The invasive pathogen Yersinia pestis disrupts host blood vasculature to spread and provoke hemorrhages.

Yersinia pestis is a powerful pathogen with a rare invasive capacity. After a flea bite, the plague bacillus can reach the bloodstream in a matter of days giving way to invade the whole organism reaching all organs and provoking disseminated hemorrhages. However, the mechanisms used by this bacterium to cross and disrupt the endothelial vascular barrier remain poorly understood. In this study, an innovative model of in vivo infection was used to focus on the interaction between Y. pestis and its host vascular system. In the draining lymph nodes and in secondary organs, bacteria provoked the porosity and disruption of blood vessels. An in vitro model of endothelial barrier showed a role in this phenotype for the pYV/pCD1 plasmid that carries a Type Three Secretion System. This work supports that the pYV/pCD1 plasmid is responsible for the powerful tissue invasiveness capacity of the plague bacillus and the hemorrhagic features of plague.

Introduction

Plague is an infection caused by the invasive Gram-negative bacillus Yersinia pestis. It is among the most dramatic bacterial diseases in human history [1]. In spite of its disappearance in most developed countries, plague still represents a significant public health problem in many regions of the world [2]. Madagascar is currently the most active plague focus worldwide, with hundreds of human cases every year and regular epidemics [3]. Since the 1990’s, the disease has also re-emerged in countries where it was thought to be extinct [4-6]. Furthermore, the discovery of natural isolates of Y. pestis carrying antibiotic-resistance plasmids [7,8] associated with the possibility of using Y. pestis as a bioweapon in the international context of terrorism is also of global concern.

Bubonic plague, the most common clinical presentation in humans and the usual form of plague in the rodent reservoir, occurs after an infectious fleabite. Studies in animal models have shown that Y. pestis may remain for various periods of time at the site of inoculation in the dermis and multiply locally. Afterwards the bacteria are drained through the lymphatic flux to the proximal lymph node where they form the pathognomonic bubo, and then to the secondary ipsilateral lymph node [9-11]. Inside the lymph nodes, Y. pestis enters through the subcapsular sinus, where the bacteria replicate and spread within the sinus. Bacteria spread into the cortex where they multiply, provoking the recruitment of numerous polymorphonuclear leukocytes. If the innate immune system of the lymph node is strong enough to contain the bacteria, the bubo suppurates and the patient recovers. However, in most instances the bacteria overpower the innate immune system of the lymph node and spread systemically [12,13]. The time elapsed between the colonization of the lymph node and the fatal outcome is very short (2.2 days on average) in the mouse experimental model [10]. Several hypotheses have been proposed for the mechanism used by Y. pestis to enter the bloodstream: (i) an invasive process by active degradation of the blood vessels of the node [11], (ii) a release of the bacteria carried by the lymphatic vessels into the blood through the subclavian vein, or (iii) a carriage of Y. pestis inside leukocytes from the lymph to the blood [14]. Once in the blood, bacteria are filtered by secondary lymphoid organs (spleen and liver), until the filtering capacity of these organs is overwhelmed, allowing the bacteria to spread and cause a severe and terminal septicemia, sometimes associated with internal and external bleedings. Massive and diffuse hemorrhages in all tissues including the lymph nodes, have been a striking feature of post-mortem pathological examination of human plague victims [13,15]. Comparable hemorrhages were also reported in wild animals that succumbed to plague [16,17]. This feature of plague pathogenesis is experimentally reproducible in various animal models [11,12,18,19]. A classically accepted view is that a Disseminated Intravascular Coagulopathy (DIC) occurs during septicemic stages and provokes these hemorrhages [18,20,21]. However, DIC usually occurs late in the pathogenesis of Gram-negative bacteria whereas hemorrhages are observable as early as day 3 post-infection with Y. pestis. Thus suggesting that Y. pestis has the ability to alter blood vessel integrity independently of coagulation defect.

Y. pestis express various virulence factors that play important roles during the invasion and colonization of its hosts. The high-pathogenicity island, or HPI, carries essential virulence genes involved in iron acquisition and is encoded on the chromosome [22]. Three plasmids carry specific virulence factors [23-24]: i) The pPla plasmid (also known as pPCP1) carries notably the plasminogen activator Pla, an adhesin with enzymatic activities capable of converting plasminogen to plasmin, thus degrading extracellular matrix and fibrin clots in vivo; ii) the pMT plasmid (also known as pFra) of Y. pestis carries the caf 1 gene that encodes for the F1 pseudocapsule, involved in the resistance to phagocytosis [25]; iii) the pYV/pCD1 plasmid carries genes necessary to synthetize a Type Three Secretion System (TTSS) that allows the injection of effectors (Yops) inside the host cell cytosol [26]. These three plasmids are associated with Y. pestis virulence. The pYV/pCD1 plasmid is shared by and essential for the virulence of the three pathogenic Yersinia species, Y. pestis, Y. pseudotuberculosis and Y. enterocolitica [27]. They might be involved in the powerful invasiveness of Y. pestis. Although the capacity of Y. pestis to alter the vascular barriers may be important during the pathological process, the interactions of the pathogen with endothelial cells remain mainly unexplored. The aim of this study was, using in vitro and in vivo models of infection, to get insight into the processes that allow Y. pestis to cross the blood vessel barrier and ultimately cause hemorrhages. The results show that Y. pestis can degrade blood vasculature within the draining lymph node in vivo, certainly causing the internal hemorrhages frequently observable. Once in the bloodstream, bacteria can also spread systematically and cross the vascular barrier from the lumen to the organs, eventually degrading the tissues in the secondary organs. The in vitro model of infection of cellular vascular barrier highlights the central role of the pYV/pCD1 plasmid in this phenomenon.

Materials and methods

Ethics statement

All animals were housed in a level 3 animal facility accredited by the French Ministry of Agriculture (accreditation B 75 15–01), and were infected in compliance with French and European regulations (EC Directive 86/609, French Law 2001–486), following the approved protocol CETEA 2014-0025/MESR 008223 by the internal Institut Pasteur ethic board. For infection of Flk-1GFP/+ mice, death of animals was never intended as outcome of the experiments and no mice died before planned sacrifice. For the LD50 measurement of the Y. pestis CO92 pFU96+ strain, death of animal was the planned outcome. Humane endpoint were defined and animals were monitored twice a day; animals appearing evidently moribund (apathetic, shivering, cold, etc.) were immediately sacrificed. EC and GM were trained by the Institut Pasteur internal training for handling and care of animals and for animal experimentation.

Bacteria and culture

The Y. pestis strains used were CO92 wild type [28] and its pYV/pCD1-cured and Δcaf derivatives [29]; 6/69 wild type and its pPla-cured derivatives [30,31]. Plasmid pFU96 [32] that confers red-fluorescence was kindly provided by P. Dersch (department of Molecular Infection Biology, Helmholtz Centre for Infection Research, Braunschweig, Germany) and was introduced into Y. pestis CO92 by electroporation. For animal experiments, bacteria were cultured at 28°C for 36h on Luria Bertani agar plates supplemented with 0.002% (w/v) hemin (LBH). For cell infection, bacteria were grown on LBH plates at 28°C for 24h, and at 37°C for another 12h. All experiments with live Y. pestis were performed in a biosafety level 3 laboratory.

Animal experiments

Flk-1GFP/+ mice [33] exhibiting a bright GFP signal in all endothelial cells due to the insertion of the gene encoding green fluorescent protein (GFP) into the VEGF receptor-2 gene locus, were kindly provided by Alexander Medvinsky (Institute for Stem Cell Research, University of Edinburgh, UK). For infection, 100 μl of 5x103 cfu/ml bacterial suspensions were injected subcutaneously into the right lateral ventral region. The LD50 of the Y. pestis CO92 pFU96+ strain was measured using six-week-old OF1 female mice (Charles River Laboratory) according to the method of Reed and Muench [34].

Cell culture & infection

Human Dermal Microvascular Endothelial Cells (Promocell) were grown in Endothelial Cell Medium MV2 with supplements (Promocell) at 37°C in 5% CO2. Five days prior to infection, cells were plated on glass coverslips in 24-well plates (for immunofluorescence experiments) or on Transwells (0.4 or 3 μm filters) coated with Type-I Collagen (R&D System) at a density of 0.25x105 cells/well. Continuous cell monolayers (≈5x105 cells/well) were infected at a Multiplicity of Infection (MOI) of 100 for 30 min to allow bacterial adhesion. After washing with PBS and addition of fresh medium to the monolayer, the incubation was continued for at least 2h. The MOI and the timing were chosen following preliminary experiments in which a serial increasing MOI was used to infect HDMEC cells (1; 10; 50; 100; 1000) for various times (30’, 60’, 90, 120’, 240’). In all conditions, the outcome was the same, only the percentage of rounding was varying depending on the MOI and time. Infection too long or with a MOI too strong resulted in detachment and/or death of the cells. MOI 100 for 150 minutes was found to be optimal for the observation of cell rounding with limited cell death/detachment, and was subsequently selected for further experiments. In a first set of permeability assays, 1 mg/ml of 4 kDa FITC-Dextran (Sigma aldrich) was added to the upper chamber of 0.4 μm Transwells covered with HDMEC that were either untreated (control) or infected for 2.5h. Every 20 min, 100 μl samples from the lower chamber were taken to measure optical density at 485 nm and 530 nm for FITC excitation and emission respectively. Fluorescence value was converted into protein mass using the fluorescence measured for a standard curve of 4 kDa FITC-Dextran. In another set of permeability assays to measure the capacity of Y. pestis to translocate through the cell monolayer, 3 μm Transwells were used. The bacteria in the upper chamber were not removed after 30 min, and 100 μl samples were taken from the lower chamber after 3.5h and 4.5h to enumerate cfu.

Bacterial filtrates were obtained from filtration (through at 0.22 μm filter) of suspensions of 5x107 bacteria cultured in cell culture medium at 28°C or 37°C for 3 hours. Bacterial sonicates were obtained from sonication of 109 bacteria cultured on Agar plate at 28°C or 37°C and resuspended in cold saline added with protease inhibitors (Roche, Complete) and PMSF (Sigma). Sonicates were filtrated through 0.22 μm filters. Cells (2x105 HDMEC) were incubated for 5 hours at 37°C with the equivalent of 5x107 bacterial filtrate or sonicate.

Confocal microscopy, histology and immunofluorescence microscopy

Right inguinal lymph nodes were recovered as draining lymph node and left inguinal lymph nodes as secondary infected organs. Lymph nodes and spleen were fixed in 4% neutral buffered paraformaldehyde (PFA) for 48h, sliced with a 200 μm-thick setting using a vibratome (Leica VT 1200S), and mounted with Vectashield mounting medium (Vector Laboratories). Slides were observed and imaged with a spinning-disc Cellvoyager CV1000 confocal system (Yokagawa) at the Imagopole of the Center for Innovation & Technological Research from the Institut Pasteur. For histopathological examination, non-infected inguinal lymph nodes were taken and fixed in 4% neutral buffered formalin for 48h, and embedded in paraffin. Four μm tissue sections were stained with hematoxylin-eosin. Histology slides were examined through an Eclipse 5Oi Nikon microscope equipped with a DSRi1 camera (Nikon).

In vitro-infected cells were fixed in 4% PFA at specified times and exposed to the following primary antibodies for 1h: mouse anti-F1 monoclonal antibody (1:500; B18-1, [35]); rabbit anti-Yersinia polyclonal antiserum (1:1000; [12]), mouse anti-human VE-Cadherin monoclonal antibody (1:100; BV9, Abcam). Afterward, they were washed 3 times with PBS and incubated with the following secondary antibodies for 1h: Alexa Fluor 488 goat anti-mouse (1:500; Invitrogen) or anti-rabbit (1:1000; Invitrogen) IgG; rhodamin-phalloidin (1:200; Invitrogen), and DAPI (1:10.000, Interchim). After three washings with PBS, stained cells were mounted in moviol and observed under a confocal microscope LSM700 upright (Zeiss) coupled to a color camera. The average number of round cells per field was determined by counting the number of cells in at least three different fields per condition for each experiment. Cell fields were chosen randomly using the DAPI channel, then the field was switched to green and the presence of holes around the cell was enumerated. All the in vitro experiments were repeated independently at least three times. Images were processed and analyzed using Photoshop (Adobe) for immunofluorescence experiments and FIJI (ImageJ) for confocal experiments. No non-linear alterations were brought to the images.

Results

Y. pestis–infected lymph node display degraded blood vessels

Hemorrhages are a classical feature of plague infection frequently observed in buboes or organs [11,12,18,19]. However, the actual mechanism involved in the disruption of blood vessels was never properly determined using classical histopathology examination of stained tissue sections. Indeed, the interactions between bacteria and the damaged blood vessels are too localized to be observed by usual histopathological sections of 5–8 μm thickness. To address this question, an in vivo model dedicated to this interaction was developed. In this model, fluorescent bacteria were used to infect mice expressing GFP-tagged vasculature. This model allowed the qualitative observation of the direct interaction between bacteria and blood vessels with minimal treatments allowing tissue quality to be preserved.

First, the plasmid pFU96 encoding the Red Fluorescent Protein (RFP) was introduced into the fully virulent Y. pestis strain CO92 [32]. The strain was monitored for the expression of the virulence factors: pYV/pCD1, pPla, pFra and the HPI locus as described in [27]. The LD50, determined according to the method of Reed and Muench [34], was 16 bacteria, showing no significant difference with the parental strain CO92.

Second, this RFP-CO92 strain was used to infect Flk1-GFP transgenic mice that express a fluorescent blood vascular system due to their Green Fluorescent Protein (GFP)-tagged VEGF receptor [33]. Fig 1A presents the green vasculature of the inguinal lymph node of a non-infected Flk1-GFP mouse (right panel). The comparison to a classically hematoxylin-eosin stained histological section of inguinal node (center panel) shows the highly vascularized medulla and the cortex of the node. In the cortex, some spots present a less dense presence of blood vessels and correspond to follicles. The capsule is clearly visible around the node.

Fig 1

Confocal microscopy, observation of the blood vessels in the draining lymph-node of non-infected and Y. pestis-infected Flk1-GFP mice.

A. Left, schematic of a lymph node (biorender.com) with legends indicating capsule, node cortex and medulla and follicles. Middle, histological section of a mouse inguinal lymph node (4 μm thick) stained with hematoxylin-eosin, Bar = 250 μm. Right, inguinal lymph node of a non-infected Flk1-GFP mouse. Image is reconstructed from 4 panels (2x2), each panel corresponding to the maximum of intensity (MIP) calculated on a 150 μm section of the bubo. Bar = 250 μm. B. Draining inguinal lymph node of a Y. pestis-infected Flk1-GFP mouse on D3 post infection. Lymph node is representative of features observed on five independent challenged mice. Image is reconstructed from 8 panels (2x4), each panel corresponding to the maximum of intensity (MIP) calculated on a 150 μm section of the bubo (15 sections separated by 10um). The red color corresponds to bacteria visible in the subcapsular sinus and deeper in the lymph node parenchyma. The blood vasculature is colored in green. Frame 1 correspond to magnification on S1 Video; frame 2 correspond to magnification on S2 Video; frame 3 correspond to panel C and S3 Video. The capsule is indicated with white arrows, the medulla with a white star, and the cortex with white crosses. Bar = 500 μm. C. The left panel shows the magnification of the central region of panel B (MIP calculated on 20μm, 40 sections separated by 0.5um). The right panel displays an interpretation of the picture: the blood vessel is discontinued in close proximity of groups of bacteria and individual cells or remnants of cells are surrounded by bacteria. Made with biorender. Bar = 10 μm.

The Flk1-GFP mice were infected subcutaneously with 500 cfu of Y. pestis RFP-CO92. RFP-CO92 infection of the Flk1-GFP mice caused bacterial loads in their spleen similar to those observed in the usual OF1 mouse model infected with the parental CO92 strain (an average of 3.104cfu/organ at 48hrs and 6.104cfu/organ at 72hrs).

The inguinal draining lymph node of the Flk1-GFP mice infected was sampled on day 3 (D3). At this stage, classic histopathological observations allow to observe important masses of bacteria in the subcapsular sinus of the draining lymph node that spread through the cortex (S1 Fig). Bacteria are surrounded by infiltrated Polynuclear neutrophils. Different studies also described clear signs of hemorrhages observable in the lymph nodes of most infected mice [11,12,18,19,22]. On D3, in the RFP-CO92-infected Flk1-GFP mice model, masses of bacteria (in red) were visible in the subcapsular sinus compartment, with bacteria infiltrating the node parenchyma (Fig 1B). These observations corresponded to the classical pattern of infection of the draining lymph node through the lymph vascular system (S1 Fig) [11,12]. In the regions of the infected-node distant from the bacteria, the green staining of the vascular system was as intense as in non-infected mice (Fig 1A, right panel). However, this fluorescence appeared weaker or missing in the areas where bacteria were visible (Fig 1B, S1 and S2 Videos). There was little or no superposition of green and red fluorescence on the same field. This phenomenon was observed in all sections examined (corresponding to the complete draining lymph nodes of five infected Flk1-GFP mice). A higher magnification of a 20 μm thick section of the parenchymal zone of the lymph node, allowing a 3D reconstruction of the zone, showed blood vessels that appeared degraded in the vicinity of bacteria (Fig 1C and S3 Video). In this field, what was interpreted as an interrupted vessel was observed next to round single cells or degraded cell remnants. The confocal microscopy allowed to investigate the fields above and below the ones presented. As opposed to the left part of the vessels (Fig 1C) that disappear out of the field (partially visible on the S3 Video), the vessel at the center is abruptly discontinued. There was no trace of the vessel in the space above or below the field presented, but instead round shapes of 10μm of diameter stained in green, strongly evocative of round endothelial cells. The absence of superposition of green and red fluorescence below the subcapsular sinus where bacteria were visible suggests that blood vessels were either repelled by the mass of bacteria, or destroyed by the invading bacteria as showed in the higher magnification. In infected draining lymph nodes at a more advanced state of infection, the red masse of bacteria eventually takes over and fill the whole bubo. Remnant of green blood vessels are scarce (S2A and S2B Fig). These results suggest that Y. pestis can disrupt blood vessels from the parenchyma of lymph nodes.

Y. pestis disseminate via the bloodstream and cross the vascular barrier

We assume that the direct degradation of blood vessels by the bacteria provokes local hemorrhages that could allow bacteria to enter the bloodstream. Once in the bloodstream, Y. pestis disseminates and colonize other lymphoid organs (spleen, liver, secondary lymph nodes) [11,26], suggesting that bacteria have the capacity to translocate from the blood vessel lumen to the parenchyma of these organs, to form secondary infectious foci.

The presence of bacteria in the blood vessels of the spleen on D3 post-infection was confirmed by confocal microscopy (Fig 2) and cfu counting (an average of 6.104cfu/organ). Contrary to the draining lymph node, colonized via the afferent lymph, where it was not possible to observe superposition of bacteria and blood vessels staining, the spleen of infected Flk-1 mice presented clear superposition of red and green (appearing yellow) demonstrating the presence of bacteria within the vessels. Bacteria were visible superposed to the blood vessels surrounding the white pulp of the spleen.

Fig 2

Confocal microscopy observation of blood vessels in the spleen of non-infected and Y. pestis-infected Flk1-GFP mice.

For all images, the red color corresponds to bacteria and the blood vasculature is colored in green. A. Non-infected Flk1-GFP mouse spleen. Left, white field picture showing the visual structure of the spleen with white pulp appearing clear grey and red pulp appearing darker. Right, organization of blood vasculature in the red pulp and around the white pulp. B. Spleen from two Y. pestis-infected Flk1-GFP mice on D3 post-infection. Clear presence of red Y. pestis bacteria in blood vasculature, particularly around the white pulp. On the left panel blood vessels around white pulp appear strongly disorganized and partially disrupted. C. Y. pestis-infected mouse on D3 post-infection. Bacteria (left panel) and blood vessels (right panel) are merged in the middle panel where bacteria fluoresce in red and blood vessels in green. Images are reconstructed from panels each corresponding to the maximum of intensity (MIP) calculated on 200 μm thick sections (A: 10 panels (2x5); B left: 24 panels (4x6); B right: 18 panels (3x6); C: 6 panels (2x3) calculated on 150 μm thick sections). Bar = 500 μm.

Similarly, in secondary infected lymph nodes (i.e. left inguinal lymph nodes that were not drained by the lymphatic flux coming from the right primary draining lymph node) dense spots of bacteria were frequently observed inside blood vessel lumens (Figs 3 and S3). These spots were observable in the cortex of the lymph node, distant from the subcapsular sinus, consistent with a penetration of the bacteria in the lymph node via the bloodstream rather than via lymphatic flux. Higher magnification confirmed the presence of red bacteria inside the blood vessel lumen (Fig 3B and S4 Video). Furthermore, at this magnification we observed bacteria in the process of leaking out of the vessel into the lymph node parenchyma, strengthening the premise that Y. pestis has the capacity to degrade the blood barrier from the lumen to penetrate into the organ. In infected lymph nodes at a more advanced state of infection, the red spot of bacteria spread from the initial blood vessel through the cortex (S2C Fig). Similarly to the draining lymph nodes, blood vessels within the cortex seem to disappear while masse of bacteria progresses. Altogether, these data show that Y. pestis colonizes secondary organs by penetrating through the blood circulation and subsequently by degrading the blood vessel barrier.

Fig 3

Confocal microscopy, observation of the blood vessels and bacteria in the non-draining lymph node of a Y. pestis-infected Flk1-GFP mouse.

A. Bacteria (left panel) and blood vessels (right panel) are merged in the middle panel. Bar = 100 μm. Image are reconstructed from 4 panels (2x2), each panel corresponding to the maximum of intensity (MIP) calculated on a 200 μm section of the bubo (20 sections separated by 10um). Frame correspond to panel B and S4 Video. B. Magnifications of a blood vessel filled with bacteria. The panel a. represent the MIP calculated on a 50 μm thick section (100 sections separated by 0.5um); panels b-c-d are successive sections of panel a. White plain arrowheads point at red bacteria within and outside the vessel, white empty arrowheads point at some holes in the vessel. Bar = 10 μm.

Y. pestis disrupts confluent monolayers of microvascular endothelial cells

An in vitro model was established to further investigate the interaction between Y. pestis and blood vessels using confluent monolayers of Human Dermal Microvascular Endothelial Cells (HDMEC). While non-infected HDMEC formed a confluent monolayer with tight junctions (Fig 4A), holes were visible at the junction between cells after 2.5h of infection with Y. pestis CO92 at an MOI of 100 (Fig 4B). These holes were observed around approximately 40% of the cells, and only around cells that were directly in contact with bacteria (Fig 4C), suggesting that a direct interaction between the cells and the bacteria was needed to induce this effect. This hypothesis was further examined by incubating HDMEC monolayers for 5h with either i) the filtrate of the supernatant of Y. pestis cultured at 28°C or 37°C for 3h or ii) with a sonicate of 107 bacteria cultured at 28°C or 37°C for 3h. No holes were observed between cells in either case (S4 Fig). This result strengthens the premise that no toxins or bacterial cytoplasmic molecules secreted outside the host cell cytoplasm were directly responsible for the observed phenotype and that a direct contact of live bacteria with the cells was necessary. The number of holes increased over time: after 8h of infection most cells were infected and started to separate from adjacent cells, and at 24h post-infection all cells were completely round (S5D Fig) and started to detach from the surface of the well.

Fig 4

Disruption of a monolayer of vascular endothelial cells by Y. pestis.

HDMEC monolayers were either uninfected (A, control), infected for 2.5h with the CO92 strain of Y. pestis at a MOI of 100 (B and C) or with a CO92 strain cured of the pYV/pCD1 plasmid (D); Panel C is a 2x magnification for better visualization. bars = 10 μm. The cells were fixed and stained to visualize their DNA (DAPI, blue), actin (phalloidin, red), and tight junctions (anti-VE-cadherin antibody, green). Bacterial cells on the monolayers were stained with an antibody directed against the F1 pseudocapsule (green) and their DNA was labeled with DAPI. Holes at the cell junctions were observed in the monolayers infected with Y. pestis (white empty arrowheads point at some holes in panels B and C). Contrast on the green panel B was enhanced to improve visualization of the holes. Bacterial DNA (DAPI) and Y. pestis were visible on the HDMEC monolayer at high magnification (white plain arrowheads).

The pYV/pCD1 plasmid is necessary to increase the permeability and cross the vascular barrier

Since a direct interaction between bacteria and cells seemed necessary for Y. pestis to disrupt the vascular endothelial cell monolayer, the role of three major virulence factors exposed at the surface of the bacillus was investigated. First, the role of the plasminogen activator Pla carried by the pPla plasmid and the F1 pseudocapsule carried on the pMT plasmid of Y. pestis were tested [25]. Both, the derivative strain of Y. pestis CO92 in which the caf operon was deleted by allelic exchange (Δcaf) [28] and the 6/69 Y. pestis strain cured of the pPla plasmid (ΔpPla) [27] retained the same ability as the parental strains to cause holes at the cell junctions (S5 Fig). Second, a pYV/pCD1-cured derivative of strain CO92 was used to infect confluent HDMEC. As shown in Fig 4D, when infected with the pYV/pCD1-cured Y. pestis, no holes were observed in the HDMEC monolayer despite the presence of numerous bacteria in contact with the cells.

To quantify this phenomenon, we measured the permeability of the vascular barrier following infection using a Transwell system (Fig 5). The non-infected monolayer showed little permeability to FITC-Dextran (<0,1ug/min/cm2), while the confluent monolayer became permeable over time after addition of Mannitol (2,3 ug/min/cm2 after 3h), a hyperosmolar osmotic molecule known to induce the opening of cellular tight junctions. The infection with Y. pestis induces an even stronger increase in monolayer permeability after 3h of infection (4,18 ug/min/cm2). The permeability of the monolayer did not increase when infected with the pYV/pCD1- derivative strain (Fig 5A and 5B). In addition, the increased permeability of the endothelial cells barrier allowed the passage of bacteria through the monolayer in a pYV/pCD1 dependent manner (Fig 5C). Altogether, these data demonstrate the essential role of the pYV/pCD1 plasmid in the increased permeability of blood vessels measured in vitro and observed in vivo during Y. pestis infection. According to these results, the ability of Y. pestis to enter and leave the blood circulation, thus invading the internal organs of its host and eventually provoking bleedings, can be explained by the deterioration of the vascular barrier carried by the functions encoded on the pYV/pCD1 plasmid.

Fig 5

Permeability of Y. pestis-infected HDMEC.

HDMEC monolayers cultured on Transwells (filter 0.4 μm) were infected (MOI = 100) for 2.5h with Y. pestis CO92 or its pYV/pCD1 cured derivative. Permeability was measured using FITC-Dex (4kDa) (A) and the diffusion speed calculated (B). Mannitol was used as positive control for junction opening. C. The crossing of bacteria was measured using Transwells of lower filter capacity (3 μm).

Discussion

In order to study the interaction between Y. pestis and its host vasculature directly in vivo, a new approach was developed using a mouse strain expressing GFP-tagged blood vessels infected with a RFP-tagged Y. pestis. This technology permitted an innovative, qualitative observation of infection and of the interaction between Y. pestis and its host at the organ level (using spinning-disc confocal microscopy) and at the cellular level (using confocal microscopy). The use of a spinning-disc confocal microscopy allowed analyzing the entire lymphoid organ (3x1.5 millimeters long and up to a millimeter of thickness) with a minimal transformation and alteration of the sample due to staining or treatment. In this study, this approach showed the previously undocumented degradation of the blood vessels by Y. pestis during infection. We propose that the degradation of blood vessels is an active mechanism caused directly by Y. pestis as observed in vitro.

An alternative explanation would be an indirect effect of Y. pestis on blood vasculature through the inflammatory response to the infection. Massive recruitment of Polynuclear Neutrophils is known to provoke an increase in vascular permeability that could lead to bleeding in some cases [36]. However, neutrophils recruitment does not induce a complete physical degradation of the vessels themselves, such as the one observed in the infected lymph nodes. Additionally, an inflammatory process would have a distant effect, nevertheless, Fig 1 shows that only blood vessels in close proximity with bacteria seem to disappear, suggesting that it is the direct contact of Y. pestis with endothelial cells that provokes the degradation of the vessels, as seen in vitro, and thus excluding a major role of the inflammatory process in this phenomenon.

This point is of importance since the mechanism used to enter the bloodstream remains controversial. Previous work by Nahm and colleagues showed that the variable timing between the subcutaneous infection of Y. pestis and the outcome is due to the time to reach the lymph node [10]. Once in the draining lymph node, Y. pestis quickly spread to other internal organs, the infection is rapid and often fatal. Thus suggesting that lymph nodes are the entry door to blood circulation. The main argument against the hypothesis for a direct entry of the bacillus into the bloodstream was the absence of proof that Y. pestis enters in contact with vessels and could cross the vascular barrier. This work demonstrates that both events take place in vivo in the bubo. Though it does not invalidate other hypotheses, the data presented strongly support for an active and direct entry of Y. pestis, in a pYV/pCD1 dependent manner, into the blood circulation and explains the invasiveness of this pathogen. The increased blood vessel permeability allows the crossing of bacteria, which could also explain systemic hemorrhages associated with plague. Past studies suggested that hemorrhages due to Y. pestis were the consequences of a DIC provoked by the toxicity of the bacterial LPS [18,20]. However, this work suggests a more direct and faster effect of the bacteria on blood vasculature. Y. pestis can breach the blood barrier in vitro and in vivo inducing an increased permeability that could provoke internal bleedings.

Among the virulence factors carried by Y. pestis, the pYV/pCD1 plasmid is one genetic element crucial to the pathogenicity [25]. This ≈70 kb plasmid encodes a TTSS that mediates the injection of molecular effectors, the Yop proteins, directly into the cell cytoplasm. Inside the cells, these Yops disorganize the actin cytoskeleton (thus preventing phagocytosis), inhibit signaling cascades involved in the early innate response, and induce apoptosis of the target cell [26,37,38]. We propose that the pYV/pCD1 plasmid is responsible for the disruption of blood vessels, probably through a direct action of the Yop effectors injected into the cytoplasm of endothelial cells. The fact that neither the bacterial sonicate nor the bacterial supernatant can produce the rounding phenotype observed during infection suggests that the Yop effectors themselves cannot act from outside the eukaryotic cells. We assume that the injection into the cytoplasm is necessary for inducing the rounding. The YopT and YopE effectors are good candidates as they have been demonstrated to have a rounding capacity on Hela cells [39,40,41]. The TTSS has been showed to be crucial in Y. pestis pathogenicity and a deletion mutant of pYV/pCD1 is virtually non-pathogenic. In particular, a pYV/pCD1- strain injected by IV route is almost non-pathogenic. We propose that the pYV/pCD1 plasmid encoding the TTSS is responsible for the impressive invasiveness of Y. pestis and for its hemorrhagic features.

Magnification 1 of confocal microscopy observation of blood vessels in the bubo of Y. pestis-infected Flk1-GFP mice.

Draining inguinal lymph node of a Y. pestis-infected Flk1-GFP mouse on D3 post infection. Magnification of the frame 1 from Fig 1A. Video shows 15 successive panels separated by 10um (from top to bottom) used to reconstruct the MIP of Fig 1A. The red color corresponds to bacteria visible in the subcapsular sinus and deeper in the lymph node parenchyma. The blood vasculature is colored in green. There is little to no superposition of the staining.

(AVI)

Click here for additional data file.

Magnification 2 of confocal microscopy observation of blood vessels in the bubo of Y. pestis-infected Flk1-GFP mice.

Draining inguinal lymph node of a Y. pestis-infected Flk1-GFP mouse on D3 post infection. Magnification of the frame 2 from Fig 1A. Video shows 15 successive panels separated by 10um (from top to bottom) used to reconstruct the MIP of Fig 1A. The red color corresponds to bacteria visible in the subcapsular sinus and deeper in the lymph node parenchyma. The blood vasculature is colored in green. There is little to no superposition of the staining.

(AVI)

Click here for additional data file.

Magnification 3 of confocal microscopy observation of blood vessels in the bubo of Y. pestis-infected Flk1-GFP mice.

Draining inguinal lymph node of a Y. pestis-infected Flk1-GFP mouse on D3 post infection. Magnification of the frame 3 from Fig 1A. Video shows 26 successive panels separated by 0.5um (from top to bottom) used to reconstruct the MIP of Fig 1C. The red color corresponds to bacteria and blood vasculature is colored in green. Individual bacteria are identifiable. The continuous blood vessel in green on the right of the image appears degraded and single cells are visible on the left next to bacteria.

(AVI)

Click here for additional data file.

Magnification 4 of confocal microscopy observation of blood vessels in the secondary lymph node of Y. pestis-infected Flk1-GFP mice.

Non-draining inguinal lymph node of a Y. pestis-infected Flk1-GFP mouse on D3 post infection. Magnification of the frame from Fig 2A. Video is a 3D reconstruct corresponding to the MIP showed on Fig 2B.a. The red color corresponds to bacteria and blood vasculature is colored in green. The continuous blood vessel in the middle of the reconstruction appears full of red bacteria. Superposition of staining green and red (appearing yellow) is observable. Individual bacteria are identifiable inside and leaking outside the vessel.

(AVI)

Click here for additional data file.

Histopathological observation of an Y. pestis-infected draining lymph node of ORF1 mice.

Histological section of a mouse inguinal lymph node (4 μm thick) stained with hematoxylin-eosin. ORF1 mouse was infected subcutaneously with 500cfu CO92 Y. pestis strain for 76h prior to sacrifice. A. Full section of the lymph node. Large infiltrate of Y. pestis, colored in pink, are visible at the periphery of the lymph node (black arrowheads), progressing from the lymphatic sinus to the cortex. Edemas and large engorged blood vessels (erythrocytes are colored in red) are visible within the lymph node. Infiltrate of Polynuclear neutrophils (PMN) with the characteristic horseshoe shaped nuclei colored in blue/purple are visible in the cortex. Bar = 500 μm B. Higher magnification of the square B displaying a “sea” of bacteria (black arrowheads, black frame surrounding pink areas) and PMNs infiltrates around the bacteria (yellow five-branch stars). Hemorrhages are visible in the tissue surrounding the lymph node (six-branch red stars). C. Higher magnification of the square C displaying a “sea” of bacteria (black arrowheads), PMNs infiltrates (yellow five-branch stars) and enlarged engorged blood vessels (red). A hemorrhage is visible in the cortex of the lymph node (six-branch red stars). Bacteria are visible are close proximity of the blood vessel. Blood vessels are tampered, but not fully degraded.

(TIF)

Click here for additional data file.

Confocal microscopy observation of blood vessels in advanced draining lymph node and secondary lymph node of Y. pestis-infected Flk1-GFP mice.

Inguinal draining lymph nodes and secondary lymph node of Y. pestis-infected mice displaying an advanced state of degradation on D3 post-infection. The images are reconstructed from panels corresponding to the maximum of intensity (MIP) calculated on 150–200 um sections. Red and green panels are merged. A. and B. are two consecutive slices of the same draining lymph node. Red bacteria are filling the entire volume of the cortex, all vasculature within the cortex seems destroyed. C. is a secondary lymph node infected through blood circulation. Dense spots of bacteria spread through the cortex from the initial entry point. Blood vessels within the cortex have progressively disappeared on the side of numerous bacterial areas (top half) compared to the other side (bottom half).

(TIF)

Click here for additional data file.

Confocal microscopy observation of blood vessels in the secondary lymph node of additional Y. pestis-infected Flk1-GFP mice.

Secondary lymph nodes of 5 Y. pestis-infected mice on D3 post-infection. The images are reconstructed from panels corresponding to the maximum of intensity (MIP) calculated on 150–200 μm sections. Bacteria are tagged with RFP and appeared as red spots co-localizing with GFP-tagged blood vessels (arrowheads). Red and green panels are merged. The average size of a lymph node is 1 to 2 mm.

(TIF)

Click here for additional data file.

Monolayer of HDMEC cells treated with Y. pestis sonicate and filtrate.

HDMEC monolayers were incubated for 5h with the equivalent of 107 bacteria Y. pestis culture medium (filtrate) or 107 bacteria sonicate (see Materials and Methods section). No holes in the tight junctions are observed due to these treatments. DNA was stained with DAPI (blue; panels a,e,i). Actin was stained with Phalloidin (Red; panels b,f,j). Tight junctions were targeted with an anti-VE-cadherin antibody (green, panels c,g,k).(d,h,l) are merged panels. Bar = 10 μm.

(TIF)

Click here for additional data file.

Infection of a monolayer of HDMEC cells with various Y. pestis derivatives.

HDMEC monolayers were infected (MOI = 100) for 2.5h with Y. pestis 6/69 wild type (A), its derivative cured of pPla (B), and Y. pestis CO92 Δcaf (C). D. HDMEC monolayers were infected (MOI = 100) for 24h with Y. pestis CO92. DNA was stained with DAPI (blue; panels a,e,i,m). Actin was stained with Phalloidin (Red; panels b,f,j,n). Staining was done as described in the legend of Fig 4. Examples of bacteria in close contact with the cells are indicated with white plain arrowheads. Bar = 10 μm.

(TIF)

Click here for additional data file.

23 Jul 2021

Dear Dr. Mikaty,

Thank you very much for submitting your manuscript "The invasive pathogen Yersinia pestis disrupts host blood vasculature to spread and provoke hemorrhages." for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

I have received the reviews for your manuscript. The reviewers have evaluated your manuscript and find it valuable to the field. However, they raise a few important questions. Additional data is needed to resolve them. A fully revised manuscript will be evaluated before publication in PLOS NTD.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

R. Manjunatha Kini

Associate Editor

PLOS Neglected Tropical Diseases

Fabiano Oliveira

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

I have received the reviews for your manuscript. The reviewers have evaluated your manuscript and find it valuable to the field. However, they raise a few important questions. Additional data is needed to resolve them. A fully revised manuscript will be evaluated before publication in PLOS NTD.

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: See my comments below

Reviewer #2: Methods were appropriate to ask the relevant questions.

Reviewer #3: Methods are appropriately described.

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: See my comments below

Reviewer #2: The results are presented clearly and concisely, and Figures are of high quality.

Reviewer #3: Results are clearly described.

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: Major Concerns:

1) The authors’ main conclusion in this paper is that Y. pestis, via a T3SS-dependent mechanism, actively damages the blood vessel endothelium, thus increasing vessel permeability, and this is how the bacteria enter the bloodstream and disseminate. I do not feel that the data the authors present here allow them to rule out the other 2 potential mechanisms of dissemination to the blood described by the authors in the introduction (movement of bacteria through the lymph and eventually into bloodstream, or carriage of bacteria into the blood by phagocytes). I can see how damage to the vessels would cause leakage out of the vessels, but that does not necessarily mean bacteria would enter those damaged vessels.

2) At the time point chosen by the authors for their in vivo experiments, there are really two different phenomena being studied regarding Y. pestis dissemination and it is a bit difficult to separate the two. The first relates to how the bacteria are entering the bloodstream. The second relates to how these disseminated bacteria move from the blood into the tissue parenchyma. For the first, since the bacteria would initially be found in the subcapsular sinus area early in infection, the bacteria would need to gain access to the LN parenchyma to come in direct contact with blood vessels to increase their permeability. The authors do not address how the bacteria might be getting to the parenchyma initially. Because of the later time point (d3), the blood vessel damage the authors see in the dLN could have been caused from the inside out by the bacteria already spreading systemically, rather than the bacteria that arrived in the LN via afferent lymph from the injection site. For the second, it’s easier to see how the bacteria already in the bloodstream could damage the vessels and leak out into the parenchyma and cause the hemorrhages the authors discuss. I feel that this study would benefit from the addition of earlier time points for the in vivo work. Being able to show the early stages of blood vessel damage, or at least bacteria in contact with vessels and able to inject Type III effectors, prior to dissemination of the bacteria throughout the body, would greatly strengthen the support for the authors’ conclusions.

3) The authors do not address the potential role of inflammatory cells, particularly neutrophils, could be playing in the blood vessel damage they observe. At the later time point they focus on, there would potentially be large numbers of neutrophils recruited to the draining LN and these cells would likely colocalize with areas of high bacterial density, and therefore, areas of blood vessel damage. The possible role of the inflammatory response in the observed phenomena should at least be discussed, if not assessed experimentally.

4) For the in vitro experiments, the authors conclude that the Yops must be secreted into the HDMEC cells by the T3SS. This is probably a reasonable assumption, but methods are available for confirming the secretion of Yops into host cell cytosol. Determining the specific Yop or Yops responsible would also be fairly straightforward.

Reviewer #2: The authors do a good job of contextualizing the findings and highlighting the importance to understanding bubonic plague. The conclusions are largely supported by the data. There are some inherent limitations to the microscopy, even though it is the most striking and compelling aspect of the manuscript. For instance, it is suggested that disruption of blood vessels is visualized- and that is likely the case, though it is only visual analysis and there is no conclusive evidence that it is what we are seeing (ie. control with agent that causes disruption).

Reviewer #3: The conclusions are not fully supported by the data. Some experiments are missing that are required for the conclusions made, or alternative hypotheses needed to be included.

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: Line 264-277: This section refers to a supplementary figure. It seems like the figure could easily be included as regular figure in the text.

Line 274: The authors refer to “secondary organs”, but I think they can specify spleen here.

Figure 2B: I think this figure would benefit from the addition of arrows pointing to the phenomena being described.

Line 308: The authors refer to “holes in cells”, but the holes seem to be gaps in the monolayer between cells at the junctions. Thus, it is not clear how the “holes were observed in approximately 40% of the cells” was calculated

Figure 4A: It’s unclear why the Y-axis is in milligrams. The methods implied that the fluorescence of FITC-dextran was what was being measured. Is the fluorescence measurement being converted to mg? If so, shouldn’t it be expressed as a concentration instead of just total mg?

Some suggestions for improving the text:

Line 33: change to “…remain poorly understood.”

Line 38: change to “…plasmid is responsible for…”

Line 45: “It is among the most important bacterial…”

Line 59: change to “Y. pestis can multiply within lymph nodes. If the innate immune response in the lymph node is strong…”

Line 74: “…have been a striking feature…”

Line 84: “…factors that play…”

Line 89: “…activities capable of converting plasminogen to plasmin, thus degrading extracellular matrix and fibrin clots in vivo.”

Line 94: “…associated with and essential for the virulence…”

Line 98: “…may be important during…”

Line 200: delete the sentence starting “This RFP CO92…”. Redundant

Line 211: “…shows the highly vascularized medulla…”

Line 271: “…and the classical cfu…”

Line 272: “Contrary to the draining lymph node, colonized via the afferent lymph,”

Line 285: “…at this magnification we observed bacteria in the process of leaking out of the vessel into the lymph node parenchyma…”

Line 314: “…cells in either case…”

Line 315: “no toxins or bacterial…”

Line 316: “…host cell cytoplasm…”

Line 387: “…increased blood vessel permeability allows the crossing of bacteria, which could also explain systemic hemorrhages associated with plague.”

Line 392: “Y. pestis can breach…permeability that could provoke…”

Line 402: “…supernatant can produce…”

Reviewer #2: Minor concerns/comments:

Of note, pCD1 is the more common nomenclature for the Y. pestis plasmid as opposed to pYV which is typically used to denote the plasmind from Y. pseudotuberculosis or Y. enterocolitica.

The sentence beginning in 38 should likely read "pYV plasmid bears responsibility" or pYV plasmid is responsible"

Line 73- may want to slightly expand on what "bleedings" refers to specificlly (ie. internal hemorrhage?)

Lines 77-79- it is a little confusing as to how DIC (which is increased coagulation) could be responsible for hemhorage, may want to slightly clarify

Line 94- specify "inside the host cell cytosol"

Lines 94-96 state that the three plasmids are essential to virulence of the pathogenic Yersinia, but this is only true of Yersinia pestis- pCD1/pYV is the only plasmid of the 3 found in all three pathogenic Yersinia.

Introduction may benefit from 1 or 2 sentences summarizing data or conculsions.

Line 187 italicize Y. pestis.

Lines 190-191 should read "mechanisms involved....were never properly determined" or "mechanism involved...was never properly determined)

The statement in line 201 "retained the same virulence" does not have any confirmative data except for the LD50, and should perhaps read "retained the same LD50" (unless bacterial burden and survival is to be included in the supplement)- obviously a very minor concern if at all.

Figure 1- Capsule appears to point to a follicle on the GFP panel, which may be confusing

It is difficult to determine if the images shown in Figure 1C are truly showing degraded cell remnants of vessels, as is suggested. Though this reviewer is not sure if it is possible, a control with an agent that disrupts the vasculature in some way or a marker of vascular damage would strengthen the argument.

On Page 12, it is indicated that bacteria are entering secondary infected lymph nodes through only the vasculature. Though likely, it is difficult to confirm that Figure 2B is showing "bacteria leaking to from the vessel into the parenchyma", it is not clear that those bacteria in the parenchyma originated in the vasculature vs entering the node in another manner (ie lymphatic system) and degrading vessels. This question isn't asked specifically (different timing, differential labeling, etc). Therefore language in this section with that regard should perhaps be tempered.

The authors do not comment on whether or not they see any disruption of the vasulature in the non-draining lymph nodes in which the bacteria have assumedly arrived by blood vessels. Is it proposed that the disruption of blood vessels strictly occurs from "outside" of the vessel?

Supplementary Figure S2 is fairly striking, and might be included in the main text.

In the discussion, it may be useful to comment on what it might mean if the vessel destruction is exclusive to bacteria that have arrived in the lymph nodes via the lymphatic system in the draining lymph node vs the destruction seen (or not seen) in the non-draining lymph node in which the bacteria have entered presumably via the vascular system.

Reviewer #3: See minor comments below.

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: In their manuscript “The invasive pathogen Yersinia pestis disrupts host blood vasculature to spread and provoke hemorrhages” Mikaty et al. use in vivo and in vitro approaches to examine the effects of Y. pestis infection on the blood vessel endothelium. For the in vivo work, the authors used a mouse strain expressing GFP in blood vessel endothelial cells, combined with red-fluorescent Y. pestis, to image bacteria and vessels in fixed draining lymph nodes, non-draining lymph nodes, and spleen. For the in vitro work, the authors look at the effects of Y. pestis on the permeability of human endothelial cell monolayers.

Overall, the work is original and important, and the results are potentially very interesting. However, I have concerns about some of the authors’ conclusions that I feel need to be addressed. I also feel that the manuscript could use quite a bit of editing and I will suggest just some of the changes that could be made.

Reviewer #2: In this study, the authors utilize both in vivo and in vitro analysis to describe the pCD1-mediated degradation/destruction of the blood vessels in the primary draining lymph node after infection with Y. pestis. Using in vivo microscopy, they show presence of bacteria consistent with lymphatic vessel-mediate arrival in the draining lymph node, and clear disruption of blood vessels in the localized areas around the clusters of bacteria. The lack of co-localization between bacteria and vasculature in the draining lymph node is striking, as is the contrast between this finding and what is seen in the non-draining lymph node. It is proposed that this disruption/destruction of the vasculature is responsible for entry of the bacteria into the bloodstream and subsequent colonization of other lymphoid organs. The authors are able to show this in non-draining lymph nodes as well as the spleen. In the non-draining lymph nodes the authors see bacteria associated with the vasculature as well as within the parenchyma of the lymph nodes.

Overall, the manuscript is well-written and the data clearly presented. The microscopy is beautiful, and the authors do a nice job of piecing together the events that occur upon entry of Y. pestis into the draining lymph node, and the events subsequent to this critical step. This work most certainly contributes to our understanding of how Y. pestis disseminates within the body. The strengths of the manuscript are in the innovative microscopic technique, as well as in the comparison of colocalization of bacterial and vasculature within draining lymph nodes vs. non-draining lymph nodes and other organs. A weakness is that the manuscript is primarily compelling microscopy, which limits some of the conclusions. For instance, though the it is indicated that vessel destruction is observed as green blots within the staining, it can't be known for sure that this is what we are seeing, and though it is assumed that bacteria have leaked from he blood vasculature into the parenchyma in the non-draining lymph nodes, this can't be confirmed for certain. It also might have been informative to have a sort of time course to observe draining and non-draining lymph nodes simultaneously at different time points (ie. earlier timepoints where we see small numbers of bacteria arriving at the draining lymph nodes vs. presumably no bacteria in the non-draining) to confirm the model proposed by the authors. Any weaknesses are offset by the fact that the microscopy in particular gives us clear insight into how Y. pestis behaves within the lymph nodes, and the authors do a great job of piecing together a narrative as to what is happening. AS a result, I consider the findings/observations to be highly significant to the field, and the manuscript paints a very compelling story, or at least lays the ground work for testing a model of dissemination.

Reviewer #3: Summary

Yersinia pestis is a highly invasive lymphotropic pathogen that must gain access to the blood in order to be effectively transmitted from the infected mammalian host to its insect vector. However, the mechanism(s) used by the bacterium to breach the lymph node to colonize the bloodstream are not clearly defined. Here Mikaty et al. describe a genetically modified mouse model, which expresses GFP primarily in endothelial cells, to visualize blood vasculature in the lymph nodes during bubonic plague. The authors use confocal microscopy to characterize the blood vessel architecture in the draining lymph node during Y. pestis infection, showing for the first time that blood vessels near Y. pestis appear to be damaged, possibly resulting in hemorrhaging. The authors also use primary human endothelial cells in cell culture to show that infection with Y. pestis results in what appears to be loss of tight junctions, and suggests that similar effects could occur in vivo, leading to vessel damage and access to the blood stream by the bacterium. Finally, the authors demonstrate that the absence of the pCD1 virulence plasmid inhibits the ability of Y. pestis to induce morphological changes in endothelial cells and permeability of cell monolayers in vitro. This study introduces an exciting new in vivo model to help us to better understand the mechanisms used by Y. pestis to breach the lymph node and colonize the blood.

Comments:

This authors provide us with a clear characterization of the Flk-1GFP/+ mouse model in the context of bubonic plague, and implicates the Ysc T3SS and Yop effectors in causing endothelial damage and dissemination. While I think this model provides investigators with a powerful tool to study vascular damage during plague, the manuscript lacks clear mechanistic studies to justify some of the conclusions that the authors make in the manuscript. These conclusions could be significantly strengthened by additional experiments and/or a more detailed discussion of the results and data from previous published studies.

1. It is not surprising that Y. pestis infection of endothelial cells in vitro results in cell rounding and death. There are ample reports that Y. pestis will inject effectors into almost any cell in monoculture, but cell-type specific targeting is observed in vivo that demonstrates that immune cells are preferentially targeted (independently demonstrated by Marketon, Mecsas, and Pechous labs). Therefore, these monocultures studies do not demonstrate that Y. pestis is specifically targeting endothelial cells in vivo. It is still possible that the immune response to the bacteria is leading to vascular leakage/damage in vivo that is required for immune cell recruitment, which is part of the formation of the bubo, and responsible for the localized vascular damage observed. This possibility is not properly discussed. In vivo targeting of endothelial cells in the lymph by Y. pestis using B-lactamase reporters would provide better support for the authors’ hypothesis. Without it, I am not sure intoxication vs. immune response hypotheses can be differentiated.

2. The MOI used in the cell culture studies were extremely high. Justification for these numbers should be included. Better yet, a dose response might be more effective approach.

3. Are the endothelial cells used in the in vitro studies polarized? Does this matter in the context of infection? Are the cells that round up targeted/injected by the T3SS?

4. Is the response by the endothelial cells due to the action of the Yop effectors or a host response to the infection (e.g., inflammasome activation and proptosis)? Infection with a strain expression the T3SS needle and lacking the effectors would help to understand the molecular mechanisms responsible for the phenotypes observed.

5. Fig. 3: Not sure why both VE-cadherin and bacteria were both stained green. Is the increased colocalization of the green and red channels in 3B due to increased VE-cadherin at the cell membrane or are these bacteria? If bacteria, why different from 3D? The actin staining in CO92 pYV- (3D) also appears to be very different from all other samples. Is this a fair observation, or an artifact of the representative image shown?

Minor comments:

1. Line 29: Not sure what is meant by “a powerful pathogen”. Suggested rewording.

2. Yop effectors are not considered classical toxins. Suggest changing the use of “toxins” with “effectors”.

3. Line 94: Only pCD1/pYV is conserved in all three Yersinia species, pPla and pMT are not in the enteric pathogens. Please revise tis sentence.

4. The reader would benefit from including the structural references to Fig. 1B to orient them to the LN structure similar to shown in Fig. 1A.

5. Line 273: I think the “lymphatic door” needs to clearly described – not sure this term is going to be widely understood.

6. Line 277: I think your conclusions would be better supported if you include the co-localization channel. Yellow color does not show up well on these images.

7. Line 349: You should include a description of the results of your positive control (mannitol) first then the Y. pestis data.

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

21 Sep 2021

Submitted filename: Reviewer_responses.docx

Click here for additional data file.

22 Sep 2021

Dear Dr. Mikaty,

We are pleased to inform you that your manuscript 'The invasive pathogen Yersinia pestis disrupts host blood vasculature to spread and provoke hemorrhages.' has been provisionally accepted for publication in PLOS Neglected Tropical Diseases.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

R. Manjunatha Kini

Associate Editor

PLOS Neglected Tropical Diseases

Fabiano Oliveira

Deputy Editor

PLOS Neglected Tropical Diseases

***********************************************************

The authors have revised the manuscript keeping all the comments made by three reviewers. I also understand that the authors' rebuttal regarding the molecular mechanism studies as they are part of their ongoing study.

I am satisfied with their responses and the revisions.

30 Sep 2021

Dear Dr. Mikaty,

We are delighted to inform you that your manuscript, "The invasive pathogen Yersinia pestis disrupts host blood vasculature to spread and provoke hemorrhages.," has been formally accepted for publication in PLOS Neglected Tropical Diseases.

We have now passed your article onto the PLOS Production Department who will complete the rest of the publication process. All authors will receive a confirmation email upon publication.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any scientific or type-setting errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript. Note: Proofs for Front Matter articles (Editorial, Viewpoint, Symposium, Review, etc...) are generated on a different schedule and may not be made available as quickly.

Soon after your final files are uploaded, the early version of your manuscript will be published online unless you opted out of this process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Shaden Kamhawi

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Paul Brindley

co-Editor-in-Chief

PLOS Neglected Tropical Diseases