Content uploaded by Ram Prasad Aganja
Author content
All content in this area was uploaded by Ram Prasad Aganja on Jan 04, 2025
Content may be subject to copyright.
Aganjaetal. Veterinary Research (2025) 56:2
https://doi.org/10.1186/s13567-024-01413-8
RESEARCH ARTICLE Open Access
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco
mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Veterinary Research
Deletion ofpagL andarnT genes involved
inLPS structure andcharge modulation
intheSalmonella genome confer reduced
endotoxicity andretained ecient protection
againstwild-type Salmonella Gallinarum
challenge inchicken
Ram Prasad Aganja1,2†, Jun Kwon1†, Amal Senevirathne1 and John Hwa Lee1,2*
Abstract
Fowl typhoid (FT) poses a significant threat to the poultry industry and can cause substantial economic losses,
especially in developing regions. Caused by Salmonella Gallinarum (SG), vaccination can prevent FT. However, existing
vaccines, like the SG9R strain, have limitations, including residual virulence and potential reversion of pathogenicity.
This study aims to develop safer and more effective SG vaccine strains through targeted genetic modifications, focus-
ing on genes involved in lipopolysaccharide (LPS) biosynthesis and modification. We evaluated two novel mutant
SG strains, JOL3015 and JOL3016, carrying in-frame deletions in ΔlonΔrfaLΔarnT and ΔlonΔrfaLΔpagL, respectively.
Intramuscular immunisation of 4-week-old young birds with JOL3015 and JOL3016 strains showed minimal impact
on their growth. However, the immunisation significantly increased antigen-specific IgY, sIgA secretion, and CD4+
and CD8+ T-cell responses while inducing lower pro-inflammatory cytokine levels than SG9R. Histopathological
evaluations revealed substantial protection in the immunised birds, with minimal tissue damage and inflammatory
responses, thus reducing the in vivo bacterial burden. Furthermore, none of the immunised birds died. This outcome
highlights the significant safety and protection the selected genetic modifications conferred. Our results indicate
that JOL3016 provided comparable protective outcomes on par with SG9R, yet with significantly lower endotoxicity
responses during the lethal challenge with SG WT JOL422. The novel detoxified SG strains, particularly JOL3016, offer
a promising alternative to existing vaccines for FT. They provide effective protection with minimal impact on poultry
growth, thereby minimising the risks associated with reversion and endotoxicity. The study highlights the potential
of genetically engineered vaccine strains in improving poultry health and productivity, emphasising the importance
of continued research.
Keywords Fowl typhoid, Salmonella Gallinarum, genetic modification, lipopolysaccharide biosynthesis, vaccine
Handling editor: Kate Sutton.
†Ram Prasad Aganja and Jun Kwon contributed equally to this work.
*Correspondence:
John Hwa Lee
johnhlee@jbnu.ac.kr
Full list of author information is available at the end of the article
Page 2 of 20
Aganjaetal. Veterinary Research (2025) 56:2
Introduction
Fowl typhoid (FT), caused by Salmonella enterica sero-
var Gallinarum (Salmonella Gallinarum, SG), is a severe
systemic disease affecting chickens of all age groups.
According to some studies, SG has a global prevalence
of 8.54%, with Asia ranking at the top [1]. Various fac-
tors, including host age, host susceptibility, nutrition,
flock management, and bacterial virulence, influence
the severity of the disease. e disease imposes a signif-
icant threat to the poultry industry, causing up to 100%
mortality and substantial economic losses [2]. Current
control measures include strict biosecurity regulations,
antibiotic use, and vaccination. However, maintain-
ing biosecurity is costly and challenging for poultry
operations, while long-term antibiotic use can lead to
the development of multi-drug resistant strains. Con-
sequently, vaccination is one of the most effective con-
trol strategies, with options including live, inactivated,
and subunit vaccines. Although FT has been eradicated
from commercial poultry in developed countries, it
remains widespread in most developing countries. is
ongoing issue underscores the need for effective and
accessible vaccination strategies to mitigate the impact
of FT on global poultry production [3].
e live attenuated SG9R strain, a semi-rough strain
with limited information on its attenuation, serves as
a commercial vaccine for FT. However, SG9R has been
reported to cause systemic disease, liver and spleen
pathology, and bacterial persistence for several weeks
in young chickens, which could inevitably affect the
productivity of young birds [4]. Furthermore, SG9R
vaccination has been associated with residual virulence
in newly hatched chickens, limited protection, and ver-
tical transmission [5]. e disease remains prevalent
among poultry flocks despite Korea’s early adoption of
a control and eradication policy for FT in the 1970s [6].
is study proposes that a safer SG vaccine can be
developed through bacterial strain manipulation using
genetic engineering to address this gap. It is well-estab-
lished that bacterial lipopolysaccharides (LPS) initiate
pro-inflammatory immune responses and endotox-
icity, which can be lethal to the host, especially at a
young age [7]. Lipid A and its acyl chains in LPS play
a central role in triggering inflammatory cytokines.
Hexa-acylated lipid A stimulates a maximum pro-
inflammatory response via the TLR4-MD2-CD14 path-
way, while tetra- or penta-acylated species significantly
reduce immunostimulatory responses [8]. us, lipid
A-derived endotoxicity can be mitigated by its struc-
tural remodelling. In the present context, the PhoP/
PhoQ-activated gene (pagL) encodes deacylase. is
encoding modifies lipid A by removing R-3-hydroxy
myristate attached at position 3, thus maintaining
bacterial virulence. Hence, pagL deletion can confer
detoxification of lipid A, reducing endotoxicity [9].
Similarly, arnT (-Ara4N transferase) modifies LPS
by adding 4-amino-4-deoxy--arabinose (-Ara4N)
to lipid A’s phosphate groups, altering the charge and
structure of the LPS. is modification contributes to
bacterial survival and immune evasion [10]. Hence,
arnT represents a potential target for regulating the
virulence of SG strains. Furthermore, by precisely and
permanently deleting such genes, there is no risk of SG
wild-type (WT) strains reverting to a virulent form,
making this strategy safe and effective for developing
vaccine candidates. erefore, remodelling the LPS
structure holds promising potential for generating avir-
ulent SG strains for vaccine development.
Furthermore, the serological diagnosis of Salmonella
infection relies on detecting LPS-specific antibodies
against the O-antigen. However, this method can often
be hindered by field infections, making it challenging to
differentiate between infected and vaccinated animals
(DIVA). e DIVA concept is crucial for implementing
effective vaccination strategies. Monitoring salmonello-
sis and ensuring ideal vaccination necessitates the capa-
bility to differentiate infected from vaccinated animals,
a feat that can be achieved through LPS truncation via
O-antigen modification. us, targeting the deletion
of rfaL, which encodes O-antigen ligase, aims to lower
LPS-specific antibodies compared to wild-type infec-
tion [11]. Ultimately, this strategy aids in differentiating
animals that are infected from those that have been vac-
cinated by quantifying antibody levels using enzyme-
linked immunosorbent assay (ELISA). e Lon protease
serves as a global regulator that controls the expression
of virulence genes located in Salmonella pathogenicity
island I (SPI-1) during the early stages of systemic infec-
tion. e dysregulation of Lon protease, a negative regu-
lator of SPI-1 genes, causes an increase in the expression
and coordination of early virulence genes [12]. However,
attenuating SG through lon gene deletion renders the
strain hyper-immunogenic with reduced virulence [13].
is targeted genetic modification not only enhances the
immunogenicity of the strain but also contributes to its
safety profile, making it a promising candidate for vac-
cine development against Salmonella infection.
e study’s objective was to comprehensively evalu-
ate the safety and protective efficacy of attenuated SG
strains engineered through targeted deletion of the lon
gene to reduce virulence. Additionally, rfaL gene deletion
was undertaken to enhance the capability of monitor-
ing salmonellosis using the DIVA principle. e strains
underwent detoxification processes to yield SG strains
with lonrfaLpagL and lonrfaLarnT modifica-
tions. e study conducted comparative assessments
Page 3 of 20
Aganjaetal. Veterinary Research (2025) 56:2
to examine the protective potential of these engineered
strains against wild-type challenge. e results showed a
significant improvement in safety and efficacy compared
to SG9R, a commercial vaccine strain. ese findings
underscore the potential of genetically engineered SG
strains as viable candidates for advanced vaccine devel-
opment, offering enhanced safety, efficacy, and moni-
toring capabilities in combating Salmonella infection in
poultry populations.
Materials andmethods
Bacterial strains, plasmids, andgrowth conditions
All bacterial strains were routinely grown in Luria
Bertani (LB) medium (Becton Dickinson, Sparks, MD,
USA) with agitation at 37°C using appropriate antibi-
otics where applicable. e strains and plasmids used
are listed in Table 1. e SG 914 strain (Δlon) [14]
was engineered to develop the SG JOL3015 and SG
JOL3016 strain by deleting ΔrfaL ΔarnT and ΔrfaL
pagL, respectively, using the λ red recombination
technique [15]. Briefly, the parent SG strain was trans-
formed with a helper plasmid, pKD46, and induced
to express recombinase with L-arabinose for homolo-
gous recombination. e linear DNA cassette of the
catR gene flanked by a rfaL gene homologous sequence
was amplified from pKD3 and electroporated (Harvard
Apparatus, USA) in pKD46-transformed Salmonella.
e rfaL-deleted mutant colonies were screened by
plating on LB media containing chloramphenicol. Colo-
nies were confirmed by inner primers and transformed
with pCP20 plasmid to eliminate the FRT-flanked catR
through flippase production. e deletion of catR was
confirmed using flanking primers (listed in Table2).
e same procedure was repeated to include the dele-
tion of arnT and pagL in their respective strains. A
commercially available vaccine SG9R was procured
(9R VAC, Komipharm International Co. Ltd., Siheung,
Korea) for the comparative study.
Bacterial growth kinetics
oftheengineeredSalmonellaGallinarum strains
Growth kinetics were evaluated using SG mutant strains
and the commercial SG9R strain. Overnight-grown bac-
terial cultures were used as 1% v/v inoculum into 50 mL
of LB broth. e cultures were incubated at 37°C in a
shaking incubator at 200rpm. Optical density at 600nm
(OD600) was measured every 4h in a 96-well plate (200
µL) using a spectrophotometer (Tecan, Seestrasse, Swit-
zerland). Samples were withdrawn every 4h and used in
colony counting after preparing serial dilutions. Plates
containing 30–300 colonies were counted to determine
CFUs.
Auto‑aggregation assay
e clustering ability of bacteria was assessed using an
auto-aggregation assay. Overnight bacterial cultures were
prepared and inoculated at a 1:100 dilution in LB broth.
After incubating the cultures at 37°C for 24h, the upper
layer was collected without disturbing the cultures. is
layer was then used to measure the optical density at
OD600 nm. Subsequently, the culture was resuspended by
vortexing and used to take the second absorbance meas-
urements under the same absorbance conditions. e
level of auto-aggregation was determined as a percentage
using the formula: [(OD600 post-resuspension – OD600
pre-resuspension)/ OD600 post-resuspension] × 100.
Hemolysis assay
e wild-type and mutant strains were grown as over-
night cultures. Subsequently, cells were harvested
through centrifugation at a rate of 8000rpm for 10min.
Supernatants were filtered through 0.2 m membrane
filters (BioFACT, Parit Jamil, Malaysia). e sterile solu-
tions were mixed with 10% chicken red blood cell (RBC)
suspension at a 4:1 ratio and incubated in a shaking incu-
bator at 37°C for 12h. A control was prepared by add-
ing LB broth to the RBC suspension at the same ratio.
After incubation, the suspensions were centrifuged at
Table 1 List of bacterial strains and plasmids
Bacteria/plasmid Genotypic characteristics References
S. Gallinarum
JOL422 Wild type Lab stock
JOL914 Δlon Lab strain
JOL3015 Δlon ΔrfaL ΔarnT This study
JOL3016 Δlon ΔrfaL ΔpagL This study
pKD46 oriR101-repA101ts; encodes λ red genes (exo, bet, gam); native terminator (tL3); arabinose-inducible promoter
for expression (ParaB); bla
[21]
pKD3 oriR6Kgamma, bla (ampR), rgnB (Ter), catR, FRT [21]
pCP20 Helper plasmid contains a temperature-inducible flp gene for removing the FRT flanked chloramphenicol gene [36]
Page 4 of 20
Aganjaetal. Veterinary Research (2025) 56:2
2000 rpm for 5 min [16]. Hemolysis rates were deter-
mined by measuring absorbance at 570nm using a multi-
well plate reader (Tecan, Männedorf, Switzerland).
Acriavine agglutination test
e acriflavine agglutination test determined the lack of
O-antigen components and confirmed the rough phe-
notype [17]. Bacteria cultures were grown on LB agar
plates for 24h, with selected bacterial colonies collected
and mixed into 30 µL of 0.2% acriflavine solution (Sigma,
Missouri, USA) on glass slides. Cells were gently mixed,
interacted for 2min, and observed under a microscope at
40 × magnification by the naked eye.
Western blot oflipopolysaccharides
Following the manufacturer’s guidelines, bacterial
lipopolysaccharides were extracted using a phenol-based
method via an LPS extraction kit (iNtRON Biotechnol-
ogy, Seoul, South Korea). e LPS samples were sepa-
rated on 12% sodium dodecyl sulfate-polyacrylamide
gel electrophoresis (SDS-PAGE). Western blotting was
performed using a mouse monoclonal antibody against
Salmonella O-antigen at 1:1000 (cat. no. 10R-S103b,
Fitzgerald, MA, USA) and goat anti-mouse IgG-HRP
conjugate at 1:5000 dilution (cat. no. 1030-05, Southern-
Biotech, Birmingham, AL 35209 USA). All steps were
conducted according to a previously described procedure
[18].
Adhesion, invasion, andstress survival
Hela and chicken peripheral blood mononuclear cells
(PBMCs) were used invitro to evaluate the adhesion and
invasion capability of SG strains. Overnight cultures of
bacterial cells were re-inoculated to LB medium as 1%
inoculum and incubated for 3h to reach 0.4–0.6 absorb-
ance at OD600. Cells were collected by centrifugation at
12 000 × g for 5 min and washed with phosphate-buff-
ered saline (PBS). Following standard procedure, blood
was collected from the bird’s wing vein, and PBMCs were
isolated using Ficoll-Paque PLUS density gradient media
(Cytiva, Uppsala, Sweden) [19]. PBMCs (2 × 106) and Hela
cells (2 × 105) were cultured in a 12-well plate with RPMI
medium, 10% heat-inactivated foetal bovine serum (FBS),
and 1% penicillin-streptomycin. e cells were then incu-
bated at 37°C in a humidified 5% CO2 incubator.
Table 2 List of primers
Gene Primer 5′–3′sequences References
Gel deletion
lon-pKD3 Sense GGT ATG GAG CAC AGC TAT ACT ATC TGA TTA CCT GGC GGA CAC TAA ACT AAG TGT AGG CTG GAGCT This study
Antisense CGA AAT AGC CTG CCA GCC CTG TTT TTA TTA GCG CTA TTT GCG CGA GGT CAA TGG GAA TTA GCC ATG
rfaL-pKD3 Sense TTT GGA AAG ATT CAT TAA AGA GAC TCT GTC TCA TCC CAA ACC TAT TGT GGG TGT AGG CTG GAG CTG CTTC This study
Antisense CCT GAT GAT GGA AAA CGC GCT GAT ACC GTA ATA AGT ATC AGC GCG TTT TTA TGG GAA TTA GCC ATG GTCC
pagL-pKD3 Sense AAT TTT AAA TAT GTT AGC CGG TTA AAA ATA ACT ATT GAC ATT GAA ATG GTG TGT AGG CTG GAG CTG CTTC This study
Antisense CGG TGA TTA ATT ACT CCT TCA GCC AGC AAC TCG CTA ATT GTT ATT CAA CTA TGG GAA TTA GCC ATG GTCC
arnT-pKD3 Sense GAG CTG ACC GCC AAC GCT GAG CAG ACT GGC AAG CAC CAG AAT GAC GCC GAG TGT AGG CTG GAG CTG CTTC This study
Antisense ATC CCT GGC CGT GAA GGT TGG CTG GGG TGC CAA CAG GCA GCG AGC GCC TCA TGG GAA TTA GCC ATG GTCC
lon-inner Sense AAT CTG CAC GAC TAC CTC GG This study
Antisense GAT TAC CGG TCA GGC AGG AA
lon-outer Sense CAG GAG TTC TTA CAG GTA GA This study
Antisense CCA CAC TCC GCT GTA GGT GA
rfaL-inner Sense ACA AGT TTA GGA CTT CGC TGCC [15]
Antisense CAG AAT GGT ATT ATG CGG ACCG
rfaL-outer Sense GCA GCG TTT CGA GGA ACA AA [15]
Antisense TCG TAT CGG TTG ATA CCG GC
pagL-inner Sense CAG ATC TCT TTT GCT GCG GG [15]
Antisense AAA AGC CCC AAA GTT CCA GC
pagL-outer Sense TGG ATG TGC CTG AAC AAC ACT [15]
Antisense TTA GCC TCC CTG TCG CCA TA
arnT-inner Sense GCA ACG CGG TAC GTT TAT CC This study
Antisense GAA ACG CGC TAT GCC GAA AT
arnT-outer Sense GAG CTG ACC GCC AAC GCT GA This study
Antisense GAA ACG CGC TAT GCC GAA AT
Page 5 of 20
Aganjaetal. Veterinary Research (2025) 56:2
e PBMCs were incubated for 5h to facilitate attach-
ment, after which the media was changed to retain adher-
ent PBMCs. e adherent PBMCs and Hela cells were
then infected with SG WT, SG9R, and the attenuated
strains at 40 multiplicity of infection (MOI), 30min for
adhesion and 2.5h for invasion. e non-infected bacte-
ria were eliminated by 2h gentamycin treatment (100µg/
mL). Adhered or invaded cells were retrieved by lysis of
monolayers using 0.25% Triton X-100, and the bacterial
enumeration was done by counting on Brilliant Green
Agar (BGA) plates (BD, Difco). To conduct the stress sur-
vival assay, bacterial strains grown at the log phase were
subjected to acidic stress at pH 4.0 and oxidative stress
at 1 mM and 5 mM H2O2 for 30min. e survival rate
was evaluated based on the initial inoculum by plating on
BGA at 10-fold dilution.
Cell survival andcytotoxicity assay
Salmonella-induced cell cytotoxicity was assessed using
an IncuCyte live imaging system (Essen Bioscience,
MI, USA). HeLa cells were seeded at 5 × 10⁵ cells/mL in
12-well plates. Cells were infected with Salmonella at
a MOI of 40 for 2h and washed twice with PBS. Non-
infected bacterial cells were eliminated with gentamy-
cin (100µg/mL) treatment for 2h. e cells were then
treated with propidium iodide (5 µL/mL, cat. no. 556463,
BD Biosciences, California, USA) and monitored via
imaging at 6h intervals over 24h.
Safety evaluation ofdetoxied SG strains inchicken
e safety of detoxified SG strains was assessed in
female brown-layer chickens following intramuscular
(IM) injection. Four-week-old chickens (n = 12) were
inoculated with either the SG wild-type (WT JOL422),
attenuated strains JOL3015 and JOL3016, or a com-
mercial vaccine strain, SG9R. Birds were monitored for
morbidity and mortality associated with FT. Clinical
parameters such as body temperature, abnormal behav-
iour, anorexia, and feed intake were observed to detect
any adverse effects caused by SG infection. Once birds
displayed severe lethargy, immobility, and greenish diar-
rhoea, they were culled and categorised under the mor-
tality group for further analysis. All animals were handled
following guidelines set by the Animal Ethics Commit-
tee (NON2023-135-001) in compliance with the Korean
Council on Animal Care and the Korean Animal Protec-
tion Law, 2007: Article 13.
e attenuated strains, JOL3015 and JOL3016, were
administered at concentrations of 1 × 10⁷ CF U/bird
(Low) and 1 × 10⁸ CFU/bird (High) to evaluate bacterial
persistence in vital organs. SG9R and SG WT JOL422
were administered at a concentration of 1 × 10⁷. ree
chickens from each group were sacrificed at 3, 7, and 14
days post-inoculation (dpi) for sample collection. After
the chickens were euthanised, the spleen and liver were
aseptically collected, which were then homogenised in
PBS using a mechanical homogeniser (IKA T 10 basic
ULTRA-TURRAX, Germany) and plated on BGA at
10-fold serial dilutions to quantify the bacterial load.
Additionally, cloacal swabs were collected using a ster-
ilised cotton swab in 1 mL of PBS to evaluate bacterial
shedding for environmental safety. e swab samples
were thoroughly mixed, serially diluted in PBS, and then
plated on BGA. Changes in body weight were monitored
every three days for up to 15 days after inoculation to
determine how the infection affected weight gain.
Histopathological evaluation oforgan damage
Histopathological examination was undertaken on the
liver, spleen, and cecum tissue sections using haema-
toxylin and eosin (H&E) staining. ree birds per group
were sacrificed on the seventh day post-inoculation, with
the organs collected and fixed in 10% formalin. e tis-
sues were sectioned into 3m slices, fixed, and processed
according to a standard protocol for H&E staining. is
process involved dehydration, clearing, embedding, and
staining to allow for clear visualisation of tissue architec-
ture. A comprehensive investigation of potential tissue
damage was performed using a Zeiss Axio Imager.M2
microscope (Carl Zeiss AB, Stockholm, Sweden). Micro-
scopic examination was conducted to evaluate the cellu-
lar and structural integrity, and the resulting images were
documented for further analysis.
Immunisation andchallenge againstfowl typhoid using
attenuated SG strain
e immune response elicited by inoculating 4-week-old
female brown chickens with attenuated SG strains was
evaluated. Each bird (n = 8) was intramuscularly immu-
nised with JOL3015 and JOL3016 strains at a concentra-
tion of 1 × 10⁷ CFU/200 µL. A commercial vaccine strain,
SG9R, was administered intramuscularly as a compara-
tive control. Additional groups served as PBS and naïve
controls. After two weeks, the birds received a booster
immunisation with the attenuated SG strains. Serum and
cloacal swab samples were collected at intervals for five
weeks following the initial inoculation. e cloacal swabs
were collected by inserting a sterile cotton swab into the
cloaca and immediately breaking the swab section into 1
mL of PBS-containing tubes, which were stored at 4°C
during the collection. e samples were then stored at
−80°C until further analysis and later used to measure
levels of IgY and IgA antibodies.
Furthermore, blood samples were collected two weeks
after the booster immunisation, and PBMCs were iso-
lated. T-cell counts were assessed using flow cytometry
Page 6 of 20
Aganjaetal. Veterinary Research (2025) 56:2
to quantify the cell-mediated immune response. ree
weeks after booster application, the chickens were chal-
lenged with SG WT JOL422 via the oral route using
1 × 106 CFU/200 µL per bird. e post-challenge sur-
vival rate was evaluated by monitoring for up to 15 days.
Birds displaying severe lethargy, immobility, and greenish
diarrhoea were culled and categorised under the mor-
tality group for further analysis. Bacterial persistence in
the spleen and liver of immunised chickens was investi-
gated one week after the challenge to elucidate the bac-
terial load. In addition, the spleen and liver tissues were
collected for H&E staining as described elsewhere [20].
At the end of the experiment, animals were sacrificed
to examine any gross morphological distortions in their
vital organs.
ELISA forcytokines quantication andhumoral
andmucosal immune responses
Chickens inoculated with engineered SG strains under-
went endotoxicity assessment by quantifying inflam-
matory cytokines, with serum samples collected on
day three post-inoculation. e levels of inflammatory
cytokines, including TNF-α, IL-1β, and IFN-γ, were
measured using commercial sandwich-ELISA kits follow-
ing the manufacturer’s instructions. Briefly, micro-ELISA
plates pre-coated with an antibody specific to chicken
TNF-α (Cat. No. MBS2509660, MyBioSource, San Diego,
USA), IL-1β (Cat. No. MBS2702032, MyBioSource) and
IFN-γ (Cat. No. MBS2700893, MyBioSource) were incu-
bated separately with serum samples along with corre-
sponding standards. A biotinylated detection antibody
and an avidin-horseradish peroxidase (HRP) conjugate
were successively added to the microplate wells and incu-
bated. After washing, a substrate solution was added to
initiate the enzyme reaction, which was then halted with
a stop solution. e Infinite M200 spectrophotometer
(Tecan) was used to measure the optical density (OD) at
450nm. Cytokine concentrations were estimated using a
reference standard.
Salmonella-specific systemic IgY and mucosal IgA
responses in immunised birds were quantified using
an indirect ELISA. Plates were coated with 400 ng/
well of crude SG WT strain protein in carbonate-bicar-
bonate buffer and incubated overnight at 4°C. Block-
ing was done with 5% skim milk. Serum samples were
diluted 1:50 for IgY detection, while undiluted cloacal
swab samples were used for IgA detection. Samples
were added to the wells and incubated for 2h at room
temperature (RT), followed by incubation with goat
anti-chicken IgY-HRP (Bethyl Laboratories, Texas,
USA) or goat anti-chicken IgA-HRP (Bethyl Labora-
tories, Texas, USA) according to the manufacturer’s
instructions. After washing, the O-phenylenediamine
dihydrochloride substrate (Sigma, Missouri, USA) was
added for colourimetric detection. e enzyme-sub-
strate reaction was stopped with 50 µL of 2N sulfuric
acid, and OD at 492nm was measured using an Infinite
M200 microplate reader (Tecan). e obtained absorb-
ance values were used to quantify the levels of IgY and
IgA antibodies in the serum and cloacal swab samples,
respectively.
Flow cytometry
e cell-mediated immune responses were investi-
gated by evaluating T-lymphocyte subsets using flow
cytometry analysis. Two weeks after administering
the booster immunisation, blood was collected from
all groups (n = 5) to isolate PBMCs. Briefly, blood was
diluted 1:1 with PBS (pH 7.4) and carefully layered over
an equal volume of Ficoll-Paque PLUS density gradient
media. Samples were centrifuged at 400×g for 30min
at 18°C, and PBMCs separated at interface layers were
collected. Cells were then resuspended in RPMI-1640
medium supplemented with 10% FBS and 1% antibiot-
ics and seeded in 96-well plates at 1 × 105 cells/well. e
cells were stimulated with 400 ng/well of crude soluble
antigen extracted from the SG WT strain for 72h in a
5% CO2 incubator at 37°C.
Cells were collected and incubated with fluorescently
labelled antibodies: anti-CD3-FITC (Clone CT-3, Cat:
8200-02, SouthernBiotech, Birmingham, AL, USA), anti-
CD8a-PE (Clone CT-8, Cat: 8220-09, SouthernBiotech),
and anti-CD4-AF700 (Clone CT-4, Cat: 8210-27, South-
ernBiotech) (each at a concentration of 8µg/mL) at 4°C
for 30min in the dark. After incubation, the cells were
washed with FACS buffer (PBS containing 2% FBS and
0.1% sodium azide) to remove unbound antibodies. e
stained cells were then analysed with 1 × 104 cells/sample
acquisition using a MACSQuant flow cytometer (Milte-
nyi Biotec, Bergisch Gladbach, Germany). As a gating
strategy, first, total lymphocytes were selected, and CD3+
cells were gated, from which CD3+CD4+ and CD3+CD4+
cells were quantified using non-stained cells and fluores-
cence minus one control. e results were analysed using
MACSQuant analysis software (version 2.6), allowing
for a detailed assessment of the cell-mediated immune
response elicited by the immunisation.
Statistical analysis
Statistical analysis was performed using Student’s
t-test and ANOVA to evaluate statistical differences. A
p-value < 0.05 was considered significant. All analyses
were done in GraphPad Prism 9.00 software (San Diego,
CA, USA).
Page 7 of 20
Aganjaetal. Veterinary Research (2025) 56:2
Results
Development ofattenuated SG strain
e SG strains were engineered to possess defec-
tive LPS structures using the well-established lambda
red recombination method [21]. is recombineering
approach involved replacing the selected genes with
a flippase recognition target (FRT) flanked chloram-
phenicol resistance (catR) gene in the chromosome.
e targeted deletions included four specific genes: lon,
rfaL, pagL, and arnT. Confirmation of these deletions
was achieved through flanking PCR [15], as depicted
in Additional file1 (PCR results), using specific flank-
ing primers listed in Table 2. Notably, deleting rfaL
impacted the biosynthesis of the core oligosaccharide,
resulting in modified LPS lacking O-antigen attach-
ment. e arnT deletion supposedly alters the transfer
of L-Ara4N to the phosphate group, affecting the over-
all charge of the cell surface. Deleting pagL may block
lipid A’s deacylation, preventing further modifications
in the LPS structure. e conceptual framework of
these deletions is depicted in Figure1.
Phenotypic andbiological characterisation
Our study found that bacteria demonstrate self-aggre-
gation when cultured and that the hydrophobicity of
their cell surfaces may influence this behaviour. Nota-
bly, mutant strains JOL3015 and JOL3016 exhibited sig-
nificantly higher auto-aggregation abilities, with rates
of 61% and 59%, respectively, compared to WT and
SG9R strains, which had auto-aggregation rates of only
25% and 37%, respectively (Figure2A). Moreover, the
haemolytic assay revealed a remarkable reduction in
hemolysis exceeding 50% in both mutant strains com-
pared to the control (Figure2B), indicating a significant
alteration in their haemolytic properties. Additionally,
the acriflavine agglutination test demonstrated agglu-
tination in the presence of acriflavine for both mutant
strains, suggesting a rough surface phenotype (Fig-
ure2C). is attribute ensures that the lipid A core is
exposed, allowing acriflavine to interact, thus leading to
agglutination. When visualised under ultraviolet light,
clear agglutination patterns were evident to the naked
eye. Further analysis by western blotting confirmed the
absence of interaction between the mutant strains and
antibodies against Salmonella O-antigen, highlight-
ing a phenotypic change induced by the LPS mutation
in these strains (Figure2D). ese findings collectively
underscore the influence of hydrophobicity and LPS
modifications on the cell surface properties of these
bacterial strains, providing valuable insights into their
phenotypic characteristics.
Bacterial growth kinetics ofattenuated SG strains
e growth kinetics of attenuated SG strains were
evaluated and compared with the wild-type strain,
JOL422, and a commercial strain, SG9R (Figure 3A,
B). roughout the experiment, discernible differ-
ences in growth dynamics were observed between the
engineered SG strains and the wild-type counterpart.
While the wild-type JOL422 and SG9R strains exhib-
ited analogous growth patterns, significant disparities
were noted with the engineered strains, particularly
JOL3016 and JOL3015. During the initial growth phase,
both JOL3015 and JOL3016 maintained a conspicuous
gap compared to the wild-type strain, with JOL3016
displaying a slightly narrower gap than JOL3015. is
disparity persisted up to 16h of incubation, after which
the gap gradually decreased and plateaued. Notably, the
optical density at 600nm (OD600) peaked between 16
and 20h for the wild-type and SG9R strains, followed
by a decline.
In contrast, JOL3015 and JOL3016 exhibited incre-
mental bacterial growth up to 28h. At 8h, the wild-type
strain demonstrated a 3.14 and 2.39-fold increase in
OD600 compared to JOL3015 and JOL3016, respectively,
narrowing to 1.35 and 1.25-fold at 16 h. e logarith-
mic phase was observed in all four strains between 4 and
12h, with both absorbance and CFU increments being
increased, thus narrowing the gap against the WT SG
422 at 28h post-incubation. At the end of the incuba-
tion period, there was a minimal disparity in CFU growth
between JOL3016 and SG9R, suggesting comparable
growth kinetics.
In vitro characterisation ofbacterial virulence
andenvironmental stress
e assessment of adhesion and invasion using Hela
and chicken PBMC revealed JOL3016 with compara-
ble results against the SG WT 422 strain. e adhesion
(Figure3C) and invasion (Figure3D) capabilities of SG9R
and JOL3015 were significantly lower than those of the
GS WT 422 strain and JOL3016. e exposure of bac-
terial cells to acidic environments at 6.5 pH and 4.0 pH
revealed that all strains could tolerate mild acid condi-
tions at 6.5 pH. However, the increase in acidity at 4.0
pH showed that the mutants are susceptible to acidity.
JOL3015 presented the lowest tolerance, while JOL3016
was comparable to the SG9R vaccine strain (Figure3E).
Furthermore, oxidative stress conditions induced by vari-
able concentrations of H2O2 (mM) demonstrated a sig-
nificant growth suppression even at 1.0 mM. At 5.0 mM
concentration, bacterial cell growth was still present.
However, 10.0 mM concentration was lethal to all bacte-
rial strains (Figure3F).
Page 8 of 20
Aganjaetal. Veterinary Research (2025) 56:2
In vitro assessment ofcytotoxic responses
Intracellular cytotoxicity induced by each strain SG
WT422, SG9R, JOL3015, and JOL3016 was investigated
using the propidium iodide staining method. Cells were
observed in real-time using the IncuCyte (Essen Biosci-
ence, Gottingen, Germany) live imaging system (Fig-
ure4A). Visual observation over 24h showed the highest
number of red fluorescing objects in cells treated with SG
WT 422 strain. Furthermore, the matric quantification
of mean red-fluorescent objects revealed that both SG9R
and JOL3016 were comparable, while JOL3015 remained
lowest in cytotoxic responses (Figure4B).
Safety assessment ofthedetoxied strains
e bacterial load in vital organs, including the spleen
and liver, as well as in cloacal swabs, was evaluated to
estimate the burden caused by the detoxified SG strains.
Chickens were inoculated with mutant strains at two
Figure1 Structural modications of lipopolysaccharide (LPS) resulting from gene deletions. The schematic representation illustrates
the structural components of LPS in wild-type and genetically modified strains. The non-modified LPS comprises three main components:
Lipid A, Core oligosaccharide, and O-antigen. In the genetically modified strain, the rfaL gene deletion results in the absence of the O-antigen.
Additionally, pagL gene deletion leads to the lack of deacylated Lipid A. Furthermore, the arnT gene deletion prevents the addition
of 4-amino-4-deoxy-l-arabinose (l-Ara4N) to the phosphate groups of Lipid A. These gene deletions result in significant structural modifications
of the LPS, which are critical for understanding the functional and immunogenic implications of bacterial pathogenesis.
Page 9 of 20
Aganjaetal. Veterinary Research (2025) 56:2
doses, 1 × 107 and 1 × 108 CFU/bird, via the intramuscular
(IM) route and monitored over 15 days. Birds inoculated
with the WT strain JOL422, which served as the control,
displayed lethargic behaviour, which was characterised
by depression, anorexia, ruffled feathers, diarrhoea, dehy-
dration, and weight loss. In contrast, chickens from the
other groups exhibited normal behaviour with usual feed
and water intake. ey also did not show adverse signs of
inoculation or clinical symptoms, such as increased body
temperature.
e bacterial load in the spleen, liver, and cloacal swabs
indicated the dispersal of bacteria throughout all tested
organs and sites. Over time, the bacteria were gradu-
ally eliminated from their respective sites, with bacte-
rial persistence lasting for 14 days, which assured the
production of an immune response (Figure5A–C). Bac-
terial retention of the attenuated strains inoculated at
1 × 107 and 1 × 108 CFU/bird in the selected lymphoid
organs was comparable to that of SG9R injected at 1 × 107
CFU/bird. Administration of a tenfold higher bacterial
Figure2 Phenotypic and biological characterisation ofSalmonellaGallinarum strains. A Auto-aggregation. Visual observation
of auto-aggregation in bacterial cultures grown statically at 37 °C for 24 h. The percentage of auto-aggregation was calculated by comparing
the OD600 values from the upper layer of the culture with those from the resuspended culture after vortexing. B Haemolytic Activity. Haemolytic
activity was assessed using the supernatant from mutant bacterial cultures incubated with a 10% chicken red blood cell (RBC) suspension at a 4:1
dilution for 12 h at 37 °C. Haemolytic activity was quantified by measuring the OD570 and comparing the mutant strains to the wild-type. Statistical
analysis was performed using one-way ANOVA, with data presented as ***p < 0.001 and **** p < 0.0001. C Acriflavine Agglutination Test. The rough
surface phenotype of mutant strains was confirmed by agglutination formation with acriflavine. Agglutination was observed under a microscope
at 40× magnification. The scale bar represents 500 μm. D Western Blot Analysis of LPS. Lipopolysaccharide (LPS) was extracted from individual
strains and analysed with a Western blot. The LPS was probed with a mouse antibody against Salmonella O antigen (primary antibody) followed
by a goat anti-mouse IgG-HRP (secondary antibody). M denotes the protein molecular weight marker.
Page 10 of 20
Aganjaetal. Veterinary Research (2025) 56:2
concentration, comparable to SG9R, demonstrated a safe
response. As a positive control, WG WT 422 infection
displayed more than 90% mortality within 5 to 15 days
post-infection (Figure5D). Overall, the results indicate
reduced infectivity in both attenuated strains, while they
retained desirable infectivity to induce immunogenicity.
e introduction of SG as a live vaccine resulted in
a mild decrease in body weight until the third post-
inoculation day. Birds vaccinated with the commercial
SG9R vaccine exhibited more than 7% body weight loss
(Figure6A). In contrast, birds inoculated with JOL3016
lost less than 5% body weight within 15 days compared
Figure3 Characterisation of attenuatedSalmonellaGallinarum strains. A Growth Curve Based on Absorbance. Growth kinetics were
evaluated by measuring the optical density (OD) at 600 nm over time. B Growth Curve Based on CFU. Bacterial growth kinetics were assessed
by plating cultures at respective time points at different dilutions on BGA media. The colony-forming units (CFU) per mL were then evaluated. C In
Vitro Adhesion. The adhesion strengths of JOL3015, JOL3016, and SG9R strains were compared to the Salmonella JOL422 wild-type (WT) strain
using HeLa cells and peripheral blood mononuclear cells (PBMCs). Monolayer cells were infected with each strain at a multiplicity of infection (MOI)
of 40. Adhesion was assessed after 30 min of incubation. D In Vitro Invasion. The invasion capacities of JOL3015, JOL3016, and SG9R strains were
compared to the WT strain using HeLa cells and PBMCs. Monolayer cells were infected with each strain at an MOI of 40. Invasion was assessed
after 2.5 h of incubation. E pH stress survival and (F) Oxidative stress survival. The survival of bacterial strains after stress was evaluated relative
to their initial inoculum concentration. Data were analysed by multiple unpaired t-tests and are presented as *p < 0.05, **p < 0.01, and ***p < 0.001.
Page 11 of 20
Aganjaetal. Veterinary Research (2025) 56:2
Figure4 Assessment of cell cytotoxicity. A Cell Survival Assay. The attenuation level and persistence of mutant SG strains were evaluated
using a cell survival assay. A confluent monolayer of HeLa cells was infected with wild-type (WT ) JOL422, SG9R, JOL3015, and JOL3016
strains at a multiplicity of infection (MOI) of 40. Cell survival was monitored using propidium iodide staining, and cytotoxicity was assessed
by real-time observation with the IncuCyte live imaging system over 24 h. Micrographs show images captured 24 h post-infection, with the scale
bar representing 100 μm. B Cytotoxicity Observation. Higher retention of red-coloured objects was observed in WT-infected cells over the 24 h,
indicating increased cytotoxicity. The dotted lines indicate the lowest mean fluorescence intensity. The experiment was repeated three times,
with R1 and R2 representing the first and second replicates.
Page 12 of 20
Aganjaetal. Veterinary Research (2025) 56:2
to the naïve group. Modifying the LPS structure in both
specially designed SG strains helped to address endo-
toxicity, which is a significant challenge in implement-
ing live bacterial vaccines. e endotoxicity induced by
these strains was corroborated by measuring inflamma-
tory cytokines using sandwich-ELISA. e concentra-
tion of TNF-α, a major inflammatory cytokine marker,
showed a significant reduction. For instance, JOL3015
and JOL3016 exhibited 3.82- and 4.13-fold decreases,
respectively, while SG9R showed a 1.76-fold reduction
compared to the WT (Figure6B).
Notably, JOL3015 and JOL3016 induced 2.17- and
2.34-fold lower TNF-α production than the commercial
SG9R strain. is outcome underscores the significance
of the developed strains. Additionally, the production of
IL-1β was down-regulated by 4.52- and 3.90-fold in the
JOL3015 and JOL3016 groups (Figure6B), respectively,
compared to WT, which was 1.47- and 1.27-fold lower
than SG9R. Furthermore, the endotoxicity-related pro-
inflammatory cytokine IFN-γ showed elevated levels
in the WT group compared to the other groups. Both
developed strains demonstrated a down-regulation of
IFN-γ by more than 2-fold (Figure6B).
Histopathological examinations of H&E stained spleen
and liver tissues also revealed the degree of damage SG
WT 422 strain induced in spleen and liver tissues. Analy-
sis showed expanded white pulp areas of lymphatic tis-
sues in the spleen and signs of severe inflammation and
potentially necrotic regions in liver tissues. Compared to
the WT inoculation group, all vaccinated groups demon-
strated lower levels of tissue damage, particularly for the
SG9R and JOL3016 strains (Figure6C, D). is investi-
gation indicates that our strains can induce lower endo-
toxicity than the WT group. is outcome highlights
Figure5 Safety assessment of the attenuated strains. A–C Bacterial localisation. Birds were inoculated intramuscularly with the developed
strains JOL3015 and JOL3016 at 1 × 107 CFU/bird (low dose, L) and 1 × 108 CFU/bird (high dose, H) to evaluate the safety profile. Bacterial load
was enumerated in the spleen (A), liver (B), and cloacal swabs (C). Data were analysed by multiple unpaired t-tests and are presented as *p < 0.05,
**p < 0.01, and ***p < 0.001. D Kaplan-Meier Survival Curve. The survival of birds was monitored for 15 days post-inoculation to assess the safety
of the strains. The Kaplan-Meier survival curve represents the percentage of surviving birds over the observation period.
Page 13 of 20
Aganjaetal. Veterinary Research (2025) 56:2
the potential of these developed strains in minimising
inflammatory responses. ese findings support the
notion that the engineered strains JOL3015 and JOL3016
are safe and effective in eliciting an immune response
without the adverse effects typically associated with live
bacterial vaccines.
Humoral andmucosal immune response
Assessment of humoral immune responses upon immu-
nisation demonstrated an increase in IgY (Figure7A, B)
levels in blood and sIgA (Figure7C) in mucosal swabs.
e response of IgY almost doubled after receiving the
booster immunisation. Notably, the immune responses
derived by SG9R and JOL3016 were comparable in the
third, fourth, and fifth weeks of post-primary inoculation.
In contrast, JOL3015 derived slightly lower IgY responses
than SG9R and JOL3016. Peak IgY responses resulted
in three weeks of post-priming and were sustained until
the fifth week post-priming. e sIgA responses also
peaked at three weeks post-priming and sustained until
the fourth week of post-priming. A significant increase in
sIgA responses was noted on booster immunisation, and
JOL3016 was comparable to the SG9R vaccine strain.
Cell‑mediated immune responses
e cell-mediated immune response elicited by immu-
nisation was evaluated by quantifying T-cell popula-
tions using flow cytometry analysis. e primary focus
was to differentiate between the T-lymphocyte subsets,
specifically CD4+ and CD8+ T cells, within PBMCs.
Flow cytometric analysis showed a significant increase
in CD3+CD4+ and CD3+CD8+ T-cell populations in
the immunised chickens, indicating an enhanced cell-
mediated immune response (Figure 7D, E) (for gating
strategy, Additional file 2). Chickens immunised with
the SG JOL3016 strain exhibited a notable rise in CD4+
and CD8+ T cells (Figure 7D, E), comparable to the
immune response observed with the commercial vac-
cine strain SG9R. e CD3+CD4+ and CD3+CD8+ T-cell
populations for SG9R were 11.40% and 5.32%, respec-
tively. However, for JOL3016, these populations were
11.25% and 5.61%. ese results indicate that immu-
nisation with the SG JOL3016 and JOL3015 strains sig-
nificantly increases CD4+ and CD8+ T cells, comparable
to the response induced by the commercial SG9R vac-
cine strain. ese findings demonstrate the potential
Figure6 Evaluation of safety and pro‑inammatory cytokines. A Change in Body Weight. The change in body weight of chickens
was monitored following the introduction of Salmonella Gallinarum (SG) strains. B Serum Cytokine Concentration. The concentration
of pro-inflammatory cytokines in the serum was measured. Data were analysed by multiple unpaired t-tests and are presented as *p < 0.05,
**p < 0.01, ***p < 0.001, and ****p < 0.0001. C Histopathological Evaluation of Spleen. Inflammatory response in the spleen of immunised birds
was assessed. Significant tissue alteration, including degeneration and necrosis in the white pulp, was noted in chickens inoculated with the WT SG
JOL422 strain (indicated by arrows). D Histopathological Evaluation of Liver. Inflammatory response in the liver of immunised birds was evaluated.
The black arrow indicates the infiltration of immune cells in the liver of birds infected with the WT strain. The scale bar: 50 μm.
Page 14 of 20
Aganjaetal. Veterinary Research (2025) 56:2
Figure7 Humoral and cellular immune response post‑immunisation. A Immunisation Schedule. Booster immunisation was administered
in the second week following the initial immunisation, and samples were collected at respective points. B IgY Antibody Production. C IgA Antibody
Production. IgY antibody production in serum and IgA in cloacal secretions in response to immunisation was assessed using indirect ELISA
over five weeks post-immunisation. D Flow Cytometry Analysis. Representative flow cytometry scatter plots show the gating of CD4+ and CD8+
cells post-immunisation. E T-Cell Percentages. The histogram represents the percentage of CD4+ and CD8+ T cells in immunised birds. Data were
analysed by multiple unpaired t-tests, with significant differences presented as *p < 0.05, **p < 0.01, and ***p < 0.001 compared to the PBS control.
Page 15 of 20
Aganjaetal. Veterinary Research (2025) 56:2
of the engineered strains to elicit a robust cell-medi-
ated immune response, which is crucial for effective
immunoprotection.
Protection againstwild‑type challenge
Following the designated vaccination schedule, the
chickens were immunised and then exposed to the SG
WT 422 strain through intramuscular (IM) injection.
Body weight measurements and observations for poten-
tial mortality were conducted regularly throughout the
experiment. Immunisation with detoxified SG strains
did not induce adverse reactions during the study period.
e effect of detoxified SG strains on the weight gain of
chickens was especially noticeable during the sixth to
ninth weeks, displaying a higher increase in weight com-
pared to the SG9R vaccine strain. Specifically, chickens
immunised with the JOL3016 strain demonstrated body
weight gains comparable to the naïve group (Figure8A).
Upon challenge, chickens in the PBS group experienced
severe weight reduction and mortality due to SG infec-
tion, whereas all the immunised birds were protected
against the lethal challenge (Figure8B). Furthermore, the
PBS group exhibited increased body temperature (Addi-
tional file3), while the other groups demonstrated only
marginal changes.
Further evaluations revealed that liver morphology
(Figure8C) and splenomegaly (Figure8D) in immunised
groups corroborated the levels of protection provided by
both SG9R and JOL3016 detoxified strains. Post-chal-
lenge assessments showed significant yet comparable
outcomes in spleen weight and bacterial loads found in
spleen and liver tissues between the SG9R and JOL3016
immunised groups (Figure8E–G). e bacterial load in
the PBS group was around log4 CFU/g, whereas the loads
in SG9R and JOL3016 immunised groups were reduced
to less than log1 CFU/g. Immunisation with detoxified
SG strains, particularly JOL3016, prevented adverse reac-
tions, promoted significant weight gain, and provided
robust protection against lethal challenges. ese find-
ings highlight the potential of detoxified SG strains in
effectively safeguarding against SG infection while sup-
porting healthy growth in chickens.
Histopathological examination
A histopathological evaluation of the spleen, liver, and
cecum tissues (Figure 9A–C) was conducted one week
after an oral challenge with the SG WT 422 strain. e
spleen tissues of naïve birds, the white pulp (lymphatic
tissues) and the red pulp (venous sinuses) were clearly
differentiated. Immunised birds with JOL3016, JOL3015,
and SG9R strains showed substantial preservation of this
tissue architecture. In contrast, the PBS control group
exhibited a markedly expanded white pulp, indicating
severe infection and inflammation (Figure 9A). In the
liver tissues of the PBS group, severe necrotic discoloura-
tions were evident, reflecting extensive tissue damage.
Liver tissues from immunised birds were comparable to
those of the naïve group, although infiltration of Kupffer
cells was observed across all groups, suggesting an active
but controlled immune response (Figure9B).
Analysis of the cecum tissues via histopathology
showed significant erosion, crypt abscesses, and signs
of oedema in the PBS group, indicating a severe bacte-
rial infection. In contrast, immunised chickens showed
considerable protection whether vaccinated with detoxi-
fied SG strains or the SG9R vaccine strain. eir cecum
tissues were largely free from these severe pathological
signs, demonstrating the effectiveness of immunisation
in mitigating infection-induced tissue damage. ese his-
topathological findings emphasise the protective efficacy
of the JOL3016, JOL3015, and SG9R vaccine strains (Fig-
ure 9C). Compared to the non-immunised PBS group,
immunised birds maintained a closer resemblance to
naïve tissue architecture across vital organs, significantly
reducing infection-related damage and inflammatory
responses.
Discussion
Fowl typhoid remains a significant concern in the poul-
try industry, particularly in developing regions where it
inflicts substantial economic losses [22]. e causative
agent, SG, not only impacts productivity but also poses
risks to animal welfare and public health. e SG9R vac-
cine is widely used to mitigate the disease; however, con-
cerns regarding its safety and efficacy remain prevalent.
Our study aims to address these concerns by engineer-
ing SG strains that have been attenuated through tar-
geted genetic modifications, thereby increasing the safety
and effectiveness of the vaccine. While effective in many
cases, the SG9R vaccine presents several limitations
that hinder its widespread use and effectiveness. Con-
cerns about potential reversion to virulence and endo-
toxicity raise questions about its long-term efficacy and
safety [23, 24]. e risk of SG9R reversion during field
outbreaks poses a significant challenge, highlighting the
need for alternative vaccine candidates. Moreover, the
residual pathogenicity of SG9R, particularly in immuno-
compromised hosts, underscores the urgency to develop
safer vaccine options [25]. Given the pivotal role of LPS
in the pathogenesis of SG and host immune responses
[26], our study focused on modifying the structure of
LPS to improve vaccine safety and immunogenicity. LPS
is a key virulence factor and immunogen, making it an
attractive target for vaccine development. By targeting
the virulence genes and genes involved in LPS biosynthe-
sis and modification, such as lon, rfaL, pagL, and arnT,
Page 16 of 20
Aganjaetal. Veterinary Research (2025) 56:2
we aimed to attenuate SG strains while preserving their
immunogenicity. e Lon protease functions as a global regulator of
bacterial virulence. erefore, its deletion could cause
the overexpression of several invasion-related genes by
Figure8 Evaluation of immunised chicken upon challenge. A Body Weight Alteration. Changes in chicken body weight were recorded
pre- and post-challenge to assess the effect of immunisation on overall health and the degree of protection against the wild-type challenge.
B Survival Rate. The survival of immunised birds challenged with the wild-type strain was compared with that of the non-immunised group.
A Kaplan-Meier survival curve was developed using mortality records over 15 days post-challenge. C Liver Morphology. Morphological changes
in the liver were examined for hepatic lesions post-challenge. D Spleen Morphology. The spleen was examined for splenomegaly and other
morphological changes post-challenge. E Spleen Weight. Post-challenge spleen weights were measured and compared with those of naïve birds.
The bacterial load of the wild-type challenge strain was assessed at 7 days post-infection for (F) Spleen. G Liver. Data were analysed by multiple
unpaired t-tests, with significant differences from the PBS control presented as *p < 0.05, **p < 0.01, and ***p < 0.001.
Page 17 of 20
Aganjaetal. Veterinary Research (2025) 56:2
promoting antigen presentation. e rfaL gene encodes
O-antigen ligase, which is essential for properly attaching
the O-antigen component to the lipid A core component.
e lack of the rfaL gene confers a truncated version of
the LPS structure, which has proven essential in pro-
viding DIVA capability [9]. e other two gene targets,
Figure9 Histopathological changes and microscopic lesions in chickens orally infected with the wild‑type strain. Chickens were orally
infected with 1 × 106 CFU/bird of Salmonella Gallinarum wild-type strain. Histopathological analysis of the internal organs was performed using
H&E staining. A Spleen. Altered cellular alignment and tissue architecture were visualised in the spleen tissues (200×). In the PBS control group,
degeneration and necrosis in the white pulp were observed, indicated by arrows. B Liver. Altered tissue architecture and inflammatory lesions
characterised by marked infiltration of heterophils and lymphocytes with degeneration and necrosis were observed in the liver tissues (200×).
Arrows highlight inflammatory lesions in the liver. C Cecum. Tissue disturbance in the cecum with thickened and shortened villi structures
was noted in the PBS group compared to vaccinated groups (40×). Black arrows denote immune cell infiltration, and blue arrows indicate
the shortening and congestion of villi. The organs of uninfected chickens (naïve) were used as the control. Data were visualised and analysed using
light microscopy. The scale bar: 50 μm for spleen and liver, and 10 μm for cecum.
Page 18 of 20
Aganjaetal. Veterinary Research (2025) 56:2
arnT and pagL, play crucial roles in modifying lipid A,
a component of LPS, thereby influencing bacterial viru-
lence and host immune response [27–29]. e addition
of L-Ara4N by ArnT changes the structure of lipid A,
decreasing its negative charge and enhancing its bac-
terial resistance to host defences [30, 31]. Conversely,
PagL-mediated deacylation reduces LPS hydrophobic-
ity, potentially evading host immune detection [27, 32].
ese modifications highlight the complex interplay
between bacterial adaptation and host immune evasion
strategies. Figure1 represents the concept behind lipid A
modification by our selected gene targets in the present
study.
Our study employed a well-established lambda red
recombineering approach to engineer attenuated SG
strains with targeted in-frame deletions of lon, rfaL, pagL,
and arnT genes (Additional file 1) in the SG genome.
ese deletions significantly modified the LPS structure,
including changes in core oligosaccharides, O-antigen
attachment, surface charge, and lipid A composition [9,
13, 15]. Importantly, these modifications aimed to reduce
endotoxicity while maintaining vaccine efficacy. e
engineered SG strains were characterised phenotypi-
cally and biologically, revealing altered surface properties
(Figure2A) and reduced haemolytic activity (Figure2B).
Truncation of the O-antigen component was confirmed
by acriflavine agglutination assay (Figure 2C) and LPS
Western blot (Figure 2D), which revealed a complete
absence of the O-antigen component.
e modified LPS structure results in a rough surface
that increases the hydrophobicity and causes the cells to
aggregate and settle. Such modifications change the phe-
notypic features and affect biological characteristics, as
evidenced by a decreased hemolysis activity. e signifi-
cant decrease in hemolysis caused by the mutant strains
indicates a reduction in virulence. is decrease needs to
be considered when developing the vaccine strain, espe-
cially as the hemolysins of Salmonella play an essential
role in intra-macrophage survival, killing cells, and pro-
longed systemic salmonellosis [33]. Moreover, examining
bacterial growth kinetics offers insights into the differ-
entiated physiological state of bacteria [34]. Our growth
assessment in this study revealed distinctive growth
kinetics compared to wild-type and commercial SG9R
strains. e complete elimination of three genes from
each detoxified SG strain, namely, lon, rfaL, and arnT
from JOL 3015 and lon, rfaL, and pagL from JOL3016,
resulted in a comparatively lower growth rate than the
wild-type and SG9R vaccine strain at early time points
of growth, however reducing the gap with an increase in
incubation time (Figure3A, B).
In particular, JOL3016 was almost equal in bacte-
rial number to WT and SG9R within a 28h incubation
period, demonstrating that the strain was not overly
attenuated. e selected genetic markers did not signifi-
cantly affect bacterial adhesion or virulence, especially
for the JOL3016 strain that carries pagL deletion. is
outcome ensures that these strains retain their capability
to invade host cells, which is essential for better antigen
presentation (Figure 3C, D) [35]. Acidic and oxidative
stress survival assays also revealed that JOL3016 is com-
parable to the SG9R vaccine strain. However, JOL3015
was found to have a slightly lower tolerance to acidity
and oxidative conditions than the SG9R and JOL3016
(Figure3E, F). ese findings enable the detoxified SG
strains to potentially undergo rapid clearance from
the intracellular oxidative stress without persisting as
a chronic infection, which may be an important safety
consideration.
Further to note is that the mutant strains induced low-
ered cytotoxic responses without significantly damaging
epithelial monolayers of Hela cells. e findings here also
showed that JOL3016 was comparable to SG9R, while the
lowest cytotoxic response was exhibited by the JOL3015
strain, exacerbating its stronger attenuation phenotype
(Figure4A, B). e safety assessments undertaken in the
study also demonstrated minimal adverse reactions and
reduced endotoxicity in inoculated chickens with detoxi-
fied strains. For example, no deaths occurred when birds
were inoculated with detoxified SG strains or SG9R,
while infection and mortality rates were significant when
inoculated with the SG WT 422 strain. A comparison of
two inoculation doses administered via the IM route, at
1 × 107 and 1 × 108 CFU/bird, was found to be completely
safe for young chickens. Additionally, this treatment did
not affect chicken growth to the same extent as SG9R.
It is worth noting that the examination of bacterial
persistence in the spleen, liver, and cloacal swabs did not
reveal any significant difference between the two inocula-
tion doses (high and low). However, by day 14 post-inoc-
ulation, bacterial persistence had reduced to less than log
2 in all organ samples, spleen, liver, and cloacal swabs
collected from challenged chicken. ese findings under-
score the safety and potential of the engineered strains
as vaccine candidates (Figure5). To further evaluate the
reduced levels of endotoxicity responses, we investi-
gated the levels of pro-inflammatory cytokines in blood
samples. e results showed significantly lower levels of
markers for pro-inflammatory cytokines, such as tumour
necrosis factor-alpha (TNF-α), Interleukin-1β (IL-1β),
and Interleukin-γ (IFN-γ), even lower than those in the
SG9R vaccine strain (Figure6). ese observations were
further exacerbated in the histopathological examination
of spleen and liver tissues. e examination showed low-
ered signs of inflammation marked by red and white pulp
Page 19 of 20
Aganjaetal. Veterinary Research (2025) 56:2
distribution in the spleen and necrotic lesions, as well as
severe inflammation in liver tissues.
e evaluation of humoral and cell-mediated immune
responses showed that the engineered SG strains elicited
a robust immune response comparable to the response
elicited by the commercial vaccine strain SG9R. As live
attenuated vaccine strains, chicken immunisation has
resulted in a significant engagement of CD3+CD4+ and
CD3+CD8+ differentiation (Figure7D, E). CD3+CD4+ T
cells also play a crucial role in activating macrophages
and CD8+ T cells, ensuring a robust and coordinated
immune response. eir role is pivotal in generating
a strong humoral response, essential for neutralising
pathogens and preventing infection spread. On the other
hand, CD3+CD8+ T cells, known as cytotoxic T cells,
are directly involved in eliminating infected cells. ey
recognise and kill cells presenting specific antigens on
their surface, typically through the major histocompat-
ibility complex class I (MHC I) pathway. is cytotoxic
activity is essential for controlling intracellular pathogens
such as SG by limiting bacterial replication and spread-
ing within the host. Furthermore, CD8+ T cells produce
various cytokines that contribute to the overall immune
response and aid in the recruitment and activation of
other immune cells.
e collective outcome and effectiveness of protective
immune responses induced by novel vaccine candidates
are clearly demonstrated in post-challenged pathologi-
cal assessments. Importantly, post-challenge survival
rates and histopathological analyses validate the protec-
tive efficacy of the engineered strains against wild-type
SG challenge (Figures8, 9). ese results emphasise the
potential of the engineered SG strains to induce protec-
tive immunity while minimising adverse reactions and
pathological manifestations.
In conclusion, this study sheds light on the promising
potential of engineered SG strains featuring modified
LPS structures as safe and efficacious vaccine candi-
dates against fowl typhoid. Notably, comparative analyses
against the commercial vaccine strain SG9R underscored
the superiority of the designed strains in terms of reduced
endotoxicity and retained protective efficacy. ese find-
ings highlight the importance of further research to
investigate the long-term efficacy and real-world applica-
tion of the engineered strains in poultry populations.
Abbreviations
FT fowl typhoid
LPS llipopolysaccharide
SG Salmonella enterica serovar Gallinarum (Salmonella Gallinarum)
pagL PhoP/PhoQ-activated gene
arnT l-Ara4N transferase gene
DIVA differentiate infected from vaccinated animals
ELISA enz yme-linked immunosorbent assay
catR chloramphenicol resistance gene
OD optical density
RBC red blood cell
PBMCs peripheral blood mononuclear cells
PBS phosphate-buffered saline
WT wild-type
dpi days post-inoculation
H&E haematoxylin and eosin staining
RT room temperature
TNF-α tumour necrosis factor-alpha
IL-1β Interleukin-1β
IFN-γ Interleukin-γ
Supplementary Information
The online version contains supplementary material available at https:// doi.
org/ 10. 1186/ s13567- 024- 01413-8.
Additional le1. Conrmation of deletion of lon , rfaL , pagL ,
and ar nT genes. Flanking primers were used to confirm the deletion
of respective genes. M = DNA marker, WT = Wild-type, and 1, 2, and
3 = Samples.
Additional le2. Gating strategy used for T‑cell subsets , a repre‑
sentative sample for the JOL3016 group. (A) Gating of Total lympho-
cytes. (B) Gating of CD3 + T cells from total lymphocytes. (C) Gating of
CD3 + CD4 + and CD3 + CD + T cells from CD3 + T cells.
Additional le3. Measurement of body temperature (°C) at
post‑immunisation.
Acknowledgements
The authors would like to acknowledge the support of the National University
Development Project, Jeonbuk National University, and the CURF at Jeonbuk
National University.
Authors’ contributions
RPA: conceptualisation, investigation, methodology, validation, formal analysis,
writing—original draft, writing–review and editing. JK: formal analysis, meth-
odology, writing—review and editing. AS: writing—review and editing. JHL:
conceptualisation, resources, supervision, funding acquisition, writing–review
and editing. All authors read and approved the final manuscript.
Funding
This work was supported by the Technology Development Program
(S3383209), funded by the Ministry of SMEs and Startups (MSS, Korea) and by
the National University Development Project at Jeonbuk National Univer-
sity in 2023. The histopathological analysis was performed in the Center for
University-wide Research Facilities (CURF) at Jeonbuk National University.
Availability of data and materials
Raw data reported in the manuscript can be made available upon request
from the corresponding author.
Declarations
Ethics approval and consent to participate
All animal experiments in this study were conducted under the Jeonbuk
National University Animal Ethics Committee (NON2023-135-001) guidelines,
following the Korean Council on Animal Care and the Korean Animal Protec-
tion Law, 2007: Article 13.
Competing interests
The authors declare that they have no competing interests.
Author details
1 College of Veterinary Medicine, Jeonbuk National University, Iksan Campus,
Iksan 54596, Republic of Korea. 2 College of Veterinary Medicine and Institute
of Animal Transplantation, Jeonbuk National University Campus, Iksan 54596,
Republic of Korea.
Page 20 of 20
Aganjaetal. Veterinary Research (2025) 56:2
Received: 17 June 2024 Accepted: 16 September 2024
References
1. Zhou X, Kang X, Zhou K, Yue M (2022) A global dataset for prevalence of
Salmonella Gallinarum between 1945 and 2021. Sci Data 9:495
2. Alves Batista DF, de Freitas Neto OC, Maria de Almeida A, Maboni G, de
Carvalho TF, de Carvalho TP, Barrow PA, Berchieri AJ (2018) Evaluation of
pathogenicity of Salmonella Gallinarum strains harbouring deletions in
genes whose orthologues are conserved pseudogenes in S. Pullorum.
PLoS One 13:e0200585
3. Arora D, Kumar S, Jindal N, Narang G, Kapoor PK, Mahajan NK (2015)
Prevalence and epidemiology of Salmonella enterica Serovar Gallinarum
from poultry in some parts of Haryana, India. Vet World 8:1300–1304
4. Wigley P, Hulme S, Powers C, Beal R, Smith A, Barrow P (2005) Oral
infection with the Salmonella enterica Serovar Gallinarum 9R attenuated
live vaccine as a model to characterise immunity to fowl typhoid in the
chicken. BMC Vet Res 1:2
5. Lee YJ, Mo IP, Kang MS (2007) Protective efficacy of live Salmonella gal-
linarum 9R vaccine in commercial layer flocks. Avian Pathol 36:495–498
6. Kwon YK, Kim A, Kang MS, Her M, Jung BY, Lee KM, Jeong W, An BK, Kwon
JH (2010) Prevalence and characterization of Salmonella Gallinarum in
the chicken in Korea during 2000 to 2008. Poult Sci 89:236–242
7. Huang XY, Ansari AR, Huang HB, Zhao X, Li NY, Sun ZJ, Peng KM, Zhong J,
Liu HZ (2017) Lipopolysaccharide mediates immuno-pathological altera-
tions in young chicken liver through TLR4 signaling. BMC Immunol 18:12
8. Brandenburg K, Wiese A (2004) Endotoxins: relationships between struc-
ture, function, and activity. Curr Top Med Chem 4:1127–1146
9. Senevirathne A, Hewawaduge C, Sivasankar C, Lee JH (2022) Prospective
lipid-A altered live attenuated Salmonella Gallinarum confers protectiv-
ity, DIVA capability, safety and low endotoxicity against fowl typhoid. Vet
Microbiol 274:109572
10. Breazeale SD, Ribeiro AA, McClerren AL, Raetz CR (2005) A formyltrans-
ferase required for polymyxin resistance in Escherichia coli and the
modification of lipid A with 4-Amino-4-deoxy-l-arabinose. Identification
and function oF UDP-4-deoxy-4-formamido-l-arabinose. J Biol Chem
280:14154–14167
11. Senevirathne A, Hewawaduge C, Lee JH (2022) Assessing an O-antigen
deficient, live attenuated Salmonella gallinarium strain that is DIVA com-
patible, environmentally safe, and protects chickens against fowl typhoid.
Dev Comp Immunol 133:104433
12. Kirthika P, Jawalagatti V, Senevirathne A, Lee JH (2022) Coordinated
interaction between Lon protease and catalase-peroxidase regulates
virulence and oxidative stress management during salmonellosis. Gut
Microbes 14:2064705
13. Kirthika P, Senevirathne A, Jawalagatti V, Park S, Lee JH (2020) Deletion of
the lon gene augments expression of Salmonella Pathogenicity Island
(SPI)-1 and metal ion uptake genes leading to the accumulation of bac-
tericidal hydroxyl radicals and host pro-inflammatory cytokine-mediated
rapid intracellular clearance. Gut Microbes 11:1695–1712
14. Matsuda K, Chaudhari AA, Kim SW, Lee KM, Lee JH (2010) Physiology,
pathogenicity and immunogenicity of lon and/or cpxR deleted mutants
of Salmonella Gallinarum as vaccine candidates for fowl typhoid. Vet Res
41:59
15. Aganja RP, Sivasankar C, Hewawaduge C, Lee JH (2022) Safety assess-
ment of compliant, highly invasive, lipid A-altered, O-antigen-defected
Salmonella strains as prospective vaccine delivery systems. Vet Res 53:76
16. Ishiguro A, Nishioka M, Morishige A, Kawano R, Kobayashi T, Fujinaga A,
Takagi F, Kogo T, Morikawa Y, Okayama N, Mizuno H, Aihara M, Suehiro Y,
Yamasaki T (2020) What is the best wavelength for the measurement of
hemolysis index? Clin Chim Acta 510:15–20
17. Guo R, Jiao Y, Li Z, Zhu S, Fei X, Geng S, Pan Z, Chen X, Li Q, Jiao X (2017)
Safety, protective immunity, and DIVA capability of a rough mutant
Salmonella Pullorum vaccine candidate in broilers. Front Microbiol 8:547
18. Hewawaduge C, Senevirathne A, Sivasankar C, Lee JH (2023) The impact
of lipid A modification on biofilm and related pathophysiological
phenotypes, endotoxicity, immunogenicity, and protection of Salmonella
Typhimurium. Vet Microbiol 282:109759
19. Bertram EM, Jilbert AR, Kotlarski I (1997) Optimization of an in vitro assay
which measures the proliferation of duck T lymphocytes from peripheral
blood in response to stimulation with PHA and ConA. Dev Comp Immu-
nol 21:299–310
20. Aganja RP, Sivasankar C, Lee JH (2023) AI-2 quorum sensing controlled
delivery of cytolysin-A by tryptophan auxotrophic low-endotoxic Salmo-
nella and its anticancer effects in CT26 mice with colon cancer. J Adv Res
61:83–100
21. Datsenko KA, Wanner BL (2000) One-step inactivation of chromosomal
genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S
A 97:6640–6645
22. Ojima S, Okamura M, Osawa N, Tamura A, Yoshioka K, Kashimoto T,
Haneda T, Ono HK, Hu DL (2021) Characteristics of systemic infection and
host responses in chickens experimentally infected with Salmonella enter-
ica serovar Gallinarum Biovar Gallinarum. J Vet Med Sci 83:1147–1154
23. Beylefeld A, Abolnik C (2023) Salmonella gallinarum strains from out-
breaks of fowl typhoid fever in Southern Africa closely related to SG9R
vaccines. Front Vet Sci 10:1191497
24. Van Immerseel F, Studholme DJ, Eeckhaut V, Heyndrickx M, Dewulf
J, Dewaele I, Van Hoorebeke S, Haesebrouck F, Van Meirhaeghe H,
Ducatelle R, Paszkiewicz K, Titball RW (2013) Salmonella Gallinarum field
isolates from laying hens are related to the vaccine strain SG9R. Vaccine
31:4940–4945
25. Kwon HJ, Cho SH (2011) Pathogenicity of SG 9R, a rough vaccine strain
against fowl typhoid. Vaccine 29:1311–1318
26. Yang KH, Lee MG (2008) Effects of endotoxin derived from Escherichia coli
lipopolysaccharide on the pharmacokinetics of drugs. Arch Pharm Res
31:1073–1086
27. Kawasaki K (2012) Complexity of lipopolysaccharide modifications in: its
effects on endotoxin activity, membrane permeability, and resistance to
antimicrobial peptides. Food Res Int 45:493–501
28. Kawasaki K, Ernst RK, Miller SI (2005) Inhibition of Salmonella enterica
serovar typhimurium lipopolysaccharide deacylation by aminoarabinose
membrane modification. J Bacteriol 187:2448–2457
29. Kawasaki K, Ernst RK, Miller SI (2004) 3-O-deacylation of lipid A by PagL, a
PhoP/PhoQ-regulated deacylase of Salmonella typhimurium, modulates
signaling through toll-like receptor 4. J Biol Chem 279:20044–20048
30. Ernst RK, Guina T, Miller SI (2001) Salmonella typhimurium outer mem-
brane remodeling: role in resistance to host innate immunity. Microbes
Infect 3:1327–1334
31. Raetz CR, Whitfield C (2002) Lipopolysaccharide endotoxins. Annu Rev
Biochem 71:635–700
32. Park BS, Song DH, Kim HM, Choi BS, Lee H, Lee JO (2009) The structural
basis of lipopolysaccharide recognition by the TLR4-MD-2 complex.
Nature 458:1191–1195
33. Agrawal RK, Singh BR, Babu N, Chandra M (2005) Novel haemolysins of
Salmonella enterica spp. Enterica Serovar Gallinarum. Indian J Exp Biol
43:626–630
34. Ferenci T (1999) Growth of bacterial cultures’ 50 years on: towards an
uncertainty principle instead of constants in bacterial growth kinetics.
Res Microbiol 150:431–438
35. Mukherjee S, Bassler BL (2019) Bacterial quorum sensing in complex and
dynamically changing environments. Nat Rev Microbiol 17:371–382
36. Doublet B, Douard G, Targant H, Meunier D, Madec JY, Cloeckaert A (2008)
Antibiotic marker modifications of lambda red and FLP helper plasmids,
pKD46 and pCP20, for inactivation of chromosomal genes using PCR
products in multidrug-resistant strains. J Microbiol Methods 75:359–361
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.
