Fbrio 02 1240698

TYPE Original Research
PUBLISHED 22 September 2023

DOI 10.3389/fbrio.2023.1240698
Live attenuated vaccines and

OPEN ACCESS layered defense strategies to
EDITED BY
Myron Christodoulides,
University of Southampton,
combat infections caused by
United Kingdom
REVIEWED BY
nonencapsulated Yersinia pestis
Emanuelle Mamroud,
Israel Institute for Biological Research
(IIBR), Israel Sergei S. Biryukov 1, Christopher P. Klimko 1,
Daniel Alford Powell,
University of Arizona, United States
Jennifer L. Dankmeyer 1, Ronald G. Toothman 1,
Bradley D. Jones, Jennifer L. Shoe 1, Melissa Hunter 1, Nathaniel O. Rill 1,
The University of Iowa, United States
Yuli Talyansky 1, Michael L. Davies 1, Ju Qiu 2, David P. Fetterer 2,
*CORRESPONDENCE
Christopher K. Cote Joel A. Bozue 1, Susan L. Welkos 1 and Christopher K. Cote 1*
christopher.k.cote.civ@health.mil
1
Bacteriology Division, United States Army Medical Research Institute of Infectious Diseases,
RECEIVED 15 June 2023 Frederick, MD, United States, 2 Regulated Research Administration, Biostatistics Division, United States
ACCEPTED 31 July 2023 Army Medical Research Institute of Infectious Diseases, Frederick, MD, United States
PUBLISHED 22 September 2023
CITATION
Biryukov SS, Klimko CP, Dankmeyer JL,
Toothman RG, Shoe JL, Hunter M, Introduction: Plague is an ancient disease caused by Yersinia pestis, a widely
Rill NO, Talyansky Y, Davies ML, Qiu J,
Fetterer DP, Bozue JA, Welkos SL and disseminated Tier 1 pathogen that poses significant public health and biothreat
Cote CK (2023) Live attenuated risks. The rapid course and high mortality of pneumonic plague limit the efficacy
vaccines and layered defense strategies
of antibiotic treatment and mandate the need for an effective, licensed, and
to combat infections caused by
nonencapsulated Yersinia pestis. readily available vaccine. New candidate vaccines are being developed; however,
Front. Bacteriol. 2:1240698. their efficacy in nonhuman primates, optimal vaccination schedule and immune
doi: 10.3389/fbrio.2023.1240698
response, duration of protection, and breadth of coverage against various
COPYRIGHT
virulent strains are inadequately understood. In the current work, we explored
© 2023 Biryukov, Klimko, Dankmeyer,
Toothman, Shoe, Hunter, Rill, Talyansky, homologous and heterologous vaccination schemes using the sensitive BALB/c
Davies, Qiu, Fetterer, Bozue, Welkos and mouse models of bubonic and pneumonic plague challenged with Y. pestis strain
Cote. This is an open-access article
distributed under the terms of the Creative
C12. This strain, a derivative of the wild-type strain CO92, lacks the anti-
Commons Attribution License (CC BY). The phagocytic F1 capsule yet remains highly virulent. Protection against such
use, distribution or reproduction in other nonencapsulated strains has been particularly elusive.
forums is permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication in Methods: We tested the efficacy of live attenuated vaccine (LAV) derivatives of Y.
this journal is cited, in accordance with pestis CO92 or C12 with a deletion of a type 3 secretion-associated gene (DyscN)
accepted academic practice. No use,
distribution or reproduction is permitted or the pgm pigmentation locus, and they were cured of the pPst (PCP1) plasmid
which does not comply with these terms. (CO92 pgm− pPst−). The LAVs were evaluated alone or accompanied by a dose of
a protein subunit vaccine (rF1V or rV).
Results: The most protective and immunogenic vaccination scheme, as tested

under a variety of conditions in bubonic and pneumonic plague models, was
heterologous vaccination with a LAV and the recombinant rF1V or rV protein
subunit vaccine. Furthermore, in the heterologous scheme, different LAVs and
subunit vaccines could be substituted, affording flexibility in vaccine component
selection. We also evaluated a novel intervention strategy consisting of
vaccination and post-exposure antibiotic treatment. The layering of
vaccination with the LAVs and post-exposure treatment with streptomycin was
synergistic, extending the time after the Y. pestis C12 challenge when treatment
remained effective and affording a sparing of antibiotics.
Frontiers in Bacteriology 01 frontiersin.org

Biryukov et al. 10.3389/fbrio.2023.1240698
Conclusion: The current work defined effective and flexible vaccination and
treatment interventions that successfully prevented lethal infection with virulent,
nonencapsulated Y. pestis.
KEYWORDS
plague, Yersinia pestis, vaccine, nonencapsulated, mice, bubonic, pneumonic
1 Introduction et al., 2013; Sun, 2016; Verma and Tuteja, 2016). Despite their
reactogenicity and short-lived efficacy, often requiring yearly
The etiologic agent of plague, Yersinia pestis, is a globally vaccinations, some of these live attenuated vaccines (LAVs) have
distributed pathogen that is both a considerable public health risk been employed as human vaccines, limited primarily to Russia and
and a potential biothreat agent. Plague is known as a fearsome China (Stepanov et al., 1999; Feodorova and Motin, 2012).
menace of antiquity, as illustrated by the massive Black Death The only vaccine for plague previously licensed in the United
pandemic of the 14th century, with a fatality rate of more than States was the immunogenic plague vaccine USP, commonly known
30%; yet, it remains an important current zoonotic hazard (Bertherat, as the Cutter vaccine. It contained a formalin-killed suspension of
2016; Baril et al., 2019). In many countries, such as China, virulent plague bacilli. It is no longer available but had been
Madagascar, the United States of America, the Democratic administered routinely to military personnel posted in Vietnam
Republic of the Congo, and elsewhere in South America and and certain other individuals, such as field personnel working in
Africa, the foci of Y. pestis persist, causing periodic outbreaks of plague endemic areas and laboratory personnel working with Y.
disease (Baril et al., 2019; Bevins et al., 2021). This situation is pestis (Meyer, 1970). Although it was effective in reducing the
highlighted by the unusually large outbreak of bubonic and occurrence of bubonic disease, as evidenced by the low incidence
pneumonic plague in 2017 in Madagascar, where the majority of of plague in U.S. personnel situated in Vietnam, laboratory findings
cases were the highly lethal pneumonic form (Randremanana et al., suggested that this vaccine might not protect against pneumonic
2019). Furthermore, Y. pestis is also categorized as a Tier 1 select plague (Sun, 2016; Verma and Tuteja, 2016). Also, the major
agent by the US Department of Health and Human Services (Nelson protective antigen in these vaccines was fraction 1 (F1), a surface
et al., 2021), largely due to its potential to be aersolized and produce capsule composed of the Caf1 protein encoded by the caf1 operon
and acute (in some cases rapidly fatal) disease. Attempts to harness Y. on plasmid pFra (pMT1). Such vaccines do not readily protect
pestis as a bioweapon have long been described, from the early use of against genetically engineered or naturally occurring F1-negative
plague corpses in the Middle Ages to the experimental dissemination strains, which can be highly virulent despite the absence of a capsule
of plague in China by the infamous Japanese Unit 731 during World (Welkos et al., 1995; Worsham et al., 1995; Andrews et al., 1999). A
War II and possibly beyond, as documented in detail (Cowdrey, 1984; more recent human plague vaccine candidate is rF1V, a protein
Williams and Wallace, 1989; Alibek, 1999; Worsham et al., 2018). fusion of the strongly immunogenic F1 and LcrV (referred to as V),
Yersinia pestis is susceptible to antibiotics, but due to the rapid course, the low calcium response virulence protein. V is an essential anti-
nonspecific symptomology, and high mortality rate of pneumonic host virulence factor that is required for type 3 secretion system
plague, antibiotic treatment is often ineffective. Additionally, (T3SS)-mediated translocation of the toxic Yersinia outer protein
antibiotic-resistant strains have been identified previously and in effectors (Yops) into host cells. V also stimulates the production of
the relatively recent outbreaks in Madagascar (Guiyoule et al., 2001; immunosuppressive cytokines (Heath et al., 1998; Pettersson et al.,
Hinnebusch et al., 2002; Cabanel et al., 2018). Thus, plague vaccines 1999; Mueller et al., 2005). The rF1V vaccine was shown to be
are greatly needed, yet none are licensed or widely available. efficacious in mice and some, but not all, species of nonhuman
The first plague vaccine prototype was developed by Haffkine in primates (Pitt, 2004; Powell et al., 2005; Mizel et al., 2009; Quenee
1897 and consisted of killed whole cells of Y. pestis (Meyer, 1970; et al., 2011; Adamovicz and Worsham, 2012; Williamson and
Titball and Williamson, 2004; Wang et al., 2013; Sun, 2016). Such Oyston, 2013; Verma and Tuteja, 2016). Furthermore, the
killed vaccines proved to be inefficient, providing only a short interval protection afforded by rF1V against F1-negative strains of Y.
of protection against bubonic plague (Titball and Williamson, 2004; pestis relies solely on the V antigen component. Since there is
Wang et al., 2013). Live attenuated derivatives of Y. pestis were evidence for V heterogeneity within Yersinia species (Roggenkamp
subsequently developed due to their potential ability to protect et al., 1997; Anisimov et al., 2010; Miller et al., 2012; Daniel et al.,
against both bubonic and pneumonic plague, to present multiple 2019), the potential exists that naturally occurring or engineered
antigens to the immune system, and to induce both humoral and strains harboring altered V antigens could overcome rF1V-induced
cellular immune responses, since both arms of the immune response immunity (Verma and Tuteja, 2016).
appear to be critical in anti-plague vaccine strategies (Elvin and Thus, a more reliably effective plague vaccine is needed. Recent
Williamson, 2004; Smiley, 2008; Feodorova and Motin, 2012; Wang candidates have involved various platforms, e.g., attenuated and

recombinant Yersinia strains, live recombinant viral and bacterial is C12 (Worsham et al., 1995). The recombinant strains had a
vectors constructed in adenovirus, Salmonella, Yersinia mutation in the virulence-associated gene yscN or had a deletion of
pseudotuberculosis, or other microbes, defined antigen subunit the entire pgm locus and were cured of the pPst (PCP1) plasmid
vaccines, outer membrane vesicles, DNA, mRNA, and nanoparticle encoding plasminogen activator (CO92 pgm− pPst−) (Welkos et al.,
constructs, as reviewed in detail elsewhere (Feodorova and Motin, 1997; Welkos et al., 2002; Swietnicki et al., 2011; Bozue et al., 2012;
2012; Wang et al., 2013; Sun, 2016; Verma and Tuteja, 2016; Rosario- Bozue et al., 2014). Yersinia pestis strains were grown on 5% sheep
Acevedo et al., 2021; Rosenzweig et al., 2021; Aftalion et al., 2023; Kon blood agar plates (SBAP) or tryptose blood agar (TBA) base slants
et al., 2023). for approximately 48 h at 30°C. Liquid cultures of Y. pestis C12 used
While a number of vaccine candidates have induced significant for mouse challenge studies were prepared in heart infusion broth
protection against defined wild-type strains of virulent Y. pestis in (HIB) medium supplemented with 0.2% xylose (HIBX), and the live
animal models, their optimal vaccination schedule and immune attenuated vaccine suspensions were prepared with broth cultures
response, duration of protection, and breadth of coverage, as well as incubated in HIBX supplemented with 2.5 mM CaCl2, as described
efficacy in nonhuman primates and human vaccinees, are below (Cote et al., 2021). A solution of 10 mM potassium
inadequately understood. The vaccine efforts in our laboratory are phosphate, pH 7.3–7.4 (KPhos), was used to prepare and dilute
focused on the modification and optimization of our candidate bacterial inocula. Challenge doses were determined by serial
LAVs and defined protein subunit vaccines (Swietnicki et al., 2011; dilutions in KPhos buffer and plating on SBAP. Bacteriological
Bozue et al., 2012). The LAV candidates evaluated in this study are media were obtained from Thermo Fisher-Remel (Rockville,
highly attenuated due to a deletion in the yscN gene or of the MD, USA).
pigmentation locus (pgm) accompanied by the curing of the pPst The rF1V vaccine was described previously (Heath et al., 1998;
plasmid. The yscN gene encodes for an ATPase which provides Amemiya et al., 2009; Biryukov et al., 2021). Toll-like receptor 9
energy to the T3SS and is necessary for Yop effector translocation. (TLR9) oligonucleotide (ODN) CpG ODN 2006 (CpG2006) was
We have previously demonstrated that this mutant strain does not purchased from InvivoGen (San Diego, CA, USA) and reconstituted
secrete LcrV into culture supernatant in vitro, but a small amount of in accordance with the manufacturer’s recommendations. The
LcrV can be found associated with whole-cell extracts using purified recombinant V (rV) protein was obtained from BEI
immunoassays (Bozue et al., 2012). The pgm mutation is a 102-kb Resources (Manassas, VA, USA). Alhydrogel was sourced from
deletion of the entire pgm locus that results in iron acquisition and InvivoGen. Streptomycin for injection was obtained from X-Gen
storage deficiency (Perry et al., 1990; Pendrak and Perry, 1991; Pharmaceuticals, Inc. (Horseheads, NY, USA).
Fetherston et al., 1992). The pPst plasmid encodes the plasminogen
activator enzyme that activates host fibrinolysis and permits
bacterial systemic dissemination (Ferber and Brubaker, 1979; 2.2 Preparation of candidate vaccines
Sodeinde and Goguen, 1988; McDonough and Falkow, 1989). for injection
Here, we explored homologous and heterologous vaccination
schemes utilizing both the DyscN or pgm− pPst− mutants as The day prior to vaccination, flasks were inoculated with a
standalone or paired with the rV or rF1V subunit vaccines in the suspension of colonies from a freshly inoculated SBAP and
sensitive BALB/c mouse model of protection against aerosol incubated for 24 h at 28°C–30°C with shaking at 200 rpm. The
challenge with an F1 capsule-negative Y. pestis strain. Optimal next day, the cultures were adjusted to an OD600 of 0.1 in fresh
protection against challenge with nonencapsulated Y. pestis was medium and incubated to the OD600 determined to produce the
achieved by heterologous vaccination with a live strain and a target CFU concentration, which was 107 colony forming units
protein subunit vaccine. In addition, we initiated studies for a novel (CFU) in doses of 0.2 mL (Cote et al., 2021). To confirm the actual
intervention strategy layering vaccination and post-exposure delivered dose of bacteria, the final suspensions were diluted and
antibiotic treatment. Disease outcome improved and a delayed plated for viable counts. All plates were incubated at 28°C–30°C for
initiation of antibiotics was still protective when vaccination was approximately 48 h before counting.
integrated with post-exposure antibiotic treatment. The layering of The doses of the rF1V vaccine contained 2 µg of rF1V in the
medical countermeasures will expand the treatment options available presence or absence of 5 µg of CpG2006 (CpG) and included
for the plague and potentially allow for vaccine and/or antibiotic Alhydrogel (250 µg) in a total volume of 0.1 mL (Biryukov et al.,
dose-sparing. 2021). In the studies indicated, 2 µg of rV with or without 5 µg of
CpG was administered in a total volume of 0.1 mL.
2 Materials and methods
2.1 Bacterial strains, media, and 2.3 Animals and vaccination studies
growth conditions
The animal research was conducted under an animal use
The wild-type and mutant derivatives of Y. pestis and their protocol approved by the USAMRIID Institutional Animal Care
construction and characteristics were described previously (Cote and Use Committee (IACUC) in compliance with the Animal
et al., 2021) and in the current study. These strains included the Welfare Act, PHS Policy, and other Federal statutes and
virulent F1-negative (nonencapsulated) derivative of CO92, which regulations relating to animals and experiments involving

animals. The facility where this research was conducted is dilutions in KPhos were plated in duplicate to determine sterility.
accredited by AAALAC International and adheres to the The limit of detection (LOD) was approximately 100 CFU/mL
principles stated in the Guide for the Care and Use of Laboratory blood or 5 CFU/organ. After CFU determinations, samples were
Animals (National Research Council, 2011). Female BALB/c mice radiation-inactivated, sterility-checked, and stored at −80°C for
were obtained from Charles River (Frederick, MD, USA) and were immunological analyses.
7–10 weeks of age at the time of vaccination; all groups contained 10
mice, with exceptions specified. Mice vaccinated with a single dose
of vaccine were injected subcutaneously and exposed 4 weeks later 2.7 Humoral immune response assays
by the aerosol or subcutaneous (SC) route to a lethal dose of Y.
pestis C12. As indicated, mice vaccinated twice were administered Immunoglobulin (Ig) IgG antibody responses to the vaccines
the second dose 20–28 days after the initial vaccine dose. Sera and were determined by semiquantitative endpoint ELISA using sera
spleens were collected from a cohort of mice to assess immune from vaccinated BALB/c mice, as previously described (Biryukov
responses to the vaccines. Mice were challenged 27–29 days post et al., 2021; Cote et al., 2021). The sera were collected as terminal
final vaccination. blood collections from axillary vessels and titrated against several
capture antigens: the rF1V recombinant fusion protein, the rV
protein, and the g-radiation-inactivated whole cells of Y. pestis
2.4 Exposure of vaccinated mice to virulent
strains that were temperature-switched (TS) CO92 or TS C12.
Yersinia pestis C12
The strains had been grown at 30°C for 21 h followed by a
temperature switch to 37°C and incubation for an additional 3 h
Mice were exposed to aerosolized (pneumonic) or SC (bubonic)
to upregulate the presentation of potential antigens. The cultures
challenge doses of virulent Y. pestis C12. For the bubonic challenge,
were pelleted by centrifugation and then inactivated with g-
bacteria were harvested from TBA slants. Mice exposed by the SC
radiation; temperature-shifted antigens were designated TS.
route were inoculated with 0.2 mL volumes of the suspension in
The rF1V fusion protein vaccine construct and the rV protein
KPhos (Bozue et al., 2012; Bozue et al., 2014). The bacteria used for
were diluted in 0.1 M of carbonate buffer, pH 9.5, to a
aerosol studies were prepared by using colonies from freshly
concentration of 2 mg/mL, while inactivated Y. pestis TS CO92
inoculated TBA slants which were suspended in HIBX to an
or TS C12 whole cells were plated at a concentration of 10 mg/mL
initial OD 6 2 0 of approximately 0.01 and incubated for
on 96-well Immulon 2HB plates (Thermo Fisher, Grand Island,
approximately 24 h at 28°C–30°C. For aerosol exposures, the
NY, USA). Plates were stored at 4°C overnight, then washed and
cultures were harvested by centrifugation and suspended in HIB
blocked, and samples were processed as previously described
medium (no xylose) to the concentration yielding the number of
(Biryukov et al., 2021). Two-fold dilutions of the serum were
LD50 doses indicated in the figures. Exposure to aerosolized bacteria
made in triplicate, and the results are reported as the geometric
was accomplished as previously described (Bozue et al., 2012; Bozue
mean (GM) and geometric standard error (GSE) of the reciprocal
et al., 2014; Trevino et al., 2018). Briefly, mice were transferred to
of the highest dilution giving a mean OD of at least 0.1 ± 1 SD at
wire mesh cages and were placed in a whole-body aerosol chamber
450 nm with a reference filter (570 nm). Samples with an
within a class 3 biological safety cabinet located inside a BSL-3
antibody titer of <50 were considered negative.
laboratory. Mice were exposed to aerosols of Y. pestis strain C12
created by a three-jet Collison nebulizer. Samples were collected
from the all-glass impinger vessel and analyzed by performing CFU
2.8 Cellular immune responses
calculations to determine the inhaled dose of Y. pestis.
Four weeks following vaccination, purified splenocytes were

2.5 Treatment of vaccinated mice prepared based on a previously published protocol (Amemiya et al.,
2006; Cote et al., 2021). Briefly, spleens were excised from mice (n =
A solution of streptomycin was prepared in water for injection 5 mice per group), weighed, and disaggregated in RPMI 1640
(Corning, Manassas, VA, USA), and a dose of 20 or 40 mg/kg was medium (Thermo Fisher, Grand Island, NY, USA). Red blood
administered to mice by the intraperitoneal route every 6 h for 5 cells in the spleen homogenate were lysed with Ammonium-
days. Mice were injected with the antibiotic beginning 48 h or 60 h Chloride-Potassium (ACK) Lysing Buffer (BioWhittaker,
after exposure to aerosolized Y. pestis C12. Walkersville, MD, USA) after the extract was diluted with RPMI
1640 medium and cells were pelleted by centrifugation at 335×g
for 10 min. Splenocytes were then resuspended in CTL-
2.6 Bacteriology Medium (Cellular Technology Limited, Cleveland, OH, USA)
supplemented with 1% L-glutamine and the cells were counted.
The tissues collected from necropsied mice included lung, Purified splenocytes were restimulated for ~48 h in the presence of
spleen, and blood. They were weighed and homogenized with rF1V (25 mg/mL) or irradiated Y. pestis TS C12 (5 mg/mL) in
disposable PRECISION ™ homogenizers (Covidien, Dublin, complete medium containing 10% heat-inactivated fetal calf serum
Republic of Ireland), and the CFU of the homogenate was (Thermo Fisher), 1 mM of sodium pyruvate, 0.1 mM of
determined on SBA plates. Undiluted homogenate and 10-fold nonessential amino acids, 100 U/mL of penicillin, 100 mg/mL of

streptomycin, and 50 mM of 2-mercaptoethanol. A solution of software for each stimulation condition and the medium-
phorbol 12-myristate 13-acetate (PMA; 100 ng/mL) and only control.
ionomycin (0.5 mg/mL) was used as the positive control stimulant
and resulted in uniformly strong signals (data not shown).
Following restimulation, the plates were centrifuged at 1,200×g
2.9 Statistical analyses
for 10 min at RT, and the cytokine expression in the supernatants
was evaluated by Luminex Mag Pix 36-plex mouse panel as per
Statistical analyses were performed using SAS version 9.4,
manufacturer’s directions (Thermo Fisher Scientific, Grand Island,
except as indicated. Survival curves of the vaccinated and control
NY, USA).
mice were estimated with the Kaplan–Meier method using
In addition, 3 days post-challenge, the spleen and lung
GraphPad Prism 9.0 and were compared statistically using the
homogenates were prepared by removing and homogenizing the
log-rank test. Significant differences in survival rates on days 7 and
whole organs in KPhos using disposable PRECISION ™
21 after the virulent challenge were determined using the Fisher
homogenizers (Covidien, Dublin, Republic of Ireland). Samples
exact test. The time-to-mortality (TTM) values were expressed as
were irradiated with approximately 21 kGy of g-radiation and
the median and interquartile range and compared with the log-rank
confirmed sterile by testing 10% of the sample before use.
test. The runs of aerosol-challenged mice were analyzed separately
Aliquots of the homogenates were saved for cytokine/chemokine
or combined, as indicated. The viable counts of Y. pestis from organ
determination and stored at −70°C. Cytokine and chemokine levels
and blood samples were compared using the Wilcoxon rank sum
were assayed in spleen and lung homogenates, as previously
test for pairwise comparisons test. The CFU was log-transformed
described (Trevino et al., 2018). The homogenate samples were
and the mean results were summarized as GM and geometric
thawed and centrifuged at 10,000×g for 10 min, and the supernatant
standard deviation (GSD) values, using the LOD/SQRT(2) to
was then examined for cytokine expression by Luminex Mag Pix
replace the values with no recoverable CFU.
36-plex mouse panel (Thermo Fisher Scientific, Grand Island, NY,
For the ELISA antibody titers, determined as described above,
USA) as per the manufacturer’s directions. The levels (pg/mL) were
cytokine concentrations determined in Luminex assays, and
measured, and heat maps were generated to report the fold increase
ELISpot assays, pairwise treatment groups were compared by the
[vaccine group (pg/mL)/PBS control group (pg/mL)] in cytokine or
Wilcoxon rank sum test. The median, Q1, Q3, GM, and GSE
chemokine levels.
were reported.
ELISpot assays were performed to measure interferon-gamma
The potential synergistic effects of vaccine and antibiotic on
(IFN-g) expression. Purified splenocytes, 4 weeks after vaccination,
protection were analyzed using the Bliss synergy (Demidenko and
were seeded in the presence of rF1V, TS C12, medium alone, or
Miller, 2019). Synergy score values indicated the ratio of median
PMA/ionomycin. Briefly, 96-well plates were coated overnight at 4°C
TTM and were based on a log-logistic or log-normal parametric
with 80 µL/well of capture anti-mouse IFN-g monoclonal antibody.
survival model.
The plates were washed one time with PBS. Recombinant F1V (25
mg/mL) or TS C12 (5 mg/mL) was resuspended in CTL-Medium with
1% L-glutamine and 100 µl was added to each well. The plates were 3 Results
incubated at 37°C, 9% CO2 for 15 min. Splenocytes were resuspended
in CTL-Medium with 1% L-glutamine and seeded at 8 × 105 cells per 3.1 The live attenuated vaccines protected
well for rF1V and 8 × 104 cells per well for TS C12 stimulations. The mice against bubonic plague but not
plates were incubated for 24 h at 37°C and 9% CO2, splenocytes were against pneumonic plague after challenge
removed, and the plates were washed twice with PBS alone and then with Yersinia pestis C12
twice with PBS and 0.05% Tween. Eighty microliters per well of
biotinylated detection anti-mouse IFN-g monoclonal antibody was We demonstrated previously the efficacy of live attenuated
added. After 2 h of incubation at room temperature, the plates were vaccine strains of Y. pestis for the protection of mice against
washed three times with PBS and 0.05% Tween. Eighty microliters of bubonic and pneumonic plague after exposure to the virulent
Strep-AP antibody solution was added to the wells, and the plates encapsulated Y. pestis CO92 strain (Cote et al., 2021). We
were incubated for 30 min at room temperature. Development describe here that the same strategies could not similarly protect
reagents were added and incubated for 15 min at room BALB/c mice against plague in animals exposed to Y. pestis strain
temperature according to the manufacturer’s recommendations. C12, a nonencapsulated derivative of CO92. As demonstrated in
The colorimetric reaction was stopped by washing the plates three Figure 1A, protection was afforded to mice vaccinated with a single
times with distilled water and air drying overnight. Spots were dose of Y. pestis mutant CO92 DyscN or C12 DyscN or a
scanned and analyzed using an automated ELISpot reader (CTL- combination of both DyscN vaccine strains (Combo LAV), against
ImmunoSpot S6 Analyzer, CTL, Germany). The splenocyte response a SC challenge with approximately 214 LD50s of Y. pestis C12 (40%,
was assessed as spot-forming cells (SFCs), adjusted to 106 cells per 60%, or 70%, respectively); these survival rates were
well, which was automatically calculated by the ImmunoSpot® greater or nearly greater than the mean rate of the control mice,

FIGURE 1
Survival curves of groups of vaccinated BALB/c mice challenged with Yersinia pestis strain C12. The mice were vaccinated with a single dose of
CO92 DyscN (1.3 × 107 CFU) or C12 DyscN (1.7 × 107 CFU) or both (Combo LAV), 1.54 × 107 CFU total; the control group received KPhos buffer
alone. All groups contained 10 mice. Twenty-eight days later, the animals were exposed to 1.93 × 103 CFU (214 LD50s) by the SC route (A) or to 1.71
× 106 CFU (15 LD50s) by the aerosol route (B) and monitored for 21 days after challenge.
with p = 0.087, p = 0.011, and p = 0.003, respectively. Nevertheless, 3.2 Heterologous vaccination strategies
the vaccine-mediated protection was less robust than that reported typically protected better than live
previously for mice challenged with the wild-type Y. pestis CO92 attenuated vaccines alone against a
strain, against which some of the live vaccines protected completely Yersinia pestis C12 challenge
(Cote et al., 2021). Not surprisingly, these single-dose strategies
offered no protection to mice after exposure to aerosolized Y. pestis The LAV strains are clearly immunogenic as determined by
C12 with only slight, statistically insignificant differences in mean humoral and cell-mediated immune responses that protect against Y.
TTM (Figure 1B). pestis CO92 (Cote et al., 2021), and these immune responses protect
As demonstrated in Figure 2, two doses of the LAV (prime– mice against bubonic, but not pneumonic, plague caused by Y. pestis
boost) given on days 0 and 23 yielded only a modest increase in C12. Encapsulated LAV candidate strains produce a robust anti-F1
protection against the C12 challenge compared with the single-dose humoral immune response, but due to possible immune interference,
cohorts (Figure 1). In the bubonic model, the protection afforded the anti-F1 responses may diminish the induction of a more diverse
after a booster vaccination by CO92 DyscN and C12 DyscN was immunity against other protective antigens (Cote et al., 2021).
increased by 20% (to 60% and 80%, respectively), and the extent of To enhance the availability of other protective antigens but still
protection by the Combo LAV vaccine was unchanged (70%, retain the F1-mediated protection, approximately equal parts of
Figure 2A); the mean survival rates were statistically greater than CO92 DyscN and C12 DyscN vaccine mix (Combo LAV) were
the mean survival rate of the control mice (p = 0.057 to p = 0.006). selected for further evaluation in heterologous vaccine strategies to
The two-dose homologous vaccination schemes with our live improve the protection against Y. pestis C12 infection.
attenuated DyscN strains did not improve survival after exposure In these experiments, we primed mice with a dose of the Combo
to aerosolized Y. pestis C12 (Figure 2B), but it slightly extended the LAV. The mice were then boosted (28 days later) with either the
TTMs on days 7 and 21 after challenge (p = 0.022 to p < 0.0001) as same LAV given in the first injection or with a subunit vaccine:
compared with the controls. rF1V fusion protein or rV protein (groups 2–4, 7, 8) (Figure 3). We

FIGURE 2
Survival curves of BALB/c mice vaccinated twice with the homologous LAV (days 0 and 23) and challenged 28 days later with Yersinia pestis strain
C12 by the SC route (A) or by the aerosol route (B). The doses of C12 were 2.38 × 103 CFU (264 LD50s) or 2.5 × 105 CFU (3 LD50s), respectively. The
mice were monitored for 21 days after challenge.
also tested the impact of further adjuvanting the recombinant significantly different from that of group 4 (p = 0.141). However, the
protein boost with the addition of oligonucleotide CpG2006. Mice TTM of the latter was longer than that of mice given the Combo
which received any of the four heterologous vaccines (groups 3, 4, 7, LAV and rF1V without CpG (p = 0.048). All five vaccinated groups
8) were highly protected against bubonic disease when challenged 4 survived slightly but significantly longer than did the KPhos
weeks after the boost vaccination, as illustrated in Figure 3A. controls (p ≤ 0.041), with mean TTMs for non-survivors of 3.5
Vaccination with the Combo LAV followed by rF1V + CpG, or days (controls) and 4.0 to 5.4 days for the vaccinated groups. Thus,
rV +/− CpG, resulted in 100% survival (groups 4, 7, 8), p ≤ 0.001 vs. the immunostimulant CpG enhanced the protection afforded by the
KPhos control; and 90% survival was observed in groups given heterologous Combo LAV and rF1V vaccine.
either the Combo LAV first and then rF1V without CpG (group 3) To further support these findings, mice were vaccinated again
or with two doses of the Combo LAV vaccine (group 2), p = 0.0001. with the homologous or heterologous prime/boost scheme but were
Two groups receiving KPhos and CpG alone plus the Combo LAV, challenged with a mean inhaled dose of Y. pestis C12 approximately
as either the prime or boost doses, resulted in 80% survival after two-fold greater than the dose in the study described in Figure 3B,
the SC challenge (groups 6 and 9), p = 0.0007 vs. the control, i.e., 1.65 × 106 CFU (22 LD50s). As shown in Figure 3C, the
respectively. Thus, in the bubonic model, the heterologous vaccine heterologous vaccines outperformed two doses of the Combo
combinations were generally more protective against Y. pestis C12 LAV. Whereas the latter provided no protection, the groups
than the homologous vaccine schemes. vaccinated with the LAV followed by rF1V, with or without CpG,
Furthermore, heterologous approaches significantly improved were partially protected. The survival rate of the group receiving the
the clinical outcome in mice exposed to aerosolized Y. pestis C12 heterologous vaccine with CpG was significantly greater than that of
(Figure 3B). The best survival against aerosolized Y. pestis C12 the control and Combo LAV ×2 groups on day 7 and day 21 after
challenge was observed in mice vaccinated with the heterologous challenge (p = 0.0325); however, these survival rate comparisons to
combinations of Combo LAV (prime) followed by a boost of rF1V the control and Combo LAV ×2 groups were not statistically
with CpG (Figure 3B). The survival rates of these mice (90%) at different for the mice receiving the heterologous vaccine without
both early (day 7) and final time points post-challenge (group 4) CpG. The TTMs of the two heterologous vaccination groups were
were significantly better than those of four of the other five groups greater than those of the KPhos control and LAV ×2 groups on both
(p-values ranged from p = 0.0001 to p = 0.048). Although 50% of day 7 and day 21 post-challenge (p ≤ 0.0423). Overall, vaccination
group 3 (Combo LAV and rF1V) survived, this survival rate was not with heterologous combinations of LAVs and subunit protein

FIGURE 3
A comparison of protection elicited by homologous and heterologous vaccines against Yersinia pestis strain C12 challenge. The vaccine doses
contained approximately 1 × 107 CFU total Combo LAV (ranging from 0.76 to 0.88 × 107 CFU) and/or 2 µg of protein subunit vaccine: rF1V or rV.
Two vaccinations were given to the mice (n = 10/group). Vaccinated mice received two doses 28 days apart and were exposed to C12 by the SC
route (3.62 × 103 CFU C12, 402 LD50s) 27 days later (A). Vaccinated mice received two doses 21 days apart and were then exposed to 7.93 × 105
CFU of aerosolized C12 (10 LD50s) 28 days later (B). Other groups of mice were vaccinated with two doses 21 days apart and then exposed to a
larger challenge (1.65 × 106 CFU, 22 LD50s) of aerosolized Y. pestis C12 29 days later (C). The mice were monitored daily for 21 days after challenge.
The vaccine regimens provided to the groups are described in Table 1 embedded into manuscript section 3.2.
vaccines was more strongly protective than homologous vaccine statistical analyses revealed significant vaccine-related differences.
strategies of LAVs, especially against aerosol challenge (Figures 3B, For the group vaccinated twice with the Combo LAV, there were no
C). Importantly, vaccination with the recombinant protein alone significant reductions in bacterial burden compared with the KPhos
was shown to be less effective against similar challenge doses of C12 control mice in the lung, spleen, and blood. In contrast, mice
compared with mice challenged with Y. pestis CO92 (Biryukov vaccinated with the heterologous Combo LAV/rF1V + CpG
et al., 2021). regimen had a reduced burden in the lungs (GM of 274 CFU/g),
Tissue samples and blood were collected from cohorts of mice and the bacterial load was significantly reduced compared with the
in the study described in Figure 3B to assess bacterial load and other five groups (p = 0.026 to 0.0022). In addition, a GM of 123
antibody titers on day 3 after the aerosol challenge, and the results CFU/g was isolated from the spleen samples of mice vaccinated
are shown in Tables 2, 3. Although there were large variations in with Combo LAV/rF1V + CpG, with four of six mice having no
organ and blood bacterial loads between mice in some groups, detectable CFU, and blood samples yielded a GM of 237 CFU/mL

TABLE 1 Vaccination regimens used in the experiment detailed in Figure 3. were reduced, but there were no differences compared with the
other five groups. Overall, the three vaccine groups, namely, Combo
Group # Prime vaccination Booster vaccination LAV/rF1V, KPhos/rF1V, and KPhos/rF1V + CpG, had generally
1 KPhos KPhos similar bacterial loads. The GMs of the lung, spleen, and blood CFU
2 Combo LAV Combo LAV
of mice with positive cultures were not statistically different in
comparison between these three vaccine groups. However,
3 Combo LAV 2 µg rF1V
compared with the Combo LAV ×2 burdens, two of these
4 Combo LAV 2 µg rF1V + 5 µg CpG vaccine groups had significant reductions of one to two logs in
5 5 µg CpG + KPhos Combo LAV + 2 µg rF1V lung CFU (p = 0.026 for the KPhos/rF1V group and p = 0.041 for
the Combo LAV/rF1V-vaccinated mice), at least one log lower
6 5 µg CpG + KPhos Combo LAV
spleen CFU counts, and no detectable CFU in half of the blood
7 Combo LAV 2 µg rV samples from the KPhos/rF1V and Combo LAV/rF1V-vaccinated
8 Combo LAV 2 µg rV + 5 µg CpG mice. Thus, the bacterial burdens of these three groups were
generally intermediate between those mice with the smallest
9 Combo LAV KPhos + 5 µg CpG
bacterial loads (Combo LAV/rF1V + CpG) and the KPhos controls.
10 KPhos 2 µg rF1V Antibody titers to four antigens were assessed by ELISA on sera
11 KPhos 2 µg rF1V + CpG from blood collected 3 days after challenge. The antigens included
rF1V, rV, irradiated TS CO92 whole cells, and irradiated TS C12
whole cells, and the results are shown in Table 3. The most elevated
(Table 2); no bacteria were recovered from the blood of five out of antibody titers overall were those to the rF1V fusion protein or the
six mice vaccinated with Combo LAV/rF1V + CpG. rV antigen which reiterates the strong immunogenicity of the F1
For the Combo LAV/rF1V-vaccinated mice, the lung CFUs and V antigens in BALB/c mice. All five vaccines induced anti-rF1V
were only significantly less than those recovered from the Combo and rV titers that were significantly greater than those of the KPhos
LAV ×2-vaccinated mice (p = 0.041); the spleen and blood burdens control mice, with p = 0.0159 for all groups (Table 3, Supplementary
TABLE 2 The bacterial burden of mice collected three days after challenge with Yersinia pestis C12.
Vaccine # Positive CFU/g or mLb

Prime Boost Organ samples/total #a GM (GSD)
KPhos KPhos Lung 6/6 1.08 × 109 (2.2)
Spleen 6/6 3.98 × 107 (12.3)
Blood 5/5c 1.92 × 104 (5.9)
Combo LAV Combo LAV Lung 6/6 1.43 × 109 (1.9)
Spleen 6/6 2.07 × 106 (39.3)
Blood 6/6 4.27 × 103 (10.0)
Combo LAV rF1V Lung 6/6 2.36 × 108 (4.3)
Spleen 5/6 1.84 × 104 (1,305.3)

2
Blood 3/6 8.07 × 10 (30.2)
Combo LAV rF1V + CpG Lung 4/6 2.74 × 102 (1,232.6)
Spleen 2/6 1.23 × 102 (1,577.8)
Blood 1/6 2.37 × 102 (19.3)
KPhos rF1V Lung 6/6 5.29 × 106 (1,921.7)
Spleen 4/6 4.23 × 103 (2,960.8)
Blood 3/6 9.06 × 102 (20.9)
KPhos rF1V + CpG Lung 6/6 1.16 × 108 (396.5)
Spleen 5/6 1.74 × 105 (662.4)
Blood 4/5 3.21 × 103 (26.5)

a
Number of samples with CFU recovered/total number of mice sampled.
b
GM and GSD values included all samples, with the LOD/SQRT(2) for those with no CFU recovered.
c
One mouse succumbed on day 3; the sample was not available.

TABLE 3 Total IgG antibody response against rF1V, Yersinia pestis CO92, or Y. pestis C12.
Vaccine Antibody titera

Vaccination Groupb nc rF1V rV TS CO92 TS C12
KPhos 1 4 94 (1.22) 200 (1.23) 212 (1.12) 168 (1.19)
Combo LAV ×2 (21 days) 2 5 221,106 (1.87) 3,200 (1.65) 52,947 (1.42) 9,263 (1.45)
Combo LAV ×1; rF1V (21 days) 3 6 387,944 (1.49) 114,035 (2.76) 775,889 (1.62) 741 (2.18)
Combo LAV ×1; rF1V + CpG (21 days) 4 6 870,906 (1.35) 323,050 (1.71) 29,314 (1.50) 1,008 (1.77)
KPhos ×1; rF1V (21 days) 10 6 117,579 (1.15) 57,470 (1.66) 504 (1.26) 52 (1.04)
KPhos ×1; rF1V + CpG (21 days) 11 5 265,996 (2.07) 177,726 (1.98) 1,676 (1.71) 120 (1.53)
a
Values represent GM antibody titer with GSE in parentheses directed against the rF1V protein, rV protein, temperature-shifted CO92 (TS CO92), or temperature-shifted C12 (TS C12) strains.
b
The vaccine group numbers are the same as those shown in Figure 3.
c
Number (n) of animal samples for each group. The n applies to tests with all antigens, except n = 5 (instead of 6) for the Combo LAV/rF1V + CpG group tested with TS CO92 and TS C12.
Table 1 describes statistical analyses performed). Mice vaccinated approximately 10- to 30-fold lower [Gro-a (CXCL1), LIF, IL-1b,
with Combo LAV/rF1V + CpG generated the highest anti-rV titers GM-CSF, MIP-1a, IL-1a, and IFN-g]. In contrast, the cytokine
that were two- and three-fold higher than Combo LAV/rF1V and levels were generally significantly elevated in the Combo LAV ×2-
KPhos/rF1V + CpG, although these differences did not reach vaccinated mice. The levels of IL-3, RANTES (CCL5), IL-4, and
significance. Vaccination with a priming dose of the Combo LAV IFN-g were from 5.3- to 7.1-fold greater than KPhos controls. These
and a booster dose of rF1V induced the highest titers to TS CO92, data were inversely associated in part with the vaccine-associated
whereas two doses of the Combo LAV vaccine induced the highest protection against the C12 challenge in that the Combo LAV/rF1V
titers to TS C12 (p = 0.017). Likely, the antibody responses to TS + CpG vaccine was the most protective and Combo/LAV ×2 the
CO92 whole-cell antigen were largely directed against the F1 least protective (Figures 3B, C). The results of the analysis of spleen
capsule, and adding rF1V to the Combo LAV elicited higher titer samples were similar overall to those of the lung samples, but the
to the rF1V and TS CO92 whole-cell antigens. The addition of CpG degree of differences in the levels between the vaccine groups was
to Combo + rF1V in group 4 greatly increased titers to rF1V less (Table 5, Supplementary Table 3). In contrast to the lung
compared with the mean titers of group 1 (p = 0.0095) and group 10 results, IFN-g, MCP-1, IL-1a, IP-10, and other cytokine levels were
(p = 0.0022). However, the addition of CpG was associated with repressed in all vaccine groups, and IL-23 (a Th17 cytokine) was
repressed titers when TS CO92 whole cells were used as the capture uniquely elevated in the Combo LAV/rF1V mice (but not in the
antigen and did not significantly increase the titers to TS C12 cells. Combo LAV/rF1V + CpG group). In aggregate, the spleen data do
Specifically, the anti-TS CO92 levels in group 4 mice were less than not appear to be as informative as the lung data in comparing
those in group 3 (p = 0.0043), though not in group 2 (p = 0.643). cellular immune responses to vaccination of these female BALB/c
The vaccines containing only a single dose of rF1V elicited the mice, but they suggest that early post-challenge decreases in
lowest titers overall, albeit these were still elevated against the rF1V bacterial lung burdens of vaccine-protected mice might result in
antigen. Overall, the antibody levels against the TS C12 antigen less dissemination to the spleen.
were much less than those to the rF1V and TS CO92 antigens. This
finding might suggest either that surface antigens other than the
immune-dominant F1 protein do not elicit strong humoral 3.3 Extended time between vaccination
responses or that the C12 DyscN LAV was cleared in the mice doses increased overall titers
before reaching immunostimulatory levels due to the lack of an
anti-phagocytic capsule. In an effort to further characterize heterologous plague
Lungs and spleens were obtained from animals collected on day vaccination approaches, we examined the impact of extended
3 post-challenge to analyze the cytokine responses. The lung sample times between the prime and boost vaccine injections along with
data revealed that three vaccine groups had levels of cytokines that varying durations between the prime vaccine and subsequent
were repressed or not significantly different overall relative to the booster dose(s) in the absence of CpG. We also chose to re-
KPhos group levels (heat map in Table 4, Supplementary Table 2). evaluate the ability of the CO92 pgm− pPst− LAV to protect mice
The Combo LAV/rF1V + CpG and KPhos/rF1V groups had from the C12 challenge because of its significantly different protein
responses that were relatively most repressed. The KPhos/rF1V + expression profile and the fact that it is an excluded non-select agent
CpG group cytokines also exhibited reduced expression, whereas strain. In this experiment, the vaccines were administered once only
the Combo LAV/rF1V group responses were not appreciably or with one or two additional booster dose(s) given at different
different from those of the KPhos group. For example, in the times after the first dose, as detailed in Figure 4 (The vaccine
Combo LAV/rF1V + CpG group, the levels of three cytokines regimens provided to the groups are described in Table 6).
were approximately 100-fold lower than those of the controls (IL-6, Specifically, one group received a single dose only (group 2).
MIP-2a, G-CSF), while seven additional cytokines were Single boosters were administered 21 days after the prime dose

TABLE 4 Fold change in cytokine levels of lung homogenates from vaccinated groups relative to the KPhos group.
Combo LAV ×2 Combo LAV/rF1V Combo LAV/rF1V + CpG rF1V rF1V + CpG
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 10) (Grp 11)
IL-3 7.13 2.98 1.86 1.71 2.21
IL-28 1.27 1.00 1.61 0.74 1.18
IL-10 1.24 0.99 1.55 0.83 0.99
RANTES (CCL5) 5.26 1.22 1.43 1.10 1.43
IL-23 1.28 1.01 1.31 1.24 1.00
IL-4 6.02 2.72 1.27 0.49 0.73
Eotaxin 3.39 1.11 1.27 0.78 1.46
IFN-a 1.00 1.00 1.00 1.00 1.00
IL-5 3.72 1.83 1.00 0.71 0.82
IL-15 2.64 1.24 0.87 0.91 1.42
IL-9 1.21 0.88 0.83 0.64 1.05
IL-2 2.17 1.23 0.77 0.61 0.94
IL-27 2.34 1.00 0.49 0.36 1.12
IL-13 2.13 1.07 0.48 0.37 0.71
ENA-78 (CXCL5) 1.29 0.83 0.47 0.28 0.82
IL-12p70 2.74 1.21 0.47 0.33 0.84
IL-31 1.61 1.19 0.41 1.28 1.28
M-CSF 1.95 1.30 0.38 0.67 1.16
IP-10 (CXCL10) 1.56 1.01 0.33 0.17 0.64
MCP-3 (CCL7) 1.29 0.92 0.30 0.17 0.67
IL-17A (CTLA-8) 1.79 1.62 0.24 0.08 0.65
MIP-1b (CCL4) 1.90 1.30 0.22 0.35 0.97
MCP-1 (CCL2) 1.29 0.57 0.19 0.13 0.91
IL-18 2.02 1.03 0.15 0.32 0.56
TNF-a 2.39 1.24 0.14 0.22 0.77
IL-22 3.86 2.37 0.13 0.18 0.80
IFN-g 6.46 1.19 0.10 0.12 0.43
IL-1a 1.26 0.73 0.07 0.13 0.60
MIP-1a (CCL3) 1.15 0.94 0.06 0.15 0.66
GM-CSF 0.68 0.73 0.05 0.09 0.53
IL-1b 1.92 0.80 0.04 0.07 0.44
LIF 1.18 0.95 0.04 0.10 0.45
GRO-alpha (CXCL1) 0.95 0.68 0.03 0.07 0.49
G-CSF 0.96 0.83 0.01 0.04 0.52
MIP-2a (CXCL2) 0.99 0.81 0.01 0.04 0.46
IL-6 1.50 0.92 0.01 0.04 0.39
a
The cytokine results are shown as the ratio to KPhos and are based on the GM (pg/mL). Italicized and not bolded (not significant); bolded (significant, p < 0.05).
b
Group # corresponds to Figure 3 and Table 1.

TABLE 5 Fold change in cytokine levels of spleen homogenates from vaccinated groups relative to the KPhos group.
Combo LAV ×2 Combo LAV/rF1V Combo LAV/rF1V + CpG rF1V rF1V + CpG
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 10) (Grp 11)
M-CSF 1.11 2.48 1.83 3.29 2.20
RANTES (CCL5) 1.63 0.65 1.74 0.91 2.54
IL-4 2.41 0.47 1.71 0.91 1.09
IL-3 0.94 1.33 1.58 1.38 1.33
IL-10 0.93 1.08 1.43 1.26 1.38
IL-23 0.92 5.23 1.32 2.04 1.48
IL-31 0.75 0.64 1.04 0.68 0.82
IFN-a 1.00 1.00 1.00 1.00 1.00
IL-28 0.97 0.53 0.95 1.03 0.82
IL-9 0.67 0.87 0.87 1.21 0.82
MIP-1 b (CCL4) 1.50 0.50 0.87 0.94 1.61
IL-13 0.51 1.01 0.75 1.08 0.66
IL-27 1.17 0.54 0.60 0.75 0.87
IL-12p70 0.75 0.45 0.52 0.85 0.78
IL-2 0.85 0.82 0.48 1.25 1.00
IL-22 1.71 0.88 0.42 1.27 0.81
ENA-78 (CXCL5) 0.96 0.35 0.41 0.53 0.61
IL-5 0.85 1.02 0.39 1.52 1.02
MIP-1a (CCL3) 0.89 0.30 0.39 0.87 0.96
Eotaxin 1.47 0.46 0.39 0.59 0.71
IL-15 1.35 0.76 0.37 1.30 1.35
IL-1b 1.14 0.53 0.37 0.80 0.97
GM-CSF 0.95 0.53 0.34 0.90 0.71
LIF 1.43 0.56 0.31 1.16 0.57
MIP-2a (CXCL2) 0.41 0.43 0.29 1.05 0.34
IL-17A (CTLA-8) 1.10 0.51 0.29 0.80 0.53
TNF-a 0.85 0.41 0.28 0.76 0.57
IP-10 (CXCL10) 0.49 0.18 0.24 0.32 0.42
MCP-3 (CCL7) 1.08 0.21 0.20 0.44 0.78
IL-18 0.89 0.20 0.12 0.36 0.54
IFN-g 0.42 0.19 0.12 0.33 0.36
IL-1a 0.24 0.21 0.08 0.36 0.52
MCP-1 (CCL2) 0.42 0.16 0.06 0.34 0.43
GRO-a (CXCL1) 0.64 0.33 0.05 0.24 0.34
IL-6 1.35 0.26 0.03 0.43 0.43
G-CSF 1.15 0.30 0.02 0.24 0.49
a
The cytokine results are shown as the ratio to KPhos and are based on the GM (pg/mL). Italicized and not bolded (not significant); bolded (significant, p < 0.05).
b

(group 3), 46 days after the prime (groups 4 and 9), or 90 days after Thus, a heterologous LAV/rF1V vaccination scheme appears to be
the prime (groups 5, 7, 8, 10, 11). Double boosters were injected on optimal, but further evaluation is needed.
days 46 and 90 after the prime (groups 6 and 12). Figure 4 illustrates Antibody titers to the four antigens were assessed by ELISA on
the survival rates after the aerosol challenge with a dose of 13 LD50s sera collected from a cohort of mice 1 day before the aerosol
of C12. Two groups that had been primed with a LAV (CO92 pgm− challenge. The antigens included rF1V, rV, Y. pestis TS CO92,
pPst− or CO92 DyscN, groups 7 and 11, respectively) and given one and Y. pestis TS C12. In agreement with the protection data, the two
booster of rF1V (without CpG) 90 days after the prime dose had the groups (7 and 11) with the greatest survival rates had the highest
highest survival rates (37.5%), but survival did not reach titers to rF1V and CO92 whole cells but not to C12 whole cells
significance relative to the KPhos group. A small number of (Table 7), similar to the responses to C12 described in Table 3, and
survivors (12.5%) were observed in a heterologous group of mice the differences were confirmed statistically with the analyses
initially given the CO92 pgm− pPst− LAV and boosted twice, first detailed in Supplementary Table 4. Importantly, groups 7 and 11
with CO92 pgm− pPst− at 46 days after the prime and second with did have some of the highest anti-rV titers in this study. These two
rF1V at 90 days after the initial vaccination (group 6). The mice groups had been primed with either the encapsulated CO92 pgm−
primed with CO92 DyscN and boosted on day 46 with the same pPst− or CO92 DyscN LAV and boosted once with rF1V 90 days
LAV and on day 90 with rF1V (group 12) had no survivors but had later. Similarly, mice in groups 6 and 12, which had received prime
a significantly longer mean TTM than the KPhos controls, with and boost doses at day 46 of a CO92 LAV (pgm− pPst− or DyscN)
geometric means of 5.0 and 3.8 days, respectively (p = 0.0285). The and a second boost at day 90 with rF1V, and which were partially
remaining groups of mice had no survivors and succumbed at the protected or had an extended TTM, exhibited the second highest
same time post-challenge as the KPhos controls (day 4). Albeit antibody responses to rF1V. However, this partial protection was
complete protection was not achieved in this study, the results again not clearly associated with humoral responses to either the F1+ or
suggested that heterologous vaccination strategies are superior and F1− whole-cell antigens (Table 7, Supplementary Table 4). The most
that administering one booster injection of rF1V vaccine 3 months elevated anti-TS C12 levels were observed in mice given two
after priming with a LAV could be associated with better protection homologous doses of a LAV or heterologous prime and boost
than schedules where the prime and booster doses were spaced doses of the LAVs 90 days apart, i.e., groups 5, 8, and 10 in
more closely in time. Of interest, priming and boosting 90 days later Table 7, although the GM anti-TS C12 titers of all 11 vaccine
with homologous LAVs (groups 5 and 10) were not protective. groups were significantly elevated compared with the KPhos group
FIGURE 4
The influence on survival after Yersinia pestis C12 challenge of extending the time between prime and boost doses of the vaccines. The study
design, vaccines, and their schedule of administration are detailed in (A) and Table 6. A cohort of mice was euthanized on day 119 in order to collect
blood and organ homogenates for immunological assays, and the remainder of the mice were exposed to aerosolized Y. pestis C12 on day 120. The
LAV doses contained approximately 1 × 107 CFU (0.86–1.34 × 107 CFU), and 2 µg/dose of the subunit vaccines were delivered. The CO92 pgm−
pPst− live vaccine is deleted of the 102-kb pgm locus and cured of the pPst plasmid. Each vaccine group had eight mice. The mice were challenged
28 days after the final vaccination with 1.0 × 106 CFU (13 LD50s) of aerosolized Y. pestis C12 and checked daily for 21 days after challenge (B). NA,
not applicable. The vaccine regimens provided to the groups are described in Table 6 embedded into manuscript section 3.3.

TABLE 6 Vaccination regimens used in the experiment detailed in Figure 4.
Group # Prime vaccination Day 21 booster vaccination Day 46 booster vaccination Day 90 booster vaccination
1 KPhos NA NA NA
2 CO92 pgm− pPst− NA NA NA

− − − −
3 CO92 pgm pPst CO92 pgm pPst NA NA
− − − −
4 CO92 pgm pPst NA CO92 pgm pPst NA
− −
5 CO92 pgm pPst NA NA CO92 pgm− pPst−
6 CO92 pgm− pPst− NA CO92 pgm− pPst− rF1V
7 CO92 pgm− pPst− NA NA rF1V
8 CO92 pgm− pPst− NA NA CO92 DyscN
9 CO92 DyscN NA CO92 DyscN NA
10 CO92 DyscN NA NA CO92 DyscN
11 CO92 DyscN NA NA rF1V
12 CO92 DyscN NA CO92 DyscN rF1V
(p = 0.008 to 0.029). The GM Y. pestis TS C12 antibody titer of day 90 with rF1V. These results contrast with the humoral responses
group 8 (320,000) was significantly greater than that of vaccine to Y. pestis TS C12, where the inclusion of rF1V reduced antibody
groups 2, 3, 6, 7, 9, 11, and 12 (p = 0.008 to 0.04). Apart from group titers. In addition, this may reflect a greater role for cell-mediated
2 (which received one prime dose only), these groups had received a immunity in protection associated with the DyscN LAVs relative to
final dose on day 90 of rF1V. The presence of rF1V in the vaccines the CO92 pgm− pPst− vaccine.
appears to mostly blunt the humoral response to Y. pestis TS C12. Cytokine production by stimulated splenocytes from
Generally, the humoral responses to Y. pestis TS C12 were not vaccinated groups was analyzed with Luminex bead-based
associated with protection. immunoassay. The levels of cytokines produced by splenocytes
To assess cellular immune responses, spleens were collected from stimulated with rF1V of the vaccinated groups relative to the
animals 1 day before the aerosol challenge for ELISpot and Luminex KPhos group are illustrated in Table 8 and Supplementary
restimulation assays. ELISpot assays for IFN-g were performed using Table 5. Relative to KPhos, all CO92 DyscN-vaccinated groups
unstimulated splenocytes or splenocytes stimulated with rF1V or Y. promoted a strong cytokine response, with the most notable
pestis TS C12. The unstimulated cells from the vaccine groups upregulation of MIP-1a, IL-2, MCP-3, MIP-1b, IL-3, IL-17A,
produced ≤15 IFN-g-producing cells (SFC) per 106 splenocytes MIP-2a, and RANTES. The addition of rF1V to CO92 DyscN-
(Supplementary Figure 1). In contrast, splenocytes from all the vaccinated mice drastically enhanced overall cytokine response,
vaccinated mice stimulated with rF1V secreted much higher levels while the addition of rF1V to CO92 pgm− pPst−-vaccinated
than did the KPhos controls (Figure 5A), p = 0.0034 by pairwise groups only moderately and selectively enhanced some
comparison analyses. The greatest levels of IFN-g were produced by cytokines, notably, MIP-1a, IL-2, MIP-1b, IL-4, and IL-6.
rF1V-stimulated cells from mouse groups 6, 7, 11, and 12 (p < Relative cytokine production by splenocytes stimulated with Y.
0.0001). As illustrated for the humoral immune responses, these pestis TS C12 showed an overall increased cytokine response for six
results again paralleled the differential protective efficacies afforded by groups (groups 2–7), all of which had received the CO92 pgm−
the vaccines. Interestingly, as depicted in Figure 5B, Y. pestis TS C12 pPst− LAV with or without rF1V. This increased response was most
whole cells elicited IFN-g maximally from groups 9 to 12 (p < 0.0001 pronounced for mice vaccinated with a double dose of CO92 pgm−
compared with KPhos controls), all of which had been primed with pPst− 46 days apart with or without rF1V (groups 4 and 6) (Table 9,
the LAV CO92 DyscN and not with the CO92 pgm− pPst− vaccine. Supplementary Table 6). Substantially greater responses were
We hypothesize that the difference in IFN-g production may be observed for five groups (groups 8–12), all of which were CO92
associated with the expression of V by the CO92 pgm− pPst LAV, as DyscN vaccinees (Table 9). These results agreed in part with the
V is linked to immune suppression. Additionally, the altered humoral antibody and ELISpot data and suggest that the elevated
secretion of the other Yops may also have contributed to the antibody responses to rF1V and Y. pestis TS CO92, and the
different immune responses observed. The numbers of SFCs from heightened Y. pestis TS C12-stimulated cellular immune
splenocytes of these four groups (Figure 5B) were greater than those responses following CO92 DyscN vaccination, may be generally
produced after stimulation with rF1V (Figure 5A). IFN-g expression associated with protection. However, the associations between the
by Y. pestis TS C12-stimulated cells was maximal for groups 11 and cell-mediated immune responses and survival rates are not readily
12 (but not statistically different from groups 9 and 10; p = 0.17); both apparent, likely due to the low vaccine-mediated protection against
groups 11 and 12 had been primed with CO92 DyscN and boosted on the high challenge dose of Y. pestis C12 delivered in this study.

TABLE 7 Total IgG antibody response against rF1V, rV, Yersinia pestis CO92, or Y. pestis C12.
Vaccine Antibody titera

Vaccinationb Groupb n rF1V rV TS CO92 TS C12
KPhos 1 4 50 (1.00) 63 (1.26) 50 (1.00) 79 (1.36)
− −
CO92 pgm pPst ×1 2 4 134,543 (1.20) 378 (1.24) 36,061 (1.17) 9,551 (1.65)
− −
CO92 pgm pPst ×2 (day 21) 3 4 183,792 (1.27) 919 (2.63) 121,257 (1.45) 96,242 (1.29)
CO92 pgm− pPst− ×2 (day 46) 4 4 367,583 (1.42) 1,528 (1.53) 160,000 (1.57) 83,784 (1.71)
CO92 pgm− pPst− ×2 (day 90) 5 4 557,152 (1.17) 3,510 (1.27) 367,583 (1.53) 167,567 (1.36)
− −
CO92 pgm pPst ×2 (day 46); rF1V (day 90) 6 4 1,849,578 (1.27) 270,235 (1.28) 485,029 (1.26) 60,821 (1.43)
− −
CO92 pgm pPst ×1; rF1V (day 90) 7 4 14,039,360 (1.39) 186,721 (1.27) 1,167,005 (1.26) 9,701 (1.59)
− −
CO92 pgm pPst ×1; CO92 DyscN (day 90) 8 4 367,583 (1.27) 1,008 (1.31) 485,029 (1.33) 320,000 (1.22)
CO92 DyscN ×2 (day 46) 9 5 211,121 (1.55) 764 (2.00) 145,876 (1.62) 73,169 (1.63)
CO92 DyscN ×2 (day 90) 10 5 139,288 (1.45) 504 (1.27) 201,587 (1.54) 152,775 (1.32)
CO92 DyscN ×1; rF1V (day 90) 11 5 12,241,312 (1.37) 107,243 (1.65) 2,799,007 (1.23) 8,844 (1.69)
CO92 DyscN ×2 (day 46); rF1V (day 90) 12 5c 769,936 (1.51) 204,800 (1.60) 139,288 (1.60) 44,012 (1.81)
a
Values represent GM antibody titer with GSE in parenthesis directed against the rF1V protein, rV protein, Y. pestis temperature-shifted CO92 (TS CO92), or C12 (TS C12) strains.
b
The vaccine group numbers are the same as those shown in Figure 4.
c
n = 5 for all the antigens, with the exception of rV that had n = 4.
3.4 Live attenuated Yersinia pestis vaccines compared with the unvaccinated and vaccinated control groups
could be an important component of (Figure 6B), i.e., 7.0 days vs. 5.0 days (p = 0.0044) or vs. 5.5 days (p =
layered defense strategies against plague 0.044), respectively. The benefits accrued from this vaccination–
treatment scheme are further illustrated in Figure 6C. To improve
Following possible exposure to Y. pestis, the time between the survival of unvaccinated mice treated with antibiotic at a
infection and manifestation of lethal disease is brief. During that delayed time (60 h vs. 48 h) after challenge, animals were given a
time, an individual needs to be properly diagnosed in order to higher dose of streptomycin. A total of 20% of mice receiving 40
receive the appropriate therapy; hence, any effort to prolong mg/kg of streptomycin 60 h after challenge survived. In contrast,
the window of opportunity for therapeutic intervention can 70% of the CO92 DyscN-vaccinated and antibiotic-treated mice
dramatically improve disease outcome. The protection afforded to survived, an increase that approached significance on days 7 and 21
BALB/c mice against aerosol challenge with Y. pestis C12 by a (p = 0.070). All of the mice in the two control groups succumbed by
suboptimal vaccination and delayed post-challenge antibiotic day 6, p = 0.0044 vs. CO92 DyscN-vaccinated and KPhos control
treatment strategy was examined. The mice were either groups compared with vaccinated and antibiotic-treated mice
vaccinated twice with mutant CO92 DyscN (21 days apart) or (Figure 6C). The C12 DyscN vaccine was evaluated similarly in
administered buffer alone and then challenged 4 weeks after the the context of antibiotic treatment (Figure 7). Vaccination and
second vaccine dose with a lethal aerosol dose of Y. pestis C12. They treatment with 20 mg/kg of streptomycin 60 h post-challenge with
were subsequently treated with 20 mg of streptomycin/kg (or buffer Y. pestis C12 elicited significantly better protection against aerosol
alone) beginning 48 or 60 h after challenge, and the effects on infection. In addition, the Y. pestis C12 challenge dose was higher in
survival were determined, as shown in Figures 6A, B. The this experiment (1.67 × 106 CFU, 22 LD50s) than that used in the
vaccination had no effect on the survival of mice after challenge study in Figure 6 (GM 3.62 × 105 CFU, 5 LD50s). Half of the
with Y. pestis C12, while the administration of the antibiotic 48 h vaccinated and antibiotic-treated mice survived, and none of the
following challenge was associated with 90% and 100% of the vaccinated only or antibiotic-treated only controls survived (p =
treated (only) and vaccinated mice treated with antibiotics, 0.0325 for survival on days 7 and 21 after challenge). This layered
respectively (Figure 6A). However, increasing the time between therapy revealed that the combinations of vaccination and post-
challenge and antibiotic treatment to 60 h revealed the enhanced exposure antibiotic treatment were synergistic in the challenge
effect on survival of the layered approach. Whereas all the mice experiments shown in Figure 6B (synergy score 4.84, p = 0.0013),
treated with 20 mg/kg of streptomycin alone or vaccinated alone Figure 6C (synergy score 2.66, p = 0.040), and Figure 7 (synergy
succumbed, 80% of the vaccinated and antibiotic-treated mice score 3.16, p = 0.009). These synergistic effects allowed the
survived (p = 0.0007 vs. treated and p = 0.0011 vs. vaccinated treatment to be delayed after exposure to aerosolized Y. pestis and
alone); the median TTM of the treated group was extended may facilitate a sparing effect on the amount of antibiotic required.

FIGURE 5
ELISpot assays for IFN-g-secreting splenocytes collected post-vaccination and detected after stimulation with rF1V (A) or temperature-shifted, killed
whole-cell TS C12 (B). The vaccine regimens provided to the groups are described in Table 6 embedded into manuscript section 3.3. SFC, spot-
forming unit. The samples were obtained from four or five mice/group and the data were expressed as GM and GSE. Note the difference in the
Y-axes ranges.
4 Discussion (Winter et al., 1960; Anderson et al., 1997). nonencapsulated strains

have been isolated in Indonesia and Kazakhstan (Meka-Mechenko,
Most wild-type strains of Y. pestis produce an outer capsule 2003; Eppinger et al., 2012), and while the impact on human
composed of a polymer of fraction 1 protein (F1) encoded by the populations is not well documented, a nonencapsulated or weakly
caf1 operon of the pMT1 plasmid. F1 is a highly immunogenic and encapsulated strain may have been associated with a fatal case of
protective virulence factor that appears to have a role in preventing plague in the USA in the 1950s (Winter et al., 1960). Thus, the
phagocytosis. However, the primary Y. pestis virulence factors that virulence of F1-negative strains has been widely documented, and
block uptake by phagocytosis are those encoded by the type 3 the findings support the conclusion that new-generation vaccines
secretion system (T3SS) (Du et al., 2002). In addition, strains with must protect against infection by nonencapsulated strains (Winter
mutations preventing F1 synthesis frequently retain virulence in et al., 1960; Drozdov et al., 1995; Friedlander et al., 1995; Welkos
various mammalian models of plague, especially in mouse and et al., 1995; Worsham et al., 1995; Davis et al., 1996; Samoilova et al.,
guinea pig pneumonic plague models (Drozdov et al., 1995; 1996; Heath et al., 1998; Powell et al., 2005; Quenee et al., 2008;
Friedlander et al., 1995; Welkos et al., 1995; Worsham et al., Sebbane et al., 2009; Sha et al., 2011; Du and Wang, 2016).
1995; Davis et al., 1996; Du and Wang, 2016; Biryukov et al., Nevertheless, F1-deficient strains are not as virulent as F1+ in
2021). These strains have not only been produced experimentally certain situations, and their virulence may vary depending on the Y.
but have also been recovered in natural settings of plague infection pestis strain, exposure dose and route, and animal host, i.e., strain,

TABLE 8 Fold change relative to the KPhos-vaccinated group in splenocytes stimulated with rF1V.
CO92 CO92 CO92 CO92 CO92 CO92

pgm− pgm− pgm− pgm− pgm− pgm− CO92 CO92
CO92 pPst− pPst− pPst− pPst− ×2 pPst− pPst− ×1; DyscN CO92 CO92 DyscN
pgm− ×2 ×2 ×2 (day 46); ×1; F1V CO92 ×2 DyscN DyscN ×2 (day
pPst− (day (day (day rF1V (day (day DyscN (day ×2 (day ×1; rF1V 46); rF1V
×1 21) 46) 90) 90) 90) (day 90) 46) 90) (day 90) (day 90)
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 5) (Grp 6) (Grp 7) (Grp8) (Grp 9) (Grp 10) (Grp 11) (Grp 12)
MIP-1a/
CCL3 0.61 1.10 0.92 0.92 5.56 6.54 27.19 46.80 28.38 441.81 300.77
IL-2 1.00 1.00 1.23 1.00 11.02 7.29 16.66 39.31 11.04 383.25 637.12
MCP-3/
CCL7 1.11 1.26 1.00 1.02 2.28 1.30 53.70 91.23 98.35 306.60 162.55
MIP-1b/
CCL4 0.58 0.98 0.96 0.95 3.12 3.22 23.72 32.17 25.75 144.16 125.90
IL-3 1.00 1.00 1.00 1.00 1.68 1.89 7.41 15.76 9.09 139.64 241.37
IL-17A/
CTLA-8 1.00 1.00 1.00 1.00 1.50 1.05 17.20 22.36 5.22 69.89 195.64
IL-5 1.00 1.00 1.00 1.00 3.97 2.88 1.00 1.43 1.10 68.64 360.44
IFN-g 1.00 1.00 1.00 1.00 1.04 1.00 1.76 4.69 2.33 44.96 54.61
IL-13 1.00 1.00 1.00 1.00 1.31 1.22 1.00 1.52 1.00 33.55 76.94
MCP-1/
CCL2 1.00 1.00 1.00 1.00 1.00 1.00 3.66 7.17 7.96 28.41 12.42
RANTES/
CCL5 0.88 0.88 0.88 0.88 0.97 0.91 10.07 10.92 13.90 28.36 26.56
IL-6 1.33 1.56 1.24 1.19 3.94 6.19 4.19 3.72 1.76 28.17 31.99
MIP-2a/
CXCL2 1.11 0.94 0.94 0.94 1.04 0.94 3.89 5.10 3.77 18.74 14.30
IL-4 1.00 1.38 1.26 1.00 3.05 4.03 1.43 2.32 1.34 11.97 11.51
IL-18 1.00 1.00 1.00 1.00 1.00 1.00 1.00 2.28 1.11 11.38 15.02
GM-CSF 1.00 1.00 1.00 1.00 1.00 1.00 1.12 1.95 1.00 8.76 16.40
IL-22 1.00 1.00 1.00 1.00 1.00 1.00 1.39 1.58 1.13 7.47 9.76
TNF-a 1.00 1.00 1.00 1.00 1.20 1.00 1.21 1.97 1.17 6.41 9.21
IP-10/
CXCL10 1.02 1.00 1.00 1.00 1.00 1.00 1.00 1.01 1.07 4.64 5.15
IL-10 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 2.25 3.09
LIF 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.47 2.20
IL-15 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.26 1.60
IL-12p70 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.16 1.29
IL-1b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.04 1.05
M-CSF 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.02 1.00
IL-23 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.02 1.00
IL-27 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Eotaxin/
CCL11 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IL-31 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
(Continued)

TABLE 8 Continued

pgm− pgm− pgm− pgm− pgm− pgm− CO92 CO92
CO92 pPst− pPst− pPst− pPst− ×2 pPst− pPst− ×1; DyscN CO92 CO92 DyscN
pgm− ×2 ×2 ×2 (day 46); ×1; F1V CO92 ×2 DyscN DyscN ×2 (day
×1 21) 46) 90) 90) 90) (day 90) 46) 90) (day 90) (day 90)
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 5) (Grp 6) (Grp 7) (Grp8) (Grp 9) (Grp 10) (Grp 11) (Grp 12)
G-CSF/CSF-
3 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IL-28 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IL-9 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
IL-1a 0.93 0.93 0.93 0.93 0.93 0.93 0.93 0.93 0.93 0.93 0.93
a
The cytokine results are shown as the ratio to the KPhos-vaccinated group and are based on the GM (pg/mL). Italicized and not bolded are not significant; bolded are significant (p < 0.05).
b
sex, or source (Donavan et al., 1961; Williams and Cavanaugh, not secrete the V antigen due to the disrupted T3SS but do retain
1984; Friedlander et al., 1995; Welkos et al., 2004; Sebbane et al., some V protein associated with the bacterial cells (Bozue et al.,
2009; Weening et al., 2011). The F1 antigen appears to contribute to 2012), and in the case of the C12-derived vaccine strain, no F1 was
Y. pestis virulence in some mammalian models of bubonic and produced. Thus, they allowed us to investigate the protection
sometimes pneumonic plague (Donavan et al., 1961; Williams and afforded by potentially novel antigens, such as non-capsule
Cavanaugh, 1984; Samoilova et al., 1996; Du et al., 2002; Sha et al., surface antigens displayed by the C12 DyscN, e.g., Pla, Ail, and
2011; Weening et al., 2011; Du and Wang, 2016). For example, in LPS (Wang et al., 2020).
BALB/c mice, the LD50 values of the F1+ CO92 and F1− C12 strains The current work extended these studies by characterizing
are approximately the same by the aerosol route (6.8–7.7 × 104 protection afforded by these LAVs, with or without the defined rF1V
CFU) (Heine et al., 2007; Biryukov et al., 2021), but CO92 has an subunit vaccine, against the challenge with the nonencapsulated and
approximately five-fold lower LD50 than C12 by the SC route vaccine-resistant C12 strain of Y. pestis. We found that the most
(bubonic model), ≤1.9 CFU vs. 9.1 CFU, respectively (Worsham protective vaccination scheme, as tested under various conditions and
et al., 1995; Biryukov et al., 2021; Cote et al., 2021). Sha et al. in both bubonic and pneumonic plague models, was heterologous
reported similar results with BALB/c mice from a different vendor, vaccination with a live vaccine (Combo LAV composed of the C12 and
and Weening and coworkers observed F1-associated virulence CO92 DyscN mutants, or the CO92 pgm− pPst− strain) and a separate
differences in both bubonic and pneumonic models in C57BL/J dose of a protein subunit vaccine. Although one or two identical doses
mice (Sha et al., 2011; Weening et al., 2011). The absence of F1 of either DyscN mutant alone or combined provided significant
impacted virulence to a greater extent in C57BL/6 mice than in protection against bubonic plague, they were poorly protective
BALB/c mice, showing the effect of animal background on virulence against pneumonic challenge (Figures 1, 2). With the inclusion of a
(Weening et al., 2011). In the outbred CD-1 mouse strain, the SC subunit rF1V vaccine, the heterologous vaccination approach was
LD50s of C12 and CO92 differed by 18-fold, i.e., 942 CFU and 52 optimally protective against aerosol challenge regardless of changes
CFU, respectively; similar findings were observed using outbred in vaccination schedule and challenge conditions. The latter included
Swiss–Webster mice (Sebbane et al., 2009; Biryukov et al., 2021; the spacing between prime vaccine doses and subsequent booster dose
Cote et al., 2021). (s) which was predicted to optimize antibody response (Castiglione
In previous studies, we characterized the protective efficacy of et al., 2012). For instance, heterologous vaccination increased the
several live attenuated mutant strains derived from Y. pestis strains survival rate or extended the survival time to a greater extent than
CO92 or KIM6+ against bubonic and aerosol challenges of BALB/c did two doses of Combo LAV when mice were exposed to a
mice with the wild-type encapsulated CO92 strain. Three of the significantly elevated C12 challenge. Additionally, in a heterologous
LAV strains were confirmed to be safe in high doses (Cote et al., scenario (Figures 3, 4), a different LAV, i.e., the CO92 pgm− pPst−
2021), and one dose of either F1+ CO92 mutant (the DyscN or pgm− strain, was as efficacious as the Combo LAV, when paired with rF1V
pPst − strain) protected BALB/c mice against infection by booster. This situation requires further examination, but it potentially
subcutaneous or inhalational routes to Y. pestis CO92. Full affords flexibility in the selection of vaccine components.
protection against CO92 in this model required the induction of In efforts to improve the rF1V vaccine, Amemiya et al. and
an immune response to F1 (Cote et al., 2021). Two vaccines (CO92 Biryukov et al. showed that the addition of TLR agonist CpG2006 to
DyscN, alone or combined with C12 DyscN) were down-selected for rF1V augmented vaccine efficacy and immunogenicity (Amemiya
further analyses of protection against other virulent strains and et al., 2009; Biryukov et al., 2021). A similarly protective role of CpG
especially those deficient in capsule synthesis. The DyscN strains do in mice vaccinated with LAVs or protein subunit combination

TABLE 9 Fold change relative to the KPhos-vaccinated group in splenocytes stimulated with TS C12.

pgm– pgm– pgm– pgm− pgm- pgm− CO92 CO92
CO92 pPst– pPst– pPst– pPst− ×2 pPst- pPst− ×1; DyscN CO92 CO92 DyscN
pgm− x2 ×2 ×2 (day 46); ×1; F1V CO92 ×2 DyscN DyscN ×2 (day
×1 21) 46) 90) 90) 90) (day 90) 46) 90) (day 90) (day 90)
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 5) (Grp 6) (Grp 7) (Grp 8) (Grp 9) (Grp 10) (Grp 11) (Grp 12)
IFN-gc 3.51 8.58 54.31 15.75 38.12 9.51 3,934.91 3,934.91 3,934.91 3,934.91 3,934.91
IL-17A/
CTLA-8 7.73 43.15 144.16 42.23 168.70 10.13 3,728.00 4,096.00 2,999.79 1,097.99 3,543.01
IL-22 1.15 2.55 5.86 2.15 8.19 1.47 313.26 350.16 339.07 261.92 263.57
IP-10/
CXCL10 1.26 1.28 6.91 2.58 2.70 0.96 141.82 150.03 220.79 172.34 173.59
IL-2 8.15 17.09 56.69 17.23 54.97 8.51 432.17 778.98 300.22 146.61 496.09
IL-3 1.71 4.75 13.24 4.76 11.79 1.68 419.25 731.91 468.32 138.09 389.26
IL-18 1.45 2.72 6.78 2.65 4.96 2.10 88.44 95.72 92.39 88.25 76.56
IL-13 1.06 2.03 6.97 2.12 5.32 1.00 215.21 384.91 301.46 64.48 160.89
GM-CSF 1.02 1.56 3.08 1.79 2.73 1.03 76.72 85.21 102.45 55.25 61.90
IL-6 1.07 2.24 2.15 1.57 1.54 1.66 28.88 37.45 44.94 30.97 39.55
IL-1b 1.00 1.23 1.58 1.27 2.24 1.37 16.79 17.29 27.00 26.03 15.51
IL-5 1.00 1.14 3.88 1.61 2.33 1.00 76.62 281.36 113.97 15.66 73.77
IL-4 2.78 8.68 9.93 4.46 4.42 2.43 21.29 34.74 30.44 14.40 21.26
MCP-3/
CCL7 1.15 1.40 1.61 1.79 1.49 1.19 8.98 12.51 23.95 12.36 15.34
MIP-1b/
CCL4 0.93 1.44 1.39 2.25 1.52 1.59 11.47 11.47 11.47 11.47 11.47
IL-12p70 0.99 1.13 1.51 1.09 1.20 0.98 10.54 12.07 11.22 10.96 9.75
G-CSF/CSF-
3 0.79 0.92 0.77 1.42 0.85 0.79 10.44 12.62 16.11 10.42 11.95
RANTES/
CCL5 0.65 1.05 0.85 0.93 0.86 1.00 5.62 5.49 10.37 10.05 7.66
MCP-1/
CCL2 0.61 0.65 1.17 0.63 0.49 0.48 4.23 7.50 11.11 8.54 7.14
MIP-1a/
CCL3 0.93 1.61 1.31 2.44 1.44 1.71 8.25 8.25 8.25 8.25 8.25
IL-1a 0.55 0.42 0.43 0.45 0.58 0.38 4.50 6.14 7.51 8.10 4.61
IL-10 0.66 0.73 0.78 2.31 0.93 0.66 9.80 13.70 14.36 7.25 8.91
TNF-a 0.89 1.34 1.15 1.16 1.01 1.41 5.39 6.33 7.61 6.64 6.01
LIF 0.99 1.01 1.19 0.99 0.99 0.99 7.53 10.55 9.98 5.05 8.15
MIP-2a/
CXCL2 0.89 1.04 0.86 1.27 0.95 1.39 4.84 4.84 4.84 4.84 4.84
IL-9 1.00 1.00 1.00 1.00 1.00 1.00 4.89 6.08 5.53 1.97 3.87
IL-23 0.93 0.93 0.93 0.92 0.92 0.92 1.55 1.66 1.52 1.30 1.48
IL-27 1.00 1.00 1.00 1.00 1.00 1.00 1.28 1.43 1.48 1.21 1.19
IL-15 1.00 1.06 1.58 1.15 1.28 1.00 1.68 1.20 1.02 1.10 1.29
M-CSF 1.00 1.00 1.00 1.00 1.00 1.00 1.10 1.27 1.35 1.05 1.01
(Continued)

TABLE 9 Continued

pgm– pgm– pgm– pgm− pgm- pgm− CO92 CO92
CO92 pPst– pPst– pPst– pPst− ×2 pPst- pPst− ×1; DyscN CO92 CO92 DyscN
pgm− x2 ×2 ×2 (day 46); ×1; F1V CO92 ×2 DyscN DyscN ×2 (day
×1 21) 46) 90) 90) 90) (day 90) 46) 90) (day 90) (day 90)
Cytokinea (Grp 2)b (Grp 3) (Grp 4) (Grp 5) (Grp 6) (Grp 7) (Grp 8) (Grp 9) (Grp 10) (Grp 11) (Grp 12)
IL-28 1.00 1.00 1.00 1.00 1.00 1.00 1.01 1.00 1.10 1.02 1.02
IL-31 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
Eotaxin/
CCL11 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
a
The cytokine results are shown as the ratio to the KPhos-vaccinated group and are based on the GM (pg/mL). Italicized and not bolded are not significant; bolded are significant (p < 0.05).
b
c
All vaccine groups containing CO92 DyscN reached the upper limit of detection for IFN-g resulting in a maximum fold change of 3,934.91 relative to KPhos.
vaccines was examined (Figure 3). Our results suggested that CpG most live vaccines designed to ameliorate plague) do not generate
enhanced protection afforded by the heterologous Combo LAV and a robust antibody response against secreted V antigen, there
rF1V vaccine against the Y. pestis C12 aerosol challenge (Figures 3B, appears to be enough V protein associated with the cell that can
C). Although CpG might not be required for maximal protection act as a prime vaccination against this antigen. Once boosted with
against the bubonic plague model described here, there may be rF1V (+/− CpG), there is an appreciable increase in anti-rV titers.
scenarios where the addition of CpG could improve disease These titers are greater than those observed in mice receiving a
outcome (Figure 3A). single dose of rF1V (+/− CpG), but this increase is not statistically
The impact of vaccination on accelerating bacterial clearance significant. This lack of statistical significance in anti-rV titers
or inhibition of dissemination in vivo was also determined by among groups 3, 4, 10, and 11 (Table 3, Figure 3B) is an important
quantitatively assessing bacterial loads in tissues 3 days after the point of discussion as it implies that there are other immunogens
aerosol challenge. The organ and blood levels of Y. pestis C12 in that were provided by the LAV prime vaccination that are
mice vaccinated with the heterologous vaccine (Combo LAV prime ultimately increasing vaccine efficacy in the mice (survival of
and rF1V + CpG boost) were the lowest compared with those in the groups 3 and 4 compared with groups 10 and 11 in Figure 3B),
other control and vaccinated groups. There appeared to be a and protection could not be correlated with anti-rV titers alone.
relationship between CFU and protection, i.e., lower CFU load in Accordingly, we continue to explore and characterize the immune
the lungs at day 3 was associated with vaccines’ protective efficacy, response generated by our LAVs in the mouse model.
albeit a statistical correlation was precluded due to low numbers of In a previous study with our LAVs (alone), cell-mediated
samples. Furthermore, the survival rates of the two groups immunity involving elevated Th17- and Th2-associated cytokines
vaccinated with Combo LAV plus rF1V (+/− CpG) were not was associated with protection against the encapsulated CO92
significantly different (90% and 50%, respectively), yet the strain (Cote et al., 2021). The findings were obtained using
mean TTM post-challenge of the group receiving the CpG splenocytes collected before the aerosol challenge. Although an
immunostimulant was increased significantly compared with that association with protection against the C12 strain was not clearly
of most other groups. observed with stimulated splenocytes collected pre-exposure from
We also evaluated the associations between the extent of the extended boost time study (Figure 4), none of the vaccines
protection and the immune responses elicited by the vaccine. included CpG and none were significantly protective in this high-
The heterologous vaccination groups with the highest survival dose challenge study. In contrast, with some exceptions, protection
rates had the greatest serum titers to rF1V and rV and the CO92 was associated with overall lower expression levels of cytokines in
whole-cell antigens, and their splenocytes secreted higher levels of tissue extracts obtained after challenge with C12 in heterologously
IFN-g after rF1V stimulation in ELISpot assays. These findings vaccinated mice relative to KPhos-vaccinated mice. This controlled
confirm the strong immunogenicity of rF1V in BALB/c mice as cytokine response potentially mitigates excessive Th17 cell and
demonstrated in our initial vaccine studies using the live neutrophil migration (CXCL1, CXCL2) as well as activation (G-
attenuated plague strains (Cote et al., 2021); full protection CSF, IL-17A, IL-1b, IL-6, IL-22) which may contribute to reduced
against F1 + Y. pestis CO92 required induction of a strong inflammation-related destruction of lung tissue (Table 4). In
humoral anti-F1 response. The consistently greater protection addition, the inability to control the infection may predispose
against aerosol exposure to Y. pestis C12 of the heterologous, mice to excessive alveolar macrophage and eosinophil activation
rF1V-containing vaccines compared with homologous single or and recruitment via colony-stimulating factors (G-CSF, M-CSF,
Combo LAVs (Figures 3B, C, 4) supports a role for rF1V- GM-CSF). Excessive GM-CSF production by lung epithelium may
associated immune responses in vaccine efficacy against C12. also be a direct correlate of lung damage and act as a marker of futile
Importantly, although the DyscN-based vaccines (and in general efforts to repair and restore epithelial barrier function (Trapnell and

FIGURE 6
The protection of BALB/c mice against lethal aerosol challenge with Yersinia pestis strain C12 by a vaccination and post-challenge antibiotic
treatment strategy: the impact on survival of antibiotic dose and time of treatment. In each experiment, two groups of mice were vaccinated twice
(20 days apart) with mutant CO92 DyscN (0.92 × 107 CFU and 0.48 × 107 CFU/mouse, respectively), and two groups were unvaccinated. All mice
were challenged with aerosolized Y. pestis C12 28 days after the boost dose. The challenge doses were administered in two iterations, with doses of
5.6 × 105 CFU/mouse (6 LD50s) for the vaccinated groups and 3.11 × 105 CFU (4 LD50s) for unvaccinated mice. The mice were treated with either 20
mg/kg of streptomycin starting at 48 h (A) or 60 h (B) after challenge or 40 mg/kg starting at 60 h after challenge (C). n = 8 mice/group (KPhos
alone and vaccine alone) or 10 mice/group (streptomycin alone and vaccine + streptomycin).
Whitsett, 2002; Standiford et al., 2012; Chen et al., 2016; Rosler and groups. Controlled upregulation of IFN-g, IL-17A, IL-2, and IL-4 in
Herold, 2016). the CO92 pgm− pPst− + rF1V-vaccinated mice appeared to correlate
Thus, we intend to investigate the protective role for vaccine- with protection, while controlled upregulation of Th2- (IL-4, IL-5,
associated dampening of the cytokine storm induced by exposure to IL-13) and Th17 (IL-17A, IL-22) along with the pleiotropic IL-2
virulent F1+ and F1− Y. pestis strains. We predict that a fine-tuned and IL-3 cytokines was associated with CO92 DyscN + rF1V-
modulation between a protective and damaging extent of cytokine vaccinated mice. It is important to note that our cytokine data are
expression is required for optimal vaccine efficacy. Cytokine collected from a single time point and these expression levels and
expression which was either excessively high or low was their correlation with protection may differ throughout the course
associated with poor protection, as suggested for example by the of infection. However, the anti-rF1V and anti-CO92 antibody
cytokine profiles of restimulated splenocytes described in Table 9. responses were the highest in the two most protective vaccines
Of the two most protective vaccine groups (CO92 pgm− pPst− + and may be used as correlates of protection. The cytokine response
rF1V and CO92 DyscN + rF1V) that afforded the same level of against TS C12 was much more pronounced with vaccines that
protection, there appeared to be a distinct immune response that included CO92 DyscN relative to CO92 pgm− pPst−, and yet the level
induced a relatively low anti-TS C12 antibody response in both of protection was the same. Following rF1V stimulation, the

FIGURE 7
The protection of BALB/c mice against lethal aerosol challenge with Yersinia pestis strain C12 by a vaccination and post-challenge antibiotic
treatment strategy. The mice were vaccinated twice (23 days apart) with Y. pestis C12 DyscN (0.48 × 107 and 0.70 × 107 CFU for doses 1 and 2,
respectively); controls received KPhos alone. The mice were challenged 28 days later with 1.67 × 106 CFU (22 LD50s) of aerosolized Y. pestis C12.
They were either untreated or treated with 20 mg/kg of streptomycin 60 h after challenge with Y. pestis strain C12 4 weeks after the boost dose.
n = 9 mice/group (KPhos alone and vaccine alone) or 10 mice/group (streptomycin alone and vaccine + streptomycin).
number of IFN-g-secreting splenocytes after 24 h was relatively high basis of this observation has not been fully discerned but may be
in all rF1V vaccine groups, but after 48 h, the levels of IFN-g were related to the V-producing strain of live vaccine and possibly the
substantially higher in all CO92 DyscN vaccine groups, especially in animal host. The potential role of the immune-modulatory activity of
groups that were also administered rF1V. The short half-life of IFN- V in the suppression of the host immune responses in the context of
g may be an indicator of this large discrepancy between the two time LAVs remains to be clarified (Leary et al., 1995; Brubaker, 2003;
points for the different vaccine formulations, such that the CO92 Depaolo et al., 2008; Quenee et al., 2011; Feodorova and Motin, 2012;
pgm− pPst− + rF1V vaccine group produces a strong yet transient Daniel et al., 2016; Feodorova et al., 2020). However, our data suggest
response that recedes by 48 h relative to the CO92 DyscN + rF1V that a LAV followed by an rF1V booster can overcome the low V titers
that appears to be longer in duration and more intense (Lortat- typically observed with LAVs and this heterologous strategy may
Jacob et al., 1996). Furthermore, TS C12 antigen stimulated a represent a superior vaccination strategy.
greater IFN-g response in splenocytes from mouse groups To maximize protection against infection, we also compared the
vaccinated with CO92 DyscN mutant (+/− rF1V: groups 9–12) efficacies of pre-exposure vaccination and post-exposure antibiotic
and not CO92 pgm− pPst− (+/− rF1V: groups 2–7). treatment, alone and combined, against inhalational challenge with Y.
In addition to the enhanced efficacy and potential modularity of pestis C12 strains. A layered defense strategy employing vaccination
our combination vaccine approach against both F1+ and F1− Y. with the DyscN mutant of the wild-type or F1-deficient strain,
pestis strains, an advantage inherent with the DyscN mutant LAVs is separately or combined, and treatment with streptomycin provided
their inclusion of antigens some of which are not expressed in other optimal protection and was clearly synergistic (Figures 6, 7). These
live vaccines, i.e., potential antigens encoded by the chromosomal promising results predict that the “pre–post” strategy will extend the
102-kb pgm locus, by other Y. pestis chromosomal genes, and/or by time post-exposure when antibiotic administration must be initiated
plasmids pMT1 (pFra) and pPcP1 (pPst). However, the CO92 pgm− to be successful. Our ongoing studies focus on assessing the limits of
pPst− strain was also highly protective, possibly due in part to an this strategy as well as its impact on vaccination with the heterologous
immune response to the secreted T3SS Yops. LAV/subunit constructs.
The V antigen is an essential component of the Y. pestis T3SS and The layered countermeasure strategy offers several attractive
an indispensable virulence factor of Y. pestis; thus, it is an obvious advantages to single medical countermeasure strategies. As
target for countermeasure development. However, although V-specific observed in animal studies, the efficacy of antibiotic treatment of
immunity has been correlated with protection, the adaptive immune Warfighters exposed to a bolus of aerosolized Y. pestis is predicted to
responses to V are complicated and may not be essential to an effective be enhanced in prophylactically vaccinated compared with
vaccine. In our previous work (Cote et al., 2021), the DyscN vaccines unvaccinated individuals. Vaccination is also expected to extend
did not induce a significant anti-V antibody response, as expected due the window of time post-exposure during which antibiotics are
to the requirement of the YscN ATPase for a functional T3SS (Bozue effective, i.e., before reaching the “point of no return.”
et al., 2012). In this current work, we noted a significant increase in Concomitantly, for vaccinees with waning immunity, post-exposure
anti-rV titers when the F1-V subunit vaccine was used as a booster. antibiotic treatment could potentially enhance the protective effects of
However, the lack of a robust anti-V response in animals or humans the vaccine. This might be especially important in a biothreat
vaccinated with live attenuated F1+ and F1− Y. pestis strains has been scenario where the limited time before potential exposure allowed
documented repeatedly (Williamson et al., 1995; Quenee et al., 2008; only a single administration of a plague vaccine. Finally, strategies
Brasiale et al., 2009; Qiu et al., 2010; Bozue et al., 2012; Sun et al., 2014; employing layered or combination countermeasures afford vaccine
Demeure et al., 2019; Feodorova et al., 2020; Cote et al., 2021). The antigen/antibiotic dose-sparing and potentially allow the use of

second-line antibiotics as needed, e.g., in situations involving IACUC. The studies were conducted in accordance with the local
naturally or engineered antibiotic-resistant challenge strains legislation and institutional requirements. Written informed consent
(Zauberman et al., 2019; Klimko et al., 2022b). was not obtained from the owners for the participation of their
Additional studies which evaluated combinations of vaccines and animals in this study because the mice were purchased for research.
therapeutics against critical pathogens have been reported. In a
mouse intranasal challenge model of pneumonic plague,
Zauberman and coworkers showed that post-exposure vaccination
with the live pgm− Y. pestis strain EV76 enhanced the efficacy of Author contributions
antibiotic treatment against the virulent Kim53 strain of Y. pestis.
Optimal protection was achieved when the vaccine was administered SB and CC designed and supervised the project. SB, CK, JD, RT, JS,
MH, NR, YT, MD, JB, and CC prepared the reagents, treatments, and
subcutaneously at the same time as an intranasal virulent Y. pestis
challenge, and antibiotic treatment was begun 48 h after challenge vaccines and performed the experiments. SB, CK, MD, JQ, DF, SW,
and CC performed the data analyses. SB, SW and CC wrote the first
(Zauberman et al., 2019). Also, such combination countermeasure
strategies are not restricted to one pathogen or model. The potential draft. SB, JB, SW, and CC revised and edited the manuscript. All
authors contributed to the article and approved the submitted version.
broad applicability of a similar protection strategy was exemplified in
our recent studies with Burkholderia pseudomallei, the etiologic agent
of melioidosis (Klimko et al., 2022a). Burkholderia pseudomallei is
intrinsically resistant to many commonly used antibiotics, and Funding
infections with it are notoriously difficult to eradicate. Effective
antibiotic treatment regimens are complex and of long duration. This work was funded by the US Defense Threat Reduction
Klimko et al. demonstrated that when used in a layered approach, Agency (DTRA) under projects CB10392 and CB10794.
established vaccine strategies employing recombinant protein subunit
conjugates or a LAV strain, and the antibiotic co-trimoxazole,
produced high levels of protection in mice.
Further efforts are being pursued to optimize the vaccination and Conflict of interest
integrated prophylactic/post-exposure countermeasure strategies.
These include a determination of the latest time post-exposure to The authors declare that the research was conducted in the
aerosolized bacteria when antibiotic treatment can rescue the mice and absence of any commercial or financial relationships that could be
the lowest effective dose of antibiotic; the optimal heterologous vaccine construed as a potential conflict of interest.
composition, order of delivery, and schedule; the most synergistic The authors CC and JB declared that they were an editorial
antibiotic(s) for various vaccine strategies; and, in the context of single board member of Frontiers, at the time of submission. This had no
vaccination, the longest interval between vaccination and challenge impact on the peer review process and the final decision.
after which the antibiotic treatment remains more effective than in
unvaccinated treated individuals. BALB/c mice are an appropriate
early model for both pathogenesis research and medical Publisher’s note
countermeasure development; however, due to differences in LD50
estimations and other inherent differences, their recapitulation of All claims expressed in this article are solely those of the
human disease remains equivocal, and thus, follow-on work will authors and do not necessarily represent those of their affiliated
also be required in a nonhuman primate model of pneumonic organizations, or those of the publisher, the editors and the
plague. Additionally, plague vaccine candidates have generally been reviewers. Any product that may be evaluated in this article, or
evaluated using only one virulent strain. However, many distinct claim that may be made by its manufacturer, is not guaranteed or
strains of virulent Y. pestis have been isolated from multiple endorsed by the publisher.
geographical sources (Vogler et al., 2016), and analyses of vaccines
for their performance against such strains are needed.
Author disclaimer
Data availability statement Opinions, interpretations, conclusions, and recommendations
are those of the authors and are not necessarily endorsed by the U.S.
The raw data supporting the conclusions of this article will be Army or the Department of Defense Health Agency.
made available by the authors, without undue reservation.
Supplementary material
Ethics statement
The Supplementary Material for this article can be found online
The animal studies were approved by the United States Army at: https://www.frontiersin.org/articles/10.3389/fbrio.2023.1240698/
Medical Research Institute of Infectious Diseases (USAMRIID) full#supplementary-material

References
Adamovicz, J. J., and Worsham, P. L. (2012). “Laboratory Animal Models of Plague,” protection conferred by an LcrV-secreting Lactococcus lactis in a pseudotuberculosis
in Biodefense: Research Methodology and Animal Models (Boca Raton, FL: CRC Press), mouse model. Front. Immunol. 10, 1830. doi: 10.3389/fimmu.2019.01830
113–145. Daniel, C., Titecat, M., Poiret, S., Cayet, D., Boutillier, D., Simonet, M., et al. (2016).
Aftalion, M., Tidhar, A., Vagima, Y., Gur, D., Zauberman, A., Holtzman, T., et al. Characterization of the protective immune response to Yersinia pseudotuberculosis
(2023). Rapid Induction of Protective Immunity against Pneumonic Plague by Yersinia infection in mice vaccinated with an LcrV-secreting strain of Lactococcus lactis. Vaccine
pestis Polymeric F1 and LcrV Antigens. Vaccines (Basel) 2, 11. doi: 10.3390/ 34, 5762–5767. doi: 10.1016/j.vaccine.2016.09.060
vaccines11030581 Davis, K. J., Fritz, D. L., Pitt, M. L., Welkos, S. L., Worsham, P. L., and Friedlander, A.
Alibek, K. (1999). Biohazard: The Chilling True Story of the Largest Covert Biological M. (1996). Pathology of experimental pneumonic plague produced by fraction 1-
Weapons Program in the World - Told from the Inside by the Man Who Ran It New positive and fraction 1-negative Yersinia pestis in African green monkeys
York (NY: Random House, Inc). (Cercopithecus aethiops). Arch. Pathol. Lab. Med. 120 (2), 156–163.
Amemiya, K., Meyers, J. L., Rogers, T. E., Fast, R. L., Bassett, A. D., Worsham, P. L., Demeure, C. E., Derbise, A., Guillas, C., Gerke, C., Cauchemez, S., Carniel, E., et al.
et al. (2009). CpG oligodeoxynucleotides augment the murine immune response to the (2019). Humoral and cellular immune correlates of protection against bubonic plague
Yersinia pestis F1-V vaccine in bubonic and pneumonic models of plague. Vaccine 27, by a live Yersinia pseudotuberculosis vaccine. Vaccine 37, 123–129. doi: 10.1016/
2220–2229. doi: 10.1016/j.vaccine.2009.02.016 j.vaccine.2018.11.022
Amemiya, K., Meyers, J. L., Trevino, S. R., Chanh, T. C., Norris, S. L., and Waag, D. Demidenko, E., and Miller, T. W. (2019). Statistical determination of synergy based
M. (2006). Interleukin-12 induces a Th1-like response to Burkholderia mallei and on Bliss definition of drugs independence. PloS One 14, e0224137. doi: 10.1371/
limited protection in BALB/c mice. Vaccine 24, 1413–1420. doi: 10.1016/ journal.pone.0224137
j.vaccine.2005.09.021 Depaolo, R. W., Tang, F., Kim, I., Han, M., Levin, N., Ciletti, N., et al. (2008). Toll-
Anderson, G. W. Jr., Worsham, P. L., Bolt, C. R., Andrews, G. P., Welkos, S. L., like receptor 6 drives differentiation of tolerogenic dendritic cells and contributes to
Friedlander, A. M., et al. (1997). Protection of mice from fatal bubonic and pneumonic LcrV-mediated plague pathogenesis. Cell Host Microbe 4, 350–361. doi: 10.1016/
plague by passive immunization with monoclonal antibodies against the F1 protein of j.chom.2008.09.004
Yersinia pestis. Am. J. Trop. Med. Hyg. 56, 471–473. doi: 10.4269/ajtmh.1997.56.471 Donavan, J. E., Ham, D., Fukui, G. M., and Surgalla, M. J. (1961). Role of the capsule
Andrews, G. P., Strachan, S. T., Benner, G. E., Sample, A. K., Anderson, G. W. Jr., of Pasteurella pestis in bubonic plague in the Guinea pig. J. Infect. Dis. 109, 154–157.
Adamovicz, J. J., et al. (1999). Protective efficacy of recombinant Yersinia outer proteins doi: 10.1093/infdis/109.2.154
against bubonic plague caused by encapsulated and nonencapsulated Yersinia pestis. Drozdov, I. G., Anisimov, A. P., Samoilova, S. V., Yezhov, I. N., Yeremin, S. A.,
Infect. Immun. 67, 1533–1537. doi: 10.1128/IAI.67.3.1533-1537.1999 Karlyshev, A. V., et al. (1995). Virulent non-capsulate Yersinia pestis variants
Anisimov, A. P., Dentovskaya, S. V., Panfertsev, E. A., Svetoch, T. E., Kopylov, P., constructed by insertion mutagenesis. J. Med. Microbiol. 42, 264–268. doi: 10.1099/
Segelke, B. W., et al. (2010). Amino acid and structural variability of Yersinia pestis 00222615-42-4-264
LcrV protein. Infect. Genet. Evol. 10, 137–145. doi: 10.1016/j.meegid.2009.10.003 Du, Y., Rosqvist, R., and Forsberg, A. (2002). Role of fraction 1 antigen of Yersinia
Baril, L., Valles, X., Stenseth, N. C., Rajerison, M., Ratsitorahina, M., Pizarro-Cerda, pestis in inhibition of phagocytosis. Infect. Immun. 70, 1453–1460. doi: 10.1128/
J., et al. (2019). Can we make human plague history? A call to action. BMJ Glob Health IAI.70.3.1453-1460.2002
4, e001984. doi: 10.1136/bmjgh-2019-001984 Du, Z., and Wang, X. (2016). Pathology and pathogenesis of Yersinia pestis. Adv.
Bertherat, E. (2016). “Plague around the world 2010–2015,” in Wkly Epidemiol Rec. Exp. Med. Biol. 918, 193–222. doi: 10.1007/978-94-024-0890-4_7
91 (8), 89–93. Elvin, S. J., and Williamson, E. D. (2004). Stat 4 but not Stat 6 mediated immune
Bevins, S. N., Chandler, J. C., Barrett, N., Schmit, B. S., Wiscomb, G. W., and Shriner, mechanisms are essential in protection against plague. Microb. Pathog. 37, 177–184.
S. A. (2021). Plague exposure in mamMalian wildlife across the western United States. doi: 10.1016/j.micpath.2004.06.009
Vector Borne Zoonotic Dis. 21, 667–674. doi: 10.1089/vbz.2020.2765 Eppinger, M., Radnedge, L., Andersen, G., Vietri, N., Severson, G., Mou, S., et al.
Biryukov, S., Dankmeyer, J. L., Shamsuddin, Z., Velez, I., Rill, N. O., Rosario- (2012). Novel plasmids and resistance phenotypes in Yersinia pestis: unique plasmid
Acevedo, R., et al. (2021). Impact of Toll-Like Receptor-Specific Agonists on the Host inventory of strain Java 9 mediates high levels of arsenic resistance. PLoS One 7 (3),
Immune Response to the Yersinia pestis Plague rF1V Vaccine. Front. Immunol. 12, e32911. doi: 10.1371/journal.pone.0032911
726416. doi: 10.3389/fimmu.2021.726416 Feodorova, V. A., Lyapina, A. M., Khizhnyakova, M. A., Zaitsev, S. S., Saltykov, Y. V.,
Bozue, J., Cote, C. K., Chance, T., Kugelman, J., Kern, S. J., Kijek, T. K., et al. (2014). and Motin, V. L. (2020). Yersinia pestis Antigen F1 but Not LcrV Induced Humoral and
A Yersinia pestis tat mutant is attenuated in bubonic and small-aerosol pneumonic Cellular Immune Responses in Humans Immunized with Live Plague Vaccine-
challenge models of infection but not as attenuated by intranasal challenge. PloS One 9, Comparison of Immunoinformatic and Immunological Approaches. Vaccines
e104524. doi: 10.1371/journal.pone.0104524 (Basel). 19, 8. doi: 10.3390/vaccines8040698
Bozue, J., Cote, C. K., Webster, W., Bassett, A., Tobery, S., Little, S., et al. (2012). A Feodorova, V. A., and Motin, V. L. (2012). Plague vaccines: current developments
Yersinia pestis YscN ATPase mutant functions as a live attenuated vaccine against and future perspectives. Emerg. Microbes Infect. 1 (11), e36. doi: 10.1038/emi.2012.34
bubonic plague in mice. FEMS Microbiol. Lett. 332, 113–121. doi: 10.1111/j.1574- Ferber, D. M., and Brubaker, R. R. (1979). Mode of action of pesticin: N-
6968.2012.02583.x acetylglucosaminidase activity. J. Bacteriol. 139, 495–501. doi: 10.1128/jb.139.2.495-
Brasiale, V., Nash, M., Sinha, N., Zudina, I., and Motin, V. (2009). “Correlates of 501.1979
Immunity Elicited by Live Yersinia pestis Vaccine,” in NIH volume 1, Frontiers in Fetherston, J. D., Schuetze, P., and Perry, R. D. (1992). Loss of the pigmentation
Research Totawa (NJ: Humana Press), 473–480. phenotype in Yersinia pestis is due to the spontaneous deletion of 102 kb of
Brubaker, R. R. (2003). Interleukin-10 and inhibition of innate immunity to chromosomal DNA which is flanked by a repetitive element. Mol. Microbiol 6 (18),
Yersiniae: roles of Yops and LcrV (V antigen). Infect. Immun. 71, 3673–3681. doi: 2693-704. doi: 10.1111/j.1365-2958.1992.tb01446.x
10.1128/IAI.71.7.3673-3681.2003 Friedlander, A. M., Welkos, S. L., Worsham, P. L., Andrews, G. P., Heath, D. G.,
Cabanel, N., Bouchier, C., Rajerison, M., and Carniel, E. (2018). Plasmid-mediated Anderson, G. W. Jr., et al. (1995). Relationship between virulence and immunity as
doxycycline resistance in a Yersinia pestis strain isolated from a rat. Int. J. antimicrobial. revealed in recent studies of the F1 capsule of Yersinia pestis. Clin. Infect. Dis. Suppl 2,
agents. 51, 249–254. doi: 10.1016/j.ijantimicag.2017.09.015 S178–S181. doi: 10.1093/clinids/21.supplement_2.s178
Castiglione, F., Mantile, F., De Berardinis, P., and Prisco, A. (2012). How the interval Guiyoule, A., Gerbaud, G., Buchrieser, C., Galimand, M., Rahalison, L., Chanteau, S.,
between prime and boost injection affects the immune response in a computational et al. (2001). Transferable plasmid-mediated resistance to streptomycin in a clinical
model of the immune system. Comput. Math Methods Med. 2012, 842329. doi: 10.1155/ isolate of Yersinia pestis. Emerging Infect. Dis. 7, 43. doi: 10.3201/eid0701.010106
2012/842329 Heath, D. G., Anderson, G. W. Jr., Mauro, J. M., Welkos, S. L., Andrews, G. P.,
Chen, G. H., Teitz-Tennenbaum, S., Neal, L. M., Murdock, B. J., Malachowski, A. N., Adamovicz, J., et al. (1998). Protection against experimental bubonic and pneumonic
Dils, A. J., et al. (2016). Local GM-CSF-dependent differentiation and activation of plague by a recombinant capsular F1-V antigen fusion protein vaccine. Vaccine 16,
pulmonary dendritic cells and macrophages protect against progressive cryptococcal 1131–1137. doi: 10.1016/S0264-410X(98)80110-2
lung infection in mice. J. Immunol. 196, 1810–1821. doi: 10.4049/jimmunol.1501512 Heine, H. S., Louie, A., Sorgel, F., Bassett, J., Miller, L., Sullivan, L. J., et al. (2007).
Cote, C. K., Biryukov, S. S., Klimko, C. P., Shoe, J. L., Hunter, M., Rosario-Acevedo, Comparison of 2 antibiotics that inhibit protein synthesis for the treatment of infection
R., et al. (2021). Protection Elicited by Attenuated Live Yersinia pestis Vaccine Strains with Yersinia pestis delivered by aerosol in a mouse model of pneumonic plague.
against Lethal Infection with Virulent Y. pestis. Vaccines (Basel). 16, 9. doi: 10.3390/ J. Infect. Dis. 196 (5), 782–7. doi: 10.1086/520547
vaccines9020161 Hinnebusch, B. J., Rosso, M. L., Schwan, T. G., and Carniel, E. (2002). High-
Cowdrey, A. E. (1984). "Germ warfare" and public health in the Korean conflict. frequency conjugative transfer of antibiotic resistance genes to Yersinia pestis in the flea
J. Hist Med. Allied Sci. 39, 153–172. doi: 10.1093/jhmas/39.2.153 midgut. Mol. Microbiol. 46, 349–354. doi: 10.1046/j.1365-2958.2002.03159.x
Daniel, C., Dewitte, A., Poiret, S., Marceau, M., Simonet, M., Marceau, L., et al. Klimko, C. P., Shoe, J. L., Rill, N. O., Hunter, M., Dankmeyer, J. L., Talyansky, Y.,
(2019). Polymorphism in the Yersinia LcrV antigen enables immune escape from the et al. (2022a). Layered and integrated medical countermeasures against Burkholderia

pseudomallei infections in C57BL/6 mice. Front. Microbiol. 13, 965572. doi: 10.3389/ Rosario-Acevedo, R., Biryukov, S. S., Bozue, J. A., and Cote, C. K. (2021). Plague
fmicb.2022.965572 prevention and therapy: perspectives on current and future strategies. Biomed. Oct 9
Klimko, C. P., Welkos, S. L., Shoe, J. L., Mou, S., Hunter, M., Rill, N. O., et al. (2022b). (10), 1421. doi: 10.3390/biomedicines9101421
Efficacy of Treatment with the Antibiotic Novobiocin against Infection with Bacillus Rosenzweig, J. A., Hendrix, E. K., and Chopra, A. K. (2021). Plague vaccines: new
anthracis or Burkholderia pseudomallei. Antibio. (Basel). 23, 11. doi: 10.3390/ developments in an ongoing search. Appl. Microbiol. Biotechnol. 105, 4931–4941. doi:
antibiotics11121685 10.1007/s00253-021-11389-6
Kon, E., Levy, Y., Elia, U., Cohen, H., Hazan-Halevy, I., Aftalion, M., et al. (2023). A Rosler, B., and Herold, S. (2016). Lung epithelial GM-CSF improves host defense
single-dose F1-based mRNA-LNP vaccine provides protection against the lethal plague function and epithelial repair in influenza virus pneumonia-a new therapeutic strategy?
bacterium. Sci. Adv. 9, eadg1036. doi: 10.1126/sciadv.adg1036 Mol. Cell Pediatr. 3, 29. doi: 10.1186/s40348-016-0055-5
Leary, S. E., Williamson, E. D., Griffin, K. F., Russell, P., Eley, S. M., and Titball, R. W. Samoilova, S. V., Samoilova, L. V., Yezhov, I. N., Drozdov, I. G., and Anisimov, A. P.
(1995). Active immunization with recombinant V antigen from Yersinia pestis protects (1996). Virulence of pPst+ and pPst- strains of Yersinia pestis for Guinea-pigs. J. Med.
mice against plague. Infect. Immun. 63, 2854–2858. doi: 10.1128/iai.63.8.2854- Microbiol. 45, 440–444. doi: 10.1099/00222615-45-6-440
2858.1995 Sebbane, F., Jarrett, C., Gardner, D., Long, D., and Hinnebusch, B. J. (2009). The
Lortat-Jacob, H., Baltzer, F., and Grimaud, J. A. (1996). Heparin decreases the blood Yersinia pestis caf1M1A1 fimbrial capsule operon promotes transmission by flea bite in
clearance of interferon-gamma and increases its activity by limiting the processing of its a mouse model of bubonic plague. Infect. Immun. 77, 1222–1229. doi: 10.1128/
carboxyl-terminal sequence. J. Biol. Chem 271 (27), 16139–161343. doi: 10.1074/ IAI.00950-08
jbc.271.27.16139 Sha, J., Endsley, J. J., Kirtley, M. L., Foltz, S. M., Huante, M. B., Erova, T. E., et al.
McDonough, K. A., and Falkow, S. (1989). A Yersinia pestis-specific DNA fragment (2011). Characterization of an F1 deletion mutant of Yersinia pestis CO92, pathogenic
encodes temperature-dependent coagulase and fibrinolysin-associated phenotypes. role of F1 antigen in bubonic and pneumonic plague, and evaluation of sensitivity and
Mol. Microbiol. 3, 767–775. doi: 10.1111/j.1365-2958.1989.tb00225.x specificity of F1 antigen capture-based dipsticks. J. Clin. Microbiol. 49, 1708–1715. doi:
Meka-Mechenko, T. V. (2003). F1-negative natural Y. pestis strains. Adv. Exp. Med. 10.1128/JCM.00064-11
Biol. 529, 379–81. doi: doi: 10.1007/0-306-48416-1_76 Smiley, S. T. (2008). Current challenges in the development of vaccines for
Meyer, K. F. (1970). Effectiveness of live or killed plague vaccines in man. Bull. World pneumonic plague. Expert Rev. Vaccines 7, 209–221. doi: 10.1586/14760584.7.2.209
Health Organ 42 (5), 653–666. Sodeinde, O. A., and Goguen, J. D. (1988). Genetic analysis of the 9.5-kilobase
Miller, N. C., Quenee, L. E., Elli, D., Ciletti, N. A., and Schneewind, O. (2012). virulence plasmid of Yersinia pestis. Infect. Immun. 56, 2743–2748. doi: 10.1128/
Polymorphisms in the lcrV gene of Yersinia enterocolitica and their effect on plague iai.56.10.2743-2748.1988
protective immunity. Infect. Immun. 80, 1572–1582. doi: 10.1128/IAI.05637-11 Standiford, L. R., Standiford, T. J., Newstead, M. J., Zeng, X., Ballinger, M. N.,
Mizel, S. B., Graff, A. H., SrIranganathan, N., Ervin, S., Lees, C. J., Lively, M. O., et al. Kovach, M. A., et al. (2012). TLR4-dependent GM-CSF protects against lung injury in
(2009). Flagellin-F1-V fusion protein is an effective plague vaccine in mice and two Gram-negative bacterial pneumonia. Am. J. Physiol. Lung Cell Mol. Physiol. 302, L447–
species of nonhuman primates. Clin. Vaccine Immunol. 16, 21–28. doi: 10.1128/ L454. doi: 10.1152/ajplung.00415.2010
CVI.00333-08 Stepanov, A. V., Marinin, L. I., and Vorob'ev, A. A. (1999). Aerosol vaccination
Mueller, C. A., Broz, P., Muller, S. A., Ringler, P., Erne-Brand, F., Sorg, I., et al. against dangerous infectious diseases. Vestn Ross Akad Med. Nauk. (8), 47–54.
(2005). The V-antigen of Yersinia forms a distinct structure at the tip of injectisome Sun, W. (2016). Plague vaccines: status and future. Adv. Exp. Med. Biol. 918, 313–
needles. Science 310, 674–676. doi: 10.1126/science.1118476 360. doi: 10.1007/978-94-024-0890-4_12
National Research Council (2011). Guide laboratory animals for the care and use of Sun, W., Sanapala, S., Henderson, J. C., Sam, S., Olinzock, J., Trent, M. S., et al.
eighth edition committee for the update of the guide for the care and use of laboratory (2014). LcrV delivered via type III secretion system of live attenuated Yersinia
animals institute for laboratory animal research division on earth and life studies. pseudotuberculosis enhances immunogenicity against pneumonic plague. Infect.
Washington, DC: The National Academies Press. Immun. 82, 4390–4404. doi: 10.1128/IAI.02173-14
Nelson, C. A., Meaney-Delman, D., Fleck-Derderian, S., Cooley, K. M., Yu, P. A., Swietnicki, W., Carmany, D., Retford, M., Guelta, M., Dorsey, R., Bozue, J., et al.
Mead, P. S., et al. (2021). Antimicrobial treatment and prophylaxis of plague: (2011). Identification of small-molecule inhibitors of Yersinia pestis Type III secretion
recommendations for naturally acquired infections and bioterrorism response. system YscN ATPase. PloS One 6, e19716. doi: 10.1371/journal.pone.0019716
MMWR Recomm Rep. 70, 1–27. doi: 10.15585/mmwr.rr7003a1
Titball, R. W., and Williamson, E. D. (2004). Yersinia pestis (plague) vaccines. Expert
Pendrak, M. L., and Perry, R. D. (1991). Characterization of a hemin-storage locus of Opin. Biol. Ther. 4, 965–973. doi: 10.1517/14712598.4.6.965
Yersinia pestis. Biol. Met. 4, 41–47. doi: 10.1007/BF01135556
Trapnell, B. C., and Whitsett, J. A. (2002). Gm-CSF regulates pulmonary surfactant
Perry, R. D., Pendrak, M. L., and Schuetze, P. (1990). Identification and cloning of a homeostasis and alveolar macrophage-mediated innate host defense. Annu. Rev.
hemin storage locus involved in the pigmentation phenotype of Yersinia pestis. Physiol. 64, 775–802. doi: 10.1146/annurev.physiol.64.090601.113847
J. Bacteriol. 172, 5929–5937. doi: 10.1128/jb.172.10.5929-5937.1990
Trevino, S. R., Klimko, C. P., Reed, M. C., Aponte-Cuadrado, M. J., Hunter, M., Shoe,
Pettersson, J., Holmstrom, A., Hill, J., Leary, S., Frithz-Lindsten, E., von Euler-Matell, J. L., et al. (2018). Disease progression in mice exposed to low-doses of aerosolized
A., et al. (1999). The V-antigen of Yersinia is surface exposed before target cell contact clinical isolates of Burkholderia pseudomallei. PloS One 13, e0208277. doi: 10.1371/
and involved in virulence protein translocation. Mol. Microbiol. 32, 961–976. doi: journal.pone.0208277
10.1046/j.1365-2958.1999.01408.x
Verma, S. K., and Tuteja, U. (2016). Plague vaccine development: current research
Pitt, L. M. (2004). “Non-human primates as a model for pneumonic plague; animal
and future trends. Front. Immunol. 7, 602. doi: 10.3389/fimmu.2016.00602
models and correlates of protection for plague,” in Proceedings of the Plague Vaccines
Workshop (Gaithersburg, MD, USA). Available at: https://www.fda.gov/cber/minutes/ Vogler, A. J., Keim, P., and Wagner, D. M. (2016). A review of methods for subtyping
plague101304t.pdf. Yersinia pestis: From phenotypes to whole genome sequencing. Infect. Genet. Evol. 37,
Powell, B. S., Andrews, G. P., Enama, J. T., Jendrek, S., Bolt, C., Worsham, P., et al. 21–36. doi: 10.1016/j.meegid.2015.10.024
(2005). Design and testing for a nontagged F1-V fusion protein as vaccine antigen Wang, X., Singh, A., and Sun, W. (2020). Protection and Safety Evaluation of Live
against bubonic and pneumonic plague. Biotechnol. Prog. 21, 1490–1510. doi: 10.1021/ Constructions Derived from the Pgm– and pPCP1– Yersinia pestis Strain. Vaccines 8,
bp050098r 1–14. doi: 10.3390/vaccines8010095
Qiu, Y., Liu, Y., Qi, Z., Wang, W., Kou, Z., Zhang, Q., et al. (2010). Comparison of Wang, X., Zhang, X., Zhou, D., and Yang, R. (2013). Live-attenuated Yersinia pestis
immunological responses of plague vaccines F1+rV270 and EV76 in Chinese-origin vaccines. Expert Rev. Vaccines 12, 677–686. doi: 10.1586/erv.13.42
rhesus macaque, Macaca mulatta. Scand. J. Immunol. 72, 425–433. doi: 10.1111/j.1365-
Weening, E. H., Cathelyn, J. S., Kaufman, G., Lawrenz, M. B., Price, P., Goldman, W. E.,
3083.2010.02456.x
et al. (2011). The dependence of the Yersinia pestis capsule on pathogenesis is influenced by
Quenee, L. E., Ciletti, N. A., Elli, D., Hermanas, T. M., and Schneewind, O. (2011). the mouse background. Infect. Immun. 79, 644–652. doi: 10.1128/IAI.00981-10
Prevention of pneumonic plague in mice, rats, Guinea pigs and non-human primates
with clinical grade rV10, rV10-2 or F1-V vaccines. Vaccine 29, 6572–6583. doi: Welkos, S. L., Andrews, G. P., Lindler, L. E., Snellings, N. J., and Strachan, S. D.
10.1016/j.vaccine.2011.06.119 (2004). Mu dI1(Ap lac) mutagenesis of Yersinia pestis plasmid pFra and identification
of temperature-regulated loci associated with virulence. Plasmid 51, 1–11. doi: 10.1016/
Quenee, L. E., Cornelius, C. A., Ciletti, N. A., Elli, D., and Schneewind, O. (2008). j.plasmid.2003.09.003
Yersinia pestis caf1 variants and the limits of plague vaccine protection. Infect. Immun.
76, 2025–2036. doi: 10.1128/IAI.00105-08 Welkos, S. L., Davis, K. M., Pitt, L. M., Worsham, P. L., and Freidlander, A. M.
(1995). Studies on the contribution of the F1 capsule-associated plasmid pFra to the
Randremanana, R., Andrianaivoarimanana, V., Nikolay, B., Ramasindrazana, B., virulence of Yersinia pestis. Contrib Microbiol. Immunol. 13, 299–305.
Paireau, J., Ten Bosch, Q. A., et al. (2019). Epidemiological characteristics of an urban
plague epidemic in Madagascar, August-November 2017: an outbreak report. Lancet Welkos, S. L., Friedlander, A. M., and Davis, K. J. (1997). Studies on the role of
Infect. Dis. 19, 537–545. doi: 10.1016/S1473-3099(18)30730-8 plasminogen activator in systemic infection by virulent Yersinia pestis strain C092.
Microb. Pathog. 23, 211–223. doi: 10.1006/mpat.1997.0154
Roggenkamp, A., Geiger, A. M., Leitritz, L., Kessler, A., and Heesemann, J. (1997).
Passive immunity to infection with Yersinia spp. mediated by anti-recombinant V Welkos, S., Pitt, M. L., Martinez, M., Friedlander, A., Vogel, P., and Tammariello, R.
antigen is dependent on polymorphism of V antigen. Infect. Immun. 65, 446–451. doi: (2002). Determination of the virulence of the pigmentation-deficient and
10.1128/iai.65.2.446-451.1997 pigmentation-/plasminogen activator-deficient strains of Yersinia pestis in non-

human primate and mouse models of pneumonic plague. Vaccine 20, 2206–2214. doi: Winter, C. C., Cherry, W. B., and Moody, M. D. (1960). An unusual strain of
10.1016/S0264-410X(02)00119-6 Pasteurella pestis isolated from a fatal human case of plague. Bull. World Health Organ.
Williams, J. E., and Cavanaugh, D. C. (1984). Potential for rat plague from 23 (2), 408–409.
nonencapsulated variants of the plague bacillus (Yersinia pestis). Experientia 40, Worsham, P. L., McGovern, T. W., Vietri, N. J., and Bozue, J. (2018). “Plague,” in
739–740. doi: 10.1007/BF01949752 Medical Aspects of Biological Warfare (Ft. Sam Houston, TX: The Borden Institute),
Williams, P., and Wallace, D. (1989). Unit 731: Japan's Secret Biological Warfare in 247–284.
World War II (New York, NY: The Free Press). Worsham, P. L., Stein, M. P., and Welkos, S. L. (1995). Construction of defined F1
Williamson, E. D., Eley, S. M., Griffin, K. F., Green, M., Russell, P., Leary, S. E., et al. negative mutants of virulent Yersinia pestis. Contrib Microbiol. Immunol. 13, 325–328.
(1995). A new improved sub-unit vaccine for plague: the basis of protection. FEMS Zauberman, A., Gur, D., Levy, Y., Aftalion, M., Vagima, Y., Tidhar, A., et al. (2019).
Immunol. Med. Microbiol. 12, 223–230. doi: 10.1111/j.1574-695X.1995.tb00196.x Postexposure administration of a Yersinia pestis live vaccine for potentiation of second-
Williamson, E. D., and Oyston, P. C. (2013). Protecting against plague: towards a line antibiotic treatment against pneumonic plague. J. Infect. Dis. 220, 1147–1151. doi:
next-generation vaccine. Clin. Exp. Immunol. 172, 1–8. doi: 10.1111/cei.12044 10.1093/infdis/jiz260

Fbrio 02 1240698

Uploaded by

Copyright:

Available Formats

Fbrio 02 1240698

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fbrio 02 1240698

Uploaded by

Copyright:

Available Formats

TYPE Original Research

PUBLISHED 22 September 2023

Live attenuated vaccines and

Results: The most protective and immunogenic vaccination scheme, as tested

Frontiers in Bacteriology 01 frontiersin.org

plague, Yersinia pestis, vaccine, nonencapsulated, mice, bubonic, pneumonic

Frontiers in Bacteriology 02 frontiersin.org

Frontiers in Bacteriology 03 frontiersin.org

Four weeks following vaccination, puriﬁed splenocytes were

Frontiers in Bacteriology 04 frontiersin.org

Frontiers in Bacteriology 05 frontiersin.org

Frontiers in Bacteriology 06 frontiersin.org

Frontiers in Bacteriology 07 frontiersin.org

Frontiers in Bacteriology 08 frontiersin.org

Vaccine # Positive CFU/g or mLb

Spleen 6/6 3.98 × 107 (12.3)

Blood 5/5c 1.92 × 104 (5.9)

Combo LAV Combo LAV Lung 6/6 1.43 × 109 (1.9)

Spleen 6/6 2.07 × 106 (39.3)

Blood 6/6 4.27 × 103 (10.0)

Combo LAV rF1V Lung 6/6 2.36 × 108 (4.3)

Spleen 5/6 1.84 × 104 (1,305.3)

Combo LAV rF1V + CpG Lung 4/6 2.74 × 102 (1,232.6)

Spleen 2/6 1.23 × 102 (1,577.8)

Blood 1/6 2.37 × 102 (19.3)

KPhos rF1V Lung 6/6 5.29 × 106 (1,921.7)

Spleen 4/6 4.23 × 103 (2,960.8)

Blood 3/6 9.06 × 102 (20.9)

KPhos rF1V + CpG Lung 6/6 1.16 × 108 (396.5)

Spleen 5/6 1.74 × 105 (662.4)

Blood 4/5 3.21 × 103 (26.5)

Frontiers in Bacteriology 09 frontiersin.org

Vaccine Antibody titera

Frontiers in Bacteriology 10 frontiersin.org

IL-28 1.27 1.00 1.61 0.74 1.18

IL-10 1.24 0.99 1.55 0.83 0.99

RANTES (CCL5) 5.26 1.22 1.43 1.10 1.43

IL-23 1.28 1.01 1.31 1.24 1.00

IL-4 6.02 2.72 1.27 0.49 0.73

Eotaxin 3.39 1.11 1.27 0.78 1.46

IFN-a 1.00 1.00 1.00 1.00 1.00

IL-5 3.72 1.83 1.00 0.71 0.82

IL-15 2.64 1.24 0.87 0.91 1.42

IL-9 1.21 0.88 0.83 0.64 1.05

IL-2 2.17 1.23 0.77 0.61 0.94

IL-27 2.34 1.00 0.49 0.36 1.12

IL-13 2.13 1.07 0.48 0.37 0.71

ENA-78 (CXCL5) 1.29 0.83 0.47 0.28 0.82

IL-12p70 2.74 1.21 0.47 0.33 0.84

IL-31 1.61 1.19 0.41 1.28 1.28

M-CSF 1.95 1.30 0.38 0.67 1.16

IP-10 (CXCL10) 1.56 1.01 0.33 0.17 0.64

MCP-3 (CCL7) 1.29 0.92 0.30 0.17 0.67

IL-17A (CTLA-8) 1.79 1.62 0.24 0.08 0.65

MIP-1b (CCL4) 1.90 1.30 0.22 0.35 0.97

MCP-1 (CCL2) 1.29 0.57 0.19 0.13 0.91

IL-18 2.02 1.03 0.15 0.32 0.56

TNF-a 2.39 1.24 0.14 0.22 0.77

IL-22 3.86 2.37 0.13 0.18 0.80