toxic byproduct of red blood cell digestion (20).
Although it greatly enhanced bacterial survival and aggregation in the flea, Ymt activity was not absolutely essential. Proventricular aggregates were seen (rarely) following infection with the Ymt mutant in mouse blood, and with greater frequency following infections using rat blood (although lower than seen for Ymt+ Y. pestis). Fleas infected with Y. pestis in rat blood are susceptible to an exaggerated form of initial aggregation that extends into the esophagus, a phenomenon that correlates with the relative insolubility and crystallization of rat hemoglobin in the flea gut (3, 20). Thus, cholesterol and insoluble hemoglobin protein also likely contribute to depletion aggregation in the flea gut. We hypothesize that anionic phosphatidic acids produced by Ymt degradation of serum phospholipids are involved in driving the rapid depletion aggregation phenomenon in the flea gut. As phosphatidic acids were not detected in the EPS associated with the aggregates, their enhancement of depletion aggregation may be related to their negative charge. Another unresolved issue is that recombinant Ymt 61 kDa protein produced and purified from E. coli, although it has in vitro PLD activity (30), did not protect Ymt– Y. pestis when added to an infectious blood meal (19); and unlike native Ymt purified from Y. pestis by affinity chromatography, is nontoxic to mice (31, 32). Original Ymt fractions purified from Y. pestis consisted of two forms of 120 and 240 kDa, and it has been suggested that synergism between Ymt and other Y. pestis factors is responsible for murine toxicity (33, 34). The same may be true for the role of Ymt in the flea.
In the normal flea-borne route of transmission, Y. pestis enters the mammalian dermis in a flea-specific phenotype (35, 36) and is likely associated with EPS. Since most of the EPS surrounding the bacteria is composed of host-derived blood proteins and lipids, the initial infectious material may be seen as “self” by the host. This may delay the innate immune response, allowing time for Y. pestis to multiply at the dermal infection site (5) and to express virulence factors such as the Type 3-secretion system that are induced upon the shift to mammalian body temperature. A related phenomenon has been reported for Leishmania, which synthesizes a mucin-like glycoprotein matrix that attaches to and blocks the foregut of their sandfly vectors. Like Y. pestis, Leishmania is transmitted by regurgitation from their blocked vectors during blood feeding. The gel-like glycoprotein matrix egested along with the parasites, which may include sandfly-derived components, has been shown to enhance infection in the mammal following transmission (37, 38).
MATERIALS AND METHODS
Bacterial strains
Y. pestis KIM6+, which lacks the virulence plasmid pCD1, and KIM6+ ymtH188N, which synthesizes a mutated Ymt that lacks phospholipase D (PLD) activity (19, 30) were used. KIM6+ ymtH188N was complemented with the recombinant plasmid pCH16, which contains a wild-type copy of ymt (31), or transformed with pACYC177 (the parent plasmid of pCH16) to serve as an empty vector control for the complemented mutant. E. coli K-12 and E. coli (pCH16) were also used. E. coli and wild type and Ymt mutant KIM6+ contained pAcGFP1 (Clontech/Takara Bio) for constitutive green fluorescence; the complemented Ymt mutant and E. coli (pCH16) contained a modified pAcGFP1 in which the ampicillin resistance gene was replaced by a kanamycin resistance gene (Supplemental Methods).
Flea infection and examination
Xenopsylla cheopis from a colony maintained at the Rocky Mountain Laboratories were infected using an artificial feeding system (39). Bacteria grown overnight in brain-heart infusion (BHI) broth at 37°C were centrifuged, resuspended in 1 mL phosphate-buffered saline, pH 7.4 (PBS) and added to 5 mL of fresh heparinized mouse blood, defibrinated rat blood (BioIVT, NY), or to mouse serum, washed mouse red blood cells, or artificial plasma (prepared as described in Supplemental Methods) to a bacterial density of approximately 5 × 108 CFU/mL. Fleas (starved for 4 days from their last maintenance mouse blood meal) were allowed to feed for 1 h and then examined microscopically; those that had fed to repletion (fully expanded midgut and abdomen) were maintained at 21°C and 75% relative humidity. Samples of 12 to 20 infected fleas (approximately equal numbers of males and females) were routinely collected immediately after the infectious blood meal (T = 0) and stored at −80°C for later enumeration of bacteria per flea. Two other samples of 18 to 20 mixed-sex fleas were collected 24 h after infection (T = 24) for dissection or enumeration of bacteria per flea. Wet mounts of dissected digestive tracts were examined by differential interference contrast (DIC) and fluorescence microscopy for localization of bacterial aggregates to the proventriculus (PV+; Fig. 1D through G). The bacterial load in fleas at T = 0 and T = 24 h after infection was determined by trituration and plating of individual fleas as previously described (39).
In other experiments, fleas were fed lysates of Y. pestis grown as described above. Bacteria were concentrated 100-fold in PBS and passed twice through a French pressure cell at 15,000 psi or placed at 60°C for 30 min. The lysates were then centrifuged at 15,000 × g for 15 min, passed through a 0.22-µm filter, and normalized to 1 mg/mL total protein content. For flea feeding, lysates were added to mouse blood or serum at a 1 to 6 ratio, an amount approximately corresponding to the number of cells added to a standard blood meal. In another experimental series, fleas were infected with Y. pestis KIM6+ in mouse blood containing 0, 200, or 1,000 U/mL of DNase I.
Immunostaining and confocal microscopy
Individual bacterial aggregates were collected from the digestive tracts of fleas 24 h after infection with the aid of a dissecting microscope and fine-tip forceps (39) and placed in a drop of blocking buffer (5% normal goat serum) within a humidified chamber for 30 min. Immunostaining of the aggregates with anti-Y. pestis, -mouse erythrocyte, and -mouse serum antibodies, and with DAPI; and confocal microscopy are described in the Supplemental Methods.
Transmission electron microscopy
Midguts, with the proventriculus attached, were dissected from infected fleas, placed in 2% paraformaldehyde, and 2.5% glutaraldehyde in Sorenson’s phosphate buffer, and stored at 4°C for 24 h. Post-fixation, staining, and examination by electron microscopy are described in the Supplemental Methods.
Protein identification by mass spectrometry
Y. pestis aggregates dissected from the posterior end of the proventriculus (Fig. 1) from ~150 fleas were pooled in 40 µL PBS containing Halt protease inhibitor (Thermo Scientific) and kept on ice. Samples were prepared and electrophoresed on a 12% SDS-PAGE gel using standard procedures (Supplemental Methods). Full lanes were excised from the gel, cut into sections, and analyzed by tandem mass spectrometry (LC-MS/MS) to identify individual proteins and their relative quantities as described in the Supplemental Methods. Comparative control samples of (i) 10 µL of sterile heparinized mouse blood; (ii) the midgut contents of 30 fleas that had fed on sterile heparinized mouse blood 24 h prior to dissection; (iii) Y. pestis in vitro biofilm growth scraped from the wells of a six-well polystyrene plate after 24 h incubation at 21°C as described (40); and (iv) planktonic Y. pestis cells harvested from a 5 mL culture of the same medium after 24-h incubation at 37°C were prepared and analyzed as described above.
Lipid profiling
Bacterial aggregates from ~200 infected fleas were collected 24 h after infection and pooled in 50 µL PBS as described above for protein identification. Lipid extraction and identification and quantitation of individual lipids in the aggregate pools by mass spectroscopy are described in the Supplemental Methods.
ACKNOWLEDGMENTS
This research was supported by the Intramural Research Program of the NIH, NIAID.
We thank J. Shannon, D. Bland, and C. Bosio for the manuscript review.
Contributor Information
B. Joseph Hinnebusch, Email: jhinnebusch@niaid.nih.gov.
Matthew Parsek, University of Washington, Seattle, Washington, USA.
ETHICS APPROVAL
Fleas were from a colony maintained according to a protocol approved by the Animal Care and Use Committee of the NIH, NIAID Rocky Mountain Laboratories. Animal work adhered to NIH guidelines and was conducted by certified staff.
DATA AVAILABILITY
The proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository (41) with the data set identifier PXD046759. The lipidomics data set is available in figshare (https://doi.org/10.6084/m9.figshare.24659454.v1).
SUPPLEMENTAL MATERIAL
The following material is available online at https://doi.org/10.1128/mbio.00124-24.
Supplemental methods.
Abundant Y. pestis proteins in flea biofilm aggregates at 24 h after infection.
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.
REFERENCES
- 1. Hinnebusch BJ, Jarrett CO, Bland DM. 2017. “Fleaing” the plague: adaptations of Yersinia pestis to its insect vector that lead to transmission. Annu Rev Microbiol 71:215–232. doi: 10.1146/annurev-micro-090816-093521 [DOI] [PubMed] [Google Scholar]
- 2. Dewitte A, Bouvenot T, Pierre F, Ricard I, Pradel E, Barois N, Hujeux A, Bontemps-Gallo S, Sebbane F. 2020. A refined model of how Yersinia pestis produces a transmissible infection in its flea vector. PLoS Pathog 16:e1008440. doi: 10.1371/journal.ppat.1008440 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bland DM, Jarrett CO, Bosio CF, Hinnebusch BJ. 2018. Infectious blood source alters early foregut infection and regurgitative transmission of Yersinia pestis by rodent fleas. PLoS Pathog 14:e1006859. doi: 10.1371/journal.ppat.1006859 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Eisen RJ, Lowell JL, Montenieri JA, Bearden SW, Gage KL. 2007. Temporal dynamics of early-phase transmission of Yersinia pestis by unblocked fleas: secondary infectious feeds prolong efficient transmission by Oropsylla montana (siphonaptera: ceratophyllidae). J Med Entomol 44:672–677. doi: 10.1603/0022-2585(2007)44[672:tdoeto]2.0.co;2 [DOI] [PubMed] [Google Scholar]
- 5. Bosio CF, Jarrett CO, Scott DP, Fintzi J, Hinnebusch BJ. 2020. Comparison of the transmission efficiency and plague progression dynamics associated with two mechanisms by which fleas transmit Yersinia pestis. PLoS Pathog 16:e1009092. doi: 10.1371/journal.ppat.1009092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Mitchell CL, Schwarzer AR, Miarinjara A, Jarrett CO, Luis AD, Hinnebusch BJ. 2022. A role for early-phase transmission in the enzootic maintenance of plague. PLoS Pathog 18:e1010996. doi: 10.1371/journal.ppat.1010996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Vetter SM, Eisen RJ, Schotthoefer AM, Montenieri JA, Holmes JL, Bobrov AG, Bearden SW, Perry RD, Gage KL. 2010. Biofilm formation is not required for early-phase transmission of Yersinia pestis. Microbiology (Reading) 156:2216–2225. doi: 10.1099/mic.0.037952-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Hinnebusch BJ, Perry RD, Schwan TG. 1996. Role of the Yersinia pestis hemin storage (hms) locus in the transmission of plague by fleas. Science 273:367–370. doi: 10.1126/science.273.5273.367 [DOI] [PubMed] [Google Scholar]
- 9. Jarrett CO, Deak E, Isherwood KE, Oyston PC, Fischer ER, Whitney AR, Kobayashi SD, DeLeo FR, Hinnebusch BJ. 2004. Transmission of Yersinia pestis from an infectious biofilm in the flea vector . J Infect Dis 190:783–792. doi: 10.1086/422695 [DOI] [PubMed] [Google Scholar]
- 10. Erickson DL, Jarrett CO, Callison JA, Fischer ER, Hinnebusch BJ. 2008. Loss of a biofilm-inhibiting glycosyl hydrolase during the emergence of Yersinia pestis. J Bacteriol 190:8163–8170. doi: 10.1128/JB.01181-08 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bobrov AG, Kirillina O, Forman S, Mack D, Perry RD. 2008. Insights into Yersinia pestis biofilm development: topology and co-interaction of hms inner membrane proteins involved in exopolysaccharide production. Environ Microbiol 10:1419–1432. doi: 10.1111/j.1462-2920.2007.01554.x [DOI] [PubMed] [Google Scholar]
- 12. Bacot AW, Martin CJ. 1914. Observations on the mechanism of the transmission of plague by fleas. J Hygiene Plague 13:423–439. [PMC free article] [PubMed] [Google Scholar]
- 13. Cavanaugh DC. 1971. Specific effect of temperature upon transmission of the plague bacillus by the oriental rat flea, Xenopsylla cheopis. Am J Trop Med Hyg 20:264–273. doi: 10.4269/ajtmh.1971.20.264 [DOI] [PubMed] [Google Scholar]
- 14. Sodeinde OA, Subrahmanyam YV, Stark K, Quan T, Bao Y, Goguen JD. 1992. A surface protease and the invasive character of plague. Science 258:1004–1007. doi: 10.1126/science.1439793 [DOI] [PubMed] [Google Scholar]
- 15. Hinnebusch BJ, Fischer ER, Schwan TG. 1998. Evaluation of the role of the Yersinia pestis plasminogen activator and other plasmid-encoded factors in temperature-dependent blockage of the flea. J Infect Dis 178:1406–1415. doi: 10.1086/314456 [DOI] [PubMed] [Google Scholar]
- 16. Flemming H-C, van Hullebusch ED, Neu TR, Nielsen PH, Seviour T, Stoodley P, Wingender J, Wuertz S. 2023. The biofilm matrix: multitasking in a shared space. Nat Rev Microbiol 21:70–86. doi: 10.1038/s41579-022-00791-0 [DOI] [PubMed] [Google Scholar]
- 17. Loomis CR, Shipley GG, Small DM. 1979. The phase behavior of hydrated cholesterol. J Lipid Res 20:525–535. [PubMed] [Google Scholar]
- 18. Wigglesworth VB. 1972. The principles of insect physiology. Chapman & Hall, Dordrecht. [Google Scholar]
- 19. Hinnebusch BJ, Rudolph AE, Cherepanov P, Dixon JE, Schwan TG, Forsberg A. 2002. Role of Yersinia murine toxin in survival of Yersinia pestis in the midgut of the flea vector. Science 296:733–735. doi: 10.1126/science.1069972 [DOI] [PubMed] [Google Scholar]
- 20. Bland DM, Miarinjara A, Bosio CF, Calarco J, Hinnebusch BJ. 2021. Acquisition of Yersinia murine toxin enabled Yersinia pestis to expand the range of mammalian hosts that sustain flea-borne plague. PLoS Pathog 17:e1009995. doi: 10.1371/journal.ppat.1009995 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Stoodley P, Sauer K, Davies DG, Costerton JW. 2002. Biofilms as complex differentiated communities. Annu Rev Microbiol 56:187–209. doi: 10.1146/annurev.micro.56.012302.160705 [DOI] [PubMed] [Google Scholar]
- 22. Sauer K, Stoodley P, Goeres DM, Hall-Stoodley L, Burmølle M, Stewart PS, Bjarnsholt T. 2022. The biofilm life cycle: expanding the conceptual model of biofilm formation. Nat Rev Microbiol 20:608–620. doi: 10.1038/s41579-022-00767-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Eboigbodin KE, Newton JRA, Routh AF, Biggs CA. 2005. Role of nonadsorbing polymers in bacterial aggregation. Langmuir 21:12315–12319. doi: 10.1021/la051740u [DOI] [PubMed] [Google Scholar]
- 24. Marenduzzo D, Finan K, Cook PR. 2006. The depletion attraction: an underappreciated force driving cellular organization. J Cell Biol 175:681–686. doi: 10.1083/jcb.200609066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Schwarz-Linek J, Dorken G, Winkler A, Wilson LG, Pham NT, French CE, Schilling T, Poon WCK. 2010. Polymer-induced phase separation in suspensions of bacteria. Europhys. Lett 89:68003. doi: 10.1209/0295-5075/89/68003 [DOI] [Google Scholar]
- 26. Secor PR, Michaels LA, Ratjen A, Jennings LK, Singh PK. 2018. Entropically driven aggregation of bacteria by host polymers promotes antibiotic tolerance in Pseudomonas aeruginosa. Proc Natl Acad Sci U S A 115:10780–10785. doi: 10.1073/pnas.1806005115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Secor PR, Michaels LA, Bublitz DC, Jennings LK, Singh PK. 2022. The depletion mechanism actuates bacterial aggregation by exopolysaccharides and determines species distribution & composition in bacterial aggregates. Front Cell Infect Microbiol 12:869736. doi: 10.3389/fcimb.2022.869736 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Preska Steinberg A, Datta SS, Naragon T, Rolando JC, Bogatyrev SR, Ismagilov RF. 2019. High-molecular-weight polymers from dietary fiber drive aggregation of particulates in the murine small intestine. Elife 8:e40387. doi: 10.7554/eLife.40387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Dushay MS. 2009. Insect hemolymph clotting. Cell Mol Life Sci 66:2643–2650. doi: 10.1007/s00018-009-0036-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Rudolph AE, Stuckey JA, Zhao Y, Matthews HR, Patton WA, Moss J, Dixon JE. 1999. Expression, characterization, and mutagenesis of the Yersinia pestis murine toxin, a phospholipase D superfamily member. J Biol Chem 274:11824–11831. doi: 10.1074/jbc.274.17.11824 [DOI] [PubMed] [Google Scholar]
- 31. Hinnebusch J, Cherepanov P, Du Y, Rudolph A, Dixon JD, Schwan T, Forsberg A. 2000. Murine toxin of Yersinia pestis shows phospholipase D activity but is not required for virulence in mice. Int J Med Microbiol 290:483–487. doi: 10.1016/S1438-4221(00)80070-3 [DOI] [PubMed] [Google Scholar]
- 32. Hinnebusch BJ. 2003. Transmission factors: Yersinia pestis genes required to infect the flea vector of plague. Adv Exp Med Biol 529:55–62. doi: 10.1007/0-306-48416-1_11 [DOI] [PubMed] [Google Scholar]
- 33. Montie TC, Ajl SJ. 1970. Nature and synthesis of murine toxins of Pasteurella pestis, p 1–37. In Montie TC, Kadis S, Ajl SJ (ed), Microbial toxins. Vol. 3. Academic Press, New York. [Google Scholar]
- 34. Walker RV. 1967. Plague toxins- a critical review. Curr Top Microbiol Immunol 41:23–42. doi: 10.1007/978-3-642-46062-3_2 [DOI] [PubMed] [Google Scholar]
- 35. Vadyvaloo V, Jarrett C, Sturdevant DE, Sebbane F, Hinnebusch BJ. 2010. Transit through the flea vector induces a pretransmission innate immunity resistance phenotype in Yersinia pestis. PLoS Pathog 6:e1000783. doi: 10.1371/journal.ppat.1000783 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Hinnebusch BJ, Sebbane F, Vadyvaloo V. 2012. Transcriptional profiling of the Yersinia pestis life cycle. In Carniel E, Hinnebusch BJ (ed), Yersinia: systems biology and control. Horizon Scientific Press, Norwich, UK. [Google Scholar]
- 37. Rogers ME, Ilg T, Nikolaev AV, Ferguson MAJ, Bates PA. 2004. Transmission of cutaneous leishmaniasis by sand flies is enhanced by regurgitation of fPPG. Nature 430:463–467. doi: 10.1038/nature02675 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Bates PA. 2007. Transmission of Leishmania metacyclic promastigotes by phlebotomine sand flies. Int J Parasitol 37:1097–1106. doi: 10.1016/j.ijpara.2007.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Bland DM, Brown LD, Jarrett CO, Hinnebusch BJ, Macaluso KR. 2017. Methods in flea research. BEI Resources. Available from: https://www.beiresources.org/Portals/2/Methods%20in%20Flea%20Research%20v2.pdf [Google Scholar]
- 40. Sun Y-C, Koumoutsi A, Jarrett C, Lawrence K, Gherardini FC, Darby C, Hinnebusch BJ. 2011. Differential control of Yersinia pestis biofilm formation in vitro and in the flea vector by two c-di-GMP diguanylate cyclases. PLoS One 6:e19267. doi: 10.1371/journal.pone.0019267 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Perez-Riverol Y, Bai J, Bandla C, García-Seisdedos D, Hewapathirana S, Kamatchinathan S, Kundu DJ, Prakash A, Frericks-Zipper A, Eisenacher M, Walzer M, Wang S, Brazma A, Vizcaíno JA. 2022. The PRIDE database resources in 2022: a hub for mass spectrometry-based proteomics evidences. Nucleic Acids Res 50:D543–D552. doi: 10.1093/nar/gkab1038 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental methods.
Abundant Y. pestis proteins in flea biofilm aggregates at 24 h after infection.
Data Availability Statement
The proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository (41) with the data set identifier PXD046759. The lipidomics data set is available in figshare (https://doi.org/10.6084/m9.figshare.24659454.v1).