Protection and Safety Evaluation of Live Constructions Derived from the Pgm− and pPCP1− Yersinia pestis Strain

Based on a live attenuated Yersinia pestis KIM10(pCD1Ap) strain (Pgm−, pPCP1−), we attempted to engineer its lipid A species to achieve improvement of immunogenicity and safety. A mutant strain designated as YPS19(pCD1Ap), mainly synthesizing the hexa-acylated lipid A, and another mutant strain designated as YPS20(pCD1Ap), synthesizing 1-dephosphalated hexa-acylated lipid A (detoxified lipid A), presented relatively low virulence in comparison to KIM10(pCD1Ap) by intramuscular (i.m.) or subcutaneous (s.c.) administration. The i.m. administration with either the KIM10(pCD1Ap) or YPS19(pCD1Ap) strain afforded significant protection against bubonic and pneumonic plague compared to the s.c. administration, while administration with completely attenuated YPS20(pCD1Ap) strain failed to afford significant protection. Antibody analysis showed that i.m. administration induced balanced Th1 and Th2 responses but s.c. administration stimulated Th2-biased responses. Safety evaluation showed that YPS19(pCD1Ap) was relatively safer than its parent KIM10(pCD1Ap) in Hfe−/− mice manifesting iron overload in tissues, which also did not impair its protection. Therefore, the immune activity of hexa-acylated lipid A can be harnessed for rationally designing bacteria-derived vaccines.


Introduction
The live EV76 vaccine is a spontaneous pgm mutant that has been used in humans for over 80 years in regions and countries of the former Soviet Union (FSU) without any deaths reported [1][2][3]. The EV76 vaccine conferred better protection against bubonic and pneumonic plague than killed vaccines in different animals, but it sometimes caused local and systemic reactions including fever, malaise, lymphadenopathy, erythema and large induration at the injection site [4][5][6][7], as well as disease in primates [6]. In addition, the live Pgm − strain (Y. pestis KIM5) retained virulence by intranasal (i.n.) or intravenous (i.v.) administration [6,[8][9][10], caused fatal septicemic plague in an individual with hereditary hemochromatosis manifesting as iron overload in tissues [11], and restored its virulence in hemojuvelin-knockout (Hjv −/− ) mice mimicking human hereditary hemochromatosis [12]. Thus, variable virulence of the live vaccine strain in animals and humans has deterred this vaccine from gaining worldwide acceptance [13,14]. Nevertheless, the large amount of accumulated evidence suggests that the live Pgm − vaccine strain is very close to becoming a human vaccine. The WHO 2018 plague workshop still listed the live attenuated Y. pestis vaccine as one of new-generation plague vaccines [15]. Thus, it is worthwhile to perfect this vaccine by resolving certain concerns.
One known strategy used by Y. pestis to evade host innate surveillance is to produce a tetra-acylated lipid A (the nonstimulatory form of LPS) that is not recognized by Toll-like receptor λ Red recombinase expression plasmid [34] pYA4373 The cat-sacB cassette in the PstI and SacI sites of pUC18. [35] pYA4577 The P lpxL lpxL gene fragment flanked by the lpxP upstream and downstream sequence [20] pYA4578 The cat-sacB-P lpxL lpxL gene fragment flanked by the lpxP upstream and downstream sequence [20] pYA4735 The P lpp lpxE gene fragment flanked by the lacI upstream and downstream sequence [21] pYA4736 The cat-sacB-P lpp lpxE gene fragment flanked by the lacI upstream and downstream sequence [21]

Mice
Animal care and experimental protocols were in accordance with the NIH "Guide for the Care and Use of the laboratory Animals" and were approved by the Institutional Animal Care and Use Committee at Albany Medical College (IACUC protocol # 17-02004). Swiss Webster outbred mice were purchased from Charles River Laboratories. Hfe −/− mice (B6.129S6-Hfe tm2Nca /J) [36] were purchased from The Jackson Laboratory. Wild-type (Hfe +/+ ) B6.129 mice as controls were purchased from Taconic. All deficient mice were bred at the animal facility of Albany Medical College.

Virulence Analysis in Mice
A single colony of each mutant strain was inoculated into HIB supplemented with ampicillin (100 µg/mL) for selection of plasmid pCD1Ap and grown overnight at 26 • C. Bacteria were diluted into 10 mL of fresh HIB enriched with 0.2% xylose and 2.5 mM CaCl 2 to obtain an OD 620 of 0.1 and then incubated at 26 • C for intramuscular (i.m.) or subcutaneous (s.c.) infection, or at 37 • C for intranasal (i.n.) infection, to reach an OD 620 of 0.6. The cells were then harvested, and the pellet resuspended in 1 mL of isotonic PBS and then adjusted to an appropriate concentration.
Groups of Swiss Webster mice (10/group, equal males and females) were administrated by i.m. or s.c. injection with 100 µL of bacterial suspension (10 7 CFU) or by i.n. route with 40 µL of bacterial suspension (10 6 CFU). Actual numbers of colony-forming units (CFU) inoculated were determined by plating serial dilutions onto TBA agar.

Determination of Protective Efficacy
Y. pestis strains were grown as described above. Two groups of Swiss Webster mice (10/group, equal males and females) were immunized i.m. with 1 × 10 7 CFU of a mutant strain in 100 µL of PBS on day 0 and boosted i.m. with 1 × 10 7 CFU on day 21. A group of mice (10/group) were injected with 100 µL of PBS as a control. Blood was collected by sub-mandibular vein puncture at 2-and 4-weeks post immunization. At 42 days after initial immunization, animals anesthetized with a 1:5 xylazine/ketamine mixture were challenged intranasally with 5 × 10 3 CFU of Y. pestis KIM6+(pCD1Ap) in 40 µL PBS. All infected animals were observed over a 15-day period.

Immune Responses
ELISA was used to assay serum IgG antibodies against Yersinia whole cell lysates (YpL) of Y. pestis KIM5+ or purified rLcrV protein [37]. Polystyrene 96-well flat-bottom microtiter plates (Dynatech Laboratories Inc., Chantilly, VA, USA) were coated with 1 µg/well of YpL protein or rLcrV. The procedures were the same as those described previously [37].

Evaluation of Vaccine Safety
Y. pestis mutant strains were evaluated for safety using 7 wk old Hfe −/− mice (n = 10/group). Mice were administered intramuscularly with a single dose of 1 × 10 7 CFU of each Y. pestis mutant strain. These mice were observed for signs of mortality and morbidity for 30 days.

Statistical Analysis
The Graph-Pad Prism 8.0 was used to analyze data statistically. The log-rank (Mantel-Cox) test, and the two-way ANOVA were used for survival analysis, statistical analyses of spleen weight and cytokine analysis, respectively. Data were expressed as means ± standard deviation (SD). A p-value < 0.05 was considered significant.

Construction of Y. pestis Mutants with Lipid A Modification
To exclude potential virulence caused by Pla, Y. pestis KIM10 (Pgm − pPCP1 − ) was used as a parent strain to generate mutant strains. Following our previous studies [20,21], we introduced the ∆lpxP::P lpxL lpxL mutation into Y. pestis KIM10 to generate the YPS19 (∆lpxP::P lpxL lpxL Pgm − pPCP1 − ) strain (Table 1 and Figure 1), in which lpxL genes from E. coli encodes the transferase that catalyzes the acyl-oxyacyl linkage of laurate to the 3' hydroxy-myristate to synthesize hexa-acylated lipid A (Table 1 and Figure 1). Then, the ∆lacI::P lpp lpxE gene fragment was introduced into YPS19 to replace the lacI gene and form the YPS20 (∆lpxP::P lpxL lpxL ∆lacI::P lpp lpxE Pgm − pPCP1 − ) strain (Table 1 and Figure 1), in which the lpxE gene encodes the lipid A 1-phosphatase from F. novicida that catalyzes removal of the 1-phosphate group from hexa-acylated lipid A to synthesize monophosphoryl lipid A ( Table 1 and Figure 1).

Protection Efficiency Against Pulmonary Y. pestis Infection and Serum Immune Responses
Regarding the above data, we established that the YPS20(pCD1Ap) strain showed obviously low immunogenicity compared with KIM10(pCD1Ap) or YPS19(pCD1Ap), and the i.m. inoculation with KIM10(pCD1Ap) or YPS19(pCD1Ap) afforded better protection against plague than the s.c. inoculation did and also provided complete protection against s.c. challenge with 50,000 LD 50 of Y. pestis KIM6+(pCD1Ap) ( Figure 3A) and significant protection against i.n. challenge with 50 LD 50 of Y. pestis KIM6+(pCD1Ap) ( Figure 3B). Therefore, we further evaluated protective efficacy of i.m. immunization with KIM10(pCD1Ap) and YPS19(pCD1Ap) by the prime-boost regime against pulmonary Y. pestis challenge. Groups of mice (5 male and 5 female) were immunized with 1 × 10 7 CFU of KIM10(pCD1Ap) or 2 × 10 7 CFU of YPS19(pCD1Ap) and boosted at 3 weeks post initial immunization. During immunization, two mice succumbed in the KIM10(pCD1Ap)-immunized group at day 11 post immunization ( Figure 5A). Then, all immunized mice were challenged intranasally with 50 LD 50 of Y. pestis KIM6+(pCD1Ap) at 42 days after initial immunization. Both KIM10(pCD1Ap)-immunized and YPS19(pCD1Ap)-immunized groups had the same protective efficacy (survival of 70%). None of the mice administrated PBS survived the pulmonary challenge ( Figure 5B). Analysis of serum IgG responses to recombinant LcrV antigen and YpL showed that IgG titers to LcrV or YPL were similar in KIM10(pCD1Ap)-and YPS19(pCD1Ap)-immunized groups at week 2 and displayed increasing trends at week 4 ( Figure 5C,D).
intranasally with 50 LD50 of Y. pestis KIM6+(pCD1Ap) at 42 days after initial immunization. Both KIM10(pCD1Ap)-immunized and YPS19(pCD1Ap)-immunized groups had the same protective efficacy (survival of 70%). None of the mice administrated PBS survived the pulmonary challenge ( Figure 5B). Analysis of serum IgG responses to recombinant LcrV antigen and YpL showed that IgG titers to LcrV or YPL were similar in KIM10(pCD1Ap)-and YPS19(pCD1Ap)-immunized groups at week 2 and displayed increasing trends at week 4 ( Figure 5C,D).

Discussion
The success in development of a new generation of effective plague vaccines generally depends on the existence of a suitable vaccine candidate with desirable characteristics capable of eliciting a marked immunity with minimal side effects. This can be accomplished by using technologies to significantly improve both the protection and safety characteristics of the vaccine candidate by modifying particular bacterial properties.
In addition, the outbred Swiss Webster mouse strain has been used as a small-animal model for testing the immunogenicity and efficacy of the rF1V vaccine against plague [40][41][42]. The heterogeneous immune responses expected in Swiss Webster mice due to inherent genetic differences could be a more accurate reflection of the expected human response, rather than in an inbred strain of mouse [40]. Therefore, evaluation of live attenuated Y. pestis vaccine candidates in the outbred mice would be more relevant to translation of mouse studies to human application.
Intramuscularly administrated YPS20(pCD1Ap) could not generate effective protection against virulent Y. pestis(pCD1Ap) challenge as immunization with KIM10(pCD1Ap) or YPS19(pCD1Ap) did ( Figure 3A,B), although administration of YPS20(pCD1Ap) primed similar levels of anti-Yersinia IgG response to administration of KIM10(pCD1Ap) or YPS19(pCD1Ap) ( Figure 4A). It is worthy noticing that the residual virulence in the live Y. pestis EV vaccine strain is considered to be necessary for the development of adequate immunity against plague [3,46]. Previous studies showed that vaccination with attenuated Pgm − derived Y. pestis strains primed T cells in an antibody-independent system that protect mice against pneumonic plague [47,48]. Thus, another reason for this explanation is that antibody responses induced by immunization with the above Y. pestis mutants are not essential for protection. We also noticed that i.m. immunization with attenuated Y. pestis mutants generated balanced Th1/Th2 responses, while s.c. immunization with corresponding strains stimulated Th2-biased responses ( Figure 4B,C), which can explain why mice who survived s.c. administration were conferred less protection against virulent Y. pestis KIM6+(pCD1Ap) challenge than i.m. administration, except for the YPS20(pCD1Ap) strain ( Figure 3).
Mice vaccinated i.m with KIM10(pCD1Ap) by a prime-boost strategy still had two deaths without significance compared to YPS19(pCD1Ap) ( Figure 5A). Immunization with each strain in Swiss Webster mice conferred significant protection (70%) against pulmonary Y. pestis challenge without further enhancement ( Figure 5B). The variations of protective efficacy compared to Figure 3B may be due to genetic variations in those outbred mice. Further, safety evaluation suggested that YPS19(pCD1Ap) was safer than KIM10(pCD1Ap) in Hfe −/− mice. Overall, the advantageous features of the YPS19(pCD1Ap) (∆lpxP::P lpxL lpxL Pgm − pPCP1 − ) makes it a potential candidate for an improved live plague vaccine because it is high attenuation without weakening its protection. This study is paving a way to develop a safe, live Y. pestis vaccine for counteracting human plague.

Conclusions
Virulence evaluation in mice demonstrated that YPS19(pCD1Ap) mainly synthesizing the hexa-acylated lipid A was further attenuated in comparison to its parent strain, KIM10(pCD1Ap) by i.m. or s.c. administration. The i.m. immunization with either KIM10(pCD1Ap) or YPS19(pCD1Ap) strain induced similar levels of antibody response manifesting a balanced Th1/Th2 response and afforded the same protective efficacy against Y. pestis infection. Therefore, YPS19(pCD1Ap) would be a potential vaccine candidate with improved features.