Deletion of the lon gene augments expression of Salmonella Pathogenicity Island (SPI)-1 and metal ion uptake genes leading to the accumulation of bactericidal hydroxyl radicals and host pro-inflammatory cytokine-mediated rapid intracellular clearance

ABSTRACT

In the present study, we characterized the involvement of Lon protease in bacterial virulence and intracellular survival in Salmonella under abiotic stress conditions resembling the conditions of a natural infection. Wild type (JOL401) and the lon mutant (JOL909) Salmonella Typhimurium were exposed to low temperature, pH, osmotic, and oxidative stress conditions and changes in gene expression profiles related to virulence and metal ion uptake were investigated. Expression of candidate genes invF and hilC of Salmonella Pathogenicity Island (SPI)-1 and sifA and sseJ of SPI-2 revealed that Lon protease controls SPI-1 genes and not SPI-2 genes under all stress conditions tested. The lon mutant exhibited increased accumulation of hydroxyl (OH·) ions that lead to cell damage due to oxidative stress. This oxidative damage can also be linked to an unregulated influx of iron due to the upregulation of ion channel genes such as fepA in the lon mutant. The deletion of lon from the Salmonella genome causes oxidative damage and increased expression of virulence genes. It also prompts the secretion of host pro-inflammatory cytokines leading to early clearance of the bacteria from host cells. We conclude that poor bacterial recovery from mice infected with the lon mutant is a result of disrupted bacterial intracellular equilibrium and rapid activation of cytokine expression leading to bacterial lysis.

KEYWORDS:

• Salmonella
• Lon protease
• oxidative stress
• metal ion
• cytokines

Introduction

Salmonella enterica causes a spectrum of diseases in animals and humans, ranging between mild infection to severe infection that can be fatal¹. Different strains of Salmonella enterica Typhimurium cause enteric infections in humans and other warm-blooded animals, whereas the serovar Typhi is restricted to humans causing typhoid fever. All Salmonella species possess Salmonella Pathogenicity Island I (SPI-1) despite variations in virulence factors among different serotypes. Coordinated expression of genes belonging to SPI-1 and SPI-2 are important for initial entry and the subsequent establishment of infection in a target host. The SPI-1 encodes proteins including a Salmonella type III secretion system (TTSS) that is responsible for injecting effector molecules that alter cytoskeletal arrangement in the host and help bacterial internalization. Once the bacterium is in the host cell, genes in SPI-2 are expressed that encodes another type of TTSS that secretes effectors diverting endocytic trafficking to prevent lysosome recruitment and fusion of Salmonella-containing vacuoles. This process helps host-adapted Salmonella reach extraintestinal organs and establish a systemic infection. Salmonella invasion is a complex and intricate process that essentially controls virulence. Concerning invasion and virulence of Salmonella, studies have revealed that Lon protease plays a significant role during early infection and colonization of Salmonella.² As a member of the AAA+ superfamily of proteases that also includes ClpA, ClpB, and FtsH, Lon has been extensively studied in Escherichia coli.^3–5 The Lon protease plays a vital role in protein quality control by eliminating faulty proteins during stress responses and activating pathways to prevent any further damage to the bacterial cell.⁶ Moreover, recent reports indicate the involvement of Lon protease in antimicrobial resistance, further broadening its role in bacterial survival.⁷

The involvement of Lon in Salmonella SPI-1 gene expression has been previously reported, which proposed a link between the SPI-1 regulator HilA and Lon protease. HilA is an OmpR/ToxR family protein⁸ that is activated by the transcriptional regulator InvF, which in turn regulates the transcriptional activation of effector genes. In Salmonella, hilA expression is modulated by the HilC and HilD transcriptional regulators.⁹ Therefore, the interaction of Lon with the SPI-1 axis of virulence modulation is worth further elaboration under various environmental and physiological stress conditions to investigate the survival and virulence behavior of Salmonella.² During the natural infection cycle, Salmonella encounters various stress conditions including a transition to increased body temperature, acidic pH in the stomach, high salt and osmolarity in bile salts, and oxidative stress in the lumen as a result of the host inflammatory response.¹⁰ During systemic infection, the host attempts to starve Salmonella by limiting nutrient and metal ion uptake; in response, Salmonella manipulates ion and nutrient uptake gene expression to counter the host restrictions.^11,12 Further, Salmonella attempts to circumvent the recruitment of antigen-presenting cells to the site of infection by interacting with the host immune pathways¹³ and exploits the metabolic pathways of the host for its survival.¹⁴

Apart from manipulating these host mechanisms, the bacteria modulate their biosynthesis machinery that determines their fitness and virulence.^14,15 As a facultative anaerobe, Salmonella undergoes metabolic reprogramming depending on the presence or absence of oxygen. One of the principal determinants of this process is redox homeostasis.¹⁶ On the other hand, reactive oxygen species (ROS) produced during microbial metabolism and the stress response can alter the redox state of a cell.¹⁷ Gene deletions or mutations hindering intracellular iron regulation is known to cause the accumulation of excess free cytosolic iron, leading to iron-catalyzed production of ROS.^18,19 This endogenous oxidative stress is regarded as an unwanted and damaging by-product of cellular metabolism.

Here, we describe the quantification of gene expression involved in pathogenesis including genes that encode SPI effector proteins, genes regulating iron uptake, and levels of hydroxyl (OH·) radical accumulation in lon- mutant Salmonella; JOL 909 after exposure to various conditions including acidic, cold, osmotic, and oxidative stress treatments. The changes in gene expression were compared with that of parental wild type strain Salmonella Typhimurium; JOL 401 and lon complemented Salmonella strain JOL 909::lon. We analyzed gene expression after 5 h of exposure and found that JOL 909 exhibited the upregulation of SPI-1 virulence-associated and iron uptake genes, followed by increased accumulation of OH· in the lon mutant suggesting the participation of Lon in modulating bacterial internal stability. Further, a distinct pro-inflammatory cytokine response in the host shows that Lon alters bacterial interaction with the host TLR; a hypothesis that needs further validation.²⁰

Although a previous study reports the inability of a lon mutant to survive phagocytic lysis, the lon-specific molecular mechanism was not elucidated¹. Here, our observations show that lon regulates intracellular homeostasis when Salmonella maneuvers through the host cell to establish an infection. In the absence of this regulator, an unchecked influx of iron cations occurs that reacts with H₂O₂ produced within the bacteria, as well as inflow from the host cells, leading to the formation of OH· radicals. Hydroxyl ion-mediated oxidative stress leads to DNA damage and subsequent bacterial lysis. The present work adds a new level of information about Lon protease and its role in tightly regulated bacterial stress-response pathways. Here, we show that Lon in Salmonella is essential for mediating regulated virulence behavior that may ensure the successful establishment of systemic infection.

Results

Characterization of a lon mutant in Salmonella

The Lon protease is a global regulator that regulates the expression of early virulence genes located in SPI-1. Mechanistic studies have revealed Lon as a negative regulator of SPI-1 that regulates and ensures the successful establishment of Salmonella in a target host cell. Herein, we investigated the involvement of the lon gene in virulence modulation of Salmonella upon exposure to various abiotic stress conditions. To examine this phenomenon, we used wild type strain ST JOL 401 and a lon deleted strain ST JOL 909. To further confirm that the ability to adhere and invade epithelial cells and metal ion uptake associated genes in JOL 909 are modulated as a result of lon knockout, a functional lon gene was cloned in a low copy-number vector, pWSK 29, and the resultant plasmid was used to transform JOL 909, resulting in strain JOL 909::lon. To assess whether the lon deletion causes growth defects in Salmonella, a growth curve analysis of all three strains was carried out. Results demonstrated that the lon deletion did not significantly alter the growth characteristics of Salmonella (Supplementary Figure 1). The JOL 909 strain exhibited enhanced adhesion and invasion compared to JOL 401. Such enhanced adhesion and invasion were inhibited by lon complementation in JOL 909, showing that Lon plays a crucial role in regulating both adhesion and invasion. The adhesive and invasive phenotype of the lon mutant JOL 909 was demonstrated to be similar in all three cell lines HeLa, HepG2, and Caco-2 (Figure 1a). However, when the JOL 401 and JOL 909 were exposed to anaerobic conditions their adhesion and invasion ability increased drastically. The increased invasion by JOL 909 was further evident in fluorescent microscopy conducted using CellTracker-labeled Salmonella; however, their population was significantly reduced in both non-phagocytic (data not shown) and phagocytic cells (Figure 1b,c) earlier than the wild type and complemented strains.

Figure 1. Role of lon in Adhesion and Invasion of HeLa and HepG2 and macrophage survival.

(A) Adhesion and Invasion of HeLa, HepG2, and Caco-2 monolayer were observed post-infection with JOL 401, JOL 909, and JOL 909::lon (MOI = 20). The data represent the adhesion and invasion (Log CFU/mL) ± the SD from at least three independent experiments performed in duplicates. P values of <0.05 were considered significantly different, and bars with the same lowercase letter are not significantly different from one another. (B) Fluorescent microscopy images of RAW cells infected with Salmonella (JOL 401, JOL 909, and JOL 909::lon) containing pHLJ65::gfp plasmid to visualize the live (green fluorescent) bacteria within the cells. (C) Intracellular survival of Salmonella wild type, lon mutant, and complemented strain in RAW cells.

Actin rearrangement by Salmonella

Salmonella induced actin rearrangement in a host cell is essential for Salmonella entry and survival. Such manipulation helps Salmonella in better vacuolization and prevents lysosomal engagement allowing them to thrive in the host. Such phenotypic changes in the host cell induced by JOL 401 and JOL 909 were recorded during the entry phase, characterized by the trigger or zipper mechanisms. In HeLa cells (Figure 1a), both JOL 401 and JOL 909 induced lamellipodia formation characterized by membrane ruffling and formation of appendages engulfing the bacteria. However, in HepG2 cells (Figure 2b), the strains did not induce the formation of membrane extensions most likely due to the engagement of the zipper mechanism. Although no lamellipodia formation was evident in Caco-2 cells, a mild yet distinct membrane ruffling was observed when infected with JOL 401 and JOL 909 (Figure 2c). These results show that the regulation of host actin polymerization is independent of a Lon-mediated regulatory pathway. We hypothesize that this rearrangement is also partly dependent on the host cell type. To further ensure independent behavior on actin rearrangement, microscopic observations were conducted at varying time intervals of 15, 30, 45, 60, 90, and 150 min post-inoculation.

Figure 2. Actin polymerization is a distinctive attribute of individual host cells.

Microscopic images show membrane ruffling and lamellipodia formation. (A) HeLa cells (B) HepG2 cells (C) Caco-2 cells were infected with stress exposed Salmonella (JOL 401, JOL 909, and JOL 909::lon; MOI = 20) for 30 min and stained with phalloidin (for F-actin) and DAPI (nuclear stain);1-actin condensation,2-lamellipodia. These observations demonstrate that Salmonella enters the HeLa cells vigorously rearranging the host actin filaments via a trigger mechanism. The interaction with HepG2 cells results in a zipper-type of actin rearrangement. The interaction with Caco-2 cells results in a minimum ruffling of the membrane.

Effect of lon on bacterial hydroxyl ion generation

The generation of reactive oxygen species (ROS) is a hallmark of the bacterial stress response. In bacteria, ROS are readily detoxified by enzymatic conversion of ROS (OH·) into hydrogen peroxide and subsequently to water and oxygen. However, in the presence of iron ions, the Fenton reaction can generate OH· that are lethal to bacterial cells.²¹ To elucidate the kinetics of OH· generation in JOL 401, JOL909, and JOL909::lon, cells were exposed to various naturally-occurring stress conditions for 5 h. JOL 401 was able to effectively mitigate OH· generation under cold and oxidative stress; however, the production of OH· was significantly higher under osmotic and acidic pH conditions. Under all stress conditions, JOL 909 appeared to have produced higher levels of OH·. The maximum level of OH· production in JOL 909 was observed under oxidative stress, suggesting the role of Lon protease in mitigating oxidative stress. The involvement of Lon seems to be comparatively less with osmotic and acidic stress conditions. The results were corroborated with JOL 909::lon, which effectively diminished OH· production. These results imply the critical involvement of Lon protease activity in mitigating oxidative stress in the phagosomes of host cells²² (Figure 3). It is worthwhile to note that JOL 909::lon exhibited lower levels of OH· compared to JOL 401. As the overproduction of Lon can prove fatal for bacterial cells, these low levels may be due to slightly higher levels of Lon protein in JOL 909 complemented in trans by a low-copy-number plasmid.

Figure 3. Endogenous Hydroxyl radical (OH·) formation during stress treatments in the lon mutant.

JOL 401, JOL 909 and JOL 909::lon were exposed to different abiotic stress (cold, oxidative, osmotic, and acidic) for 5 h and their endogenous hydroxyl radical levels were measured and the values were compared with the bacterial cells grown at 37ºC without any stress inducers. The data was represented as relative OH• formation ± the SD from at least three independent experiments performed in duplicates. Means were compared by Tukey’s multiple comparison test. Similar lowercase letters indicate the non-significant difference of means (p > .05). Significant difference (*p < .05, ***p < .0001) compared to the wild type control.

Effect of lon deletion on metal ion uptake and virulence gene expression

To investigate the role of lon in Salmonella virulence under the aforementioned stress conditions, the expression profile of four genes encoding outer membrane proteins, namely ompD, ompF, fepA, and csgB was studied by qRT-PCR. Among these genes, csgB gene expression was significantly upregulated under cold stress, which lead us to believe that Lon protease, a member of the heat shock protein (Hsp) family, also regulates the expression of genes that are primarily activated below 37ºC. The other three outer membrane genes were downregulated in response to cold stress. Under osmotic and acidic stress conditions, only fepA expression appeared to be regulated by Lon. When exposed to oxidative stress, ompF, and fepA expression were upregulated (Figure 4).

Figure 4. Expression of bacterial cell surface protein genes during stress treatments.

JOL 401, JOL 909 and JOL 909::lon were subjected to abiotic (oxidative-1 mM H₂O₂; acidic-3.0 pH;osmotic- 5% NaCl and cold-4ºC) stress treatment for 5 h. The regulation of the ompF, ompD, fepA, and csgB were evaluated post-stress treatments. The results are expressed as mean fold change in the transcription levels of ompF, ompD, fepA, and csgB genes compared to unstimulated bacterial cells. The data was represented as relative mean fold change ± the SD from at least three independent experiments performed in duplicates. Means were compared by Tukey’s multiple comparison test. Similar lowercase letters indicate the non-significant difference of means (p > .05). Significant difference (*p < .05, **p < .01, ***p < .0001) compared to the wild type control. ompD- Outer membrane protein D; ompF- Outer membrane protein F; fepA- Ferric Enterobactin Transporter; csgB- Minor Curlin subunit.

The upregulation of iron uptake regulator genes (fur regulon and sitA) and iron sequestration gene (dps) under various stress conditions was distinct. In particular, the fur regulon was highly upregulated in JOL 909 under osmotic and oxidative stress conditions. The increase in the mRNA level was similar to the increased expression levels of fepA, which is an iron ion channel on the bacterial cell surface (Figure 5). Also, JOL 909 exhibited a surge in the expression of corA, mgtA, and mgtB compared to JOL 401 (Figure 6), all three of which are transporters of divalent magnesium ions while corA is also associated with the uptake of ferrous ions.

Figure 5. Expression of regulatory genes that participate in iron uptake and sequestration during stress exposure.

Post-exposure to the abiotic stress factors the transcription levels of sitA, dps, and fur were evaluated in JOL 401, JOL 909, JOL 909::lon, JOL 2768, and JOL 2769. The results are expressed as mean fold change in the transcription levels of sitA, dps, and fur genes compared to unstimulated bacterial cells. The data was represented as relative mean fold change ± the SD from at least three independent experiments performed 2compared by Tukey’s multiple comparison test. Similar lowercase letters indicate the non-significant difference of means (p > .05). Significant difference (*p < .05, **p < .01) compared to the wild type control.

Figure 6. Regulation of Magnesium uptake associated genes.

The mRNA expression of Magnesium uptake genes was studied after the bacterial cells were subjected to stress treatments for 5 h. The mean fold change in the levels of corA, mgtA, and mgtB in JOL 401, JOL 909, and JOL 909::lon were recorded. The data was represented as relative mean fold change ± the SD from at least three independent experiments performed in duplicates. Means were compared by Tukey’s multiple comparison test. Similar lowercase letters indicate the non-significant difference of means (p > .05). Significant difference (*p < .05, **p < .01) compared to the wild type control.

Apart from modulating gene expression levels of outer membrane proteins under various stress conditions, Lon regulates the expression of SPI-1 transcriptional factors. The invF and hilC (Figure 7a) genes were significantly upregulated under all stress conditions in JOL 909 compared to JOL 401, attributing it to be Lon-dependent. On the other hand, the SPI-2 genes sifA and sseJ were upregulated uniformly in both JOL 401 and JOL 909 (Figure 7b). The levels of sifA were observed to have increased up to 4-fold during oxidative and osmotic stress treatments. A similar trend was observed for sseJ, whose levels were upregulated 2-fold during oxidative, osmotic, and acidic stress. These observations suggest that the regulation of a cluster of genes in SPI-2 is partly independent of pathways involving Lon protease. SPI-1 is essential for invasion whereas SPI-2 is crucial for intracellular survival and systemic persistence.^23,24

Figure 7. Expression of a virulence-associated gene from SPI 1 and 2 gene clusters.

mRNA levels of (A) SPI1 genes-invF and hilC and (B)SPI2 genes- sifA and sseJ were recorded as relative mean fold changes using rrsG as the endogenous control. The data was represented as relative mean fold change ± the SD from at least three independent experiments performed in duplicates. Means were compared by Tukey’s multiple comparison test. Similar lowercase letters demonstrate the non-significant difference of means (p > .05). Significant difference (*p < .05, **p < .01, ***p < .001) or no significant difference (^nsp>.05) compared to the wild type control.

fepA knockout and lon expression

Based on the aforementioned observations, we theorized that fepA is transcriptionally regulated by lon under stress conditions especially during oxidative stress. To further understand the fepA-lon interaction, we constructed JOL 401ΔfepA and JOL 909ΔfepA and refer to them as JOL 2472 and JOL 2473, respectively (Supplementary Figure 2). To study the role of fepA in the regulation of lon expression, JOL 401 and JOL 2472 were exposed to the above-mentioned stress conditions for 5 h and their RNA was isolated to evaluate lon expression (Supplementary Figure 3). The expression of lon was upregulated in JOL 401 by ~10-fold under all the stress treatments, peaking at 15-fold changes on average during oxidative stress. On the other hand, lon expression increased up to 2-fold on average in JOL 2472 after treatment (Supplementary Figure 3). This was further followed by the downregulation of iron uptake and sequestration genes in JOL 2472 and JOL 2473 (Figure 5). Additionally, under the anaerobic condition, the lon expression decreased by 2-fold in both JOL 401 and JOL 2473.

In vitro bacterial survival and recovery from mice

The presence of colony-forming units (CFU) on BGA agar showed the ability of JOL 401 and JOL 909 to survive in vitro after exposure to abiotic stress conditions. Oxidative stress caused characteristic damage to JOL 909 wherein <4 log CFU/mL bacteria were observed compared to a log CFU close to 6 per mL for JOL 401 (Figure 8a,b). Although all the stress conditions hampered JOL 909 survival, the oxidative stress-induced damage was distinct. To study if the in vitro observations were consistent with the in vivo bacterial clearance, the spleens of infected mice were aseptically collected and bacterial load was evaluated. The results corroborated with the in vitro survival pattern wherein JOL 909 seemed to be eliminated by host phagocytosis and exerting oxidative stress on the invading bacteria. By day 12 post-infection, splenomegaly was evident, and <3 log CFU/mL JOL 909 were recovered from the spleen (Figure 8c).

Figure 8. Bacterial recovery and survival to study the involvement of lon in bacterial longevity.

(A) In vitro bacterial survival post-stress exposure was evaluated by serial dilution and plating of stress endured JOL 401, JOL 909, and JOL 909::lon on BGA plate. (B) The bacterial recovery was calculated by plating the serially diluted homogenized spleen (1 g) on BGA plates. (C) The spleens were collected aseptically from mice (n = 3; infected with JOL 401, JOL 909, and JOL 909::lon) on days 1, 6, 9, and 12 post-infection for evaluation of gross pathological lesions including splenomegaly. The data was represented as log CFU/mL± the SD from at least three independent experiments performed in duplicates. Means were compared by Tukey’s multiple comparison test. Significant difference (*p < .05, **p < .01, ***p < .001) compared to the wild-type control.

Host cytokine responses

The role of Lon in eliciting the host immune response is crucial as it determines the fate of invading bacteria. The in vitro studies showed that JOL 909 upregulated the expression of pro-inflammatory cytokines up to 48 hours post-infection (hpi). The levels of IL-2 and IFN-γ increased by about 80- and 200-fold on average in both non-phagocytic (HeLa and HepG2) and phagocytic (RAW) cells infected with JOL 909. Levels of the potent anti-inflammatory cytokine IL-10 were within the same range (up to 12-fold change) in cells infected with JOL 401 and JOL 909 (Figure 9). The levels of IL-4 increased in cells infected with JOL 909, which activates the adaptive immune response proving advantage to the host.

Figure 9. The kinetics of cytokine gene expression in the phagocytic and non-phagocytic epithelial cells.

The IL-2, IL-4, IFN-γ, and IL-10 mRNA expression in JOL 401, JOL 909, and JOL 909::lon infected HeLa, HepG2, and RAW cells were evaluated at various time points post-infection. The data was represented as relative mean fold change ± the SD from at least three independent experiments performed in duplicates. The experiment was performed using GAPDH as the endogenous control. IL- Interleukin; IFN- Interferon. Means were compared by Tukey’s multiple comparison test. Similar lowercase letters indicate the non-significant difference of means (p > .05). Significant difference (*p < .05, **p < .01, ***p < .001) compared to the wild-type control.

Discussion

The present study investigates the involvement of the Lon protease in Salmonella virulence and intracellular survival with respect to the ability to establish systemic infection. To dissect the details of Lon-mediated Salmonella virulence, we exposed a lon mutant JOL 909, wild type JOL 401, and lon complemented Salmonella JOL 909::lon to a range of abiotic stress conditions including temperature, osmolarity, oxidative and pH-induced stress and quantitatively assessed the expression of virulence and nutrient uptake genes, their interaction with host cells, and fate during an in vivo infection.¹ Survival during stress requires proficient manipulation of a myriad of genes; thus, the effect of Lon may be better elucidated under stress-induced virulence studies. Growth analysis experiments revealed that deletion of the lon gene was not detrimental to host cells and did not cause significant growth rate alteration as compared to the wild type strain. Therefore, the mutant could be utilized in subsequent studies comparable to the wild type strain. During the infection process, Salmonella will make contact with the host cell by attaching to the host cell through their receptors, and the level of adhesion and invasion are the very first virulence determinants for successful infection.^25,26 The adhesion and invasion studies conducted on non-phagocytic HeLa, HepG2 and Caco-2 cells revealed the lon deletion significantly increased (approximately 2-fold increase) both adhesion and invasion into the host cell. This was confirmed by CFU counting and by fluorescence microscopy. When we examined persistence of the various Salmonella strains within host cells, it was evident that the lon mutant was significantly defective in intracellular survival compared to the wild type strain. Such a dramatic increase in virulence yet defective intracellular fitness in lon mutant hints at a link between Lon protease and SPI-1 genes that are important for the early stage of infection,^27,28 but not with SPI-2 genes that are primarily responsible for Salmonella survival in the host cell.^29,30 The SPI-1 gene region encodes elements of Type III secretory system (TTSS) machinery that bacteria use to inject effector proteins at early stages of infection that cause actin rearrangement in the host cytoskeleton facilitating Salmonella entry into the host cell.^31,32 This phenomenon is demarcated by membrane ruffling and the formation of protrusions called lamellipodia (Trigger mechanism). In certain cells like the HepG2, actin condensation is less profound, and the process is termed the Zipper mechanism. A similar observation was evident in Caco-2 cells wherein distinct membrane ruffling was seen without the formation of lamellipodia. These studies conducted on HeLa, HepG2, and Caco-2 cells for entry mechanism did not show significant alteration due to lon deletion. This suggests, even though Lon causes enhancement of Salmonella virulence, Lon may not influence the entry mechanism into the host cell.

Under normal physiological conditions, cells produce ROS that is harmful to cellular components.^33–35 In healthy cells, the levels of ROS are maintained at harmless levels by enzymes such as superoxide dismutase, peroxidase, and catalase that convert reactive oxygen into H₂O₂ or water as the terminal product. However, in the presence of iron ions, H₂O₂ can revert to form OH· free radicals that are again harmful to cellular components. On the other hand, mammalian phagocytic cells use H₂O₂ as a defense tool against invading pathogens.^35,36 The results of the present study suggest that the lon mutant extensively produced OH· free radicals relative to the wild type counterpart, demonstrating that the mutant lost the ability to mitigate oxidative stress. It also strengthens the notion that the lon mutant may enhance internal and host-derived oxidative stress due to an enhanced virulence phenotype. One strategy bacteria utilize to avoid the accumulation of H₂O₂ is active secretion via outer membrane proteins.³⁷ The expression of genes encoding outer membrane proteins was elevated in the lon mutant as we measured four representative outer membrane protein-encoding genes ompF, ompD, fepA, and csgB.^38–40 The encoded proteins are important for nutrient and ion uptake. Interestingly, the lon mutant resulted in a sharp increase in csgB at cold shock (4°C) that encodes for a minor curling subunit in Salmonella that confirms Lon protease function as a cold-shock protein similarly to what has been demonstrated in E. coli in previous studies.^41,42

Herein, a link between the Lon protease and iron uptake regulation can be proposed due to the hyperexpression of the fepA gene in the lon mutant. To investigate the connection between lon and iron uptake regulation, we evaluated the level of expression of iron homeostasis related genes fur (ferric uptake regulator), sitA (Iron/manganese ABS transporter substrate-binding protein), and dps (DNA protection during starvation protein) genes. Expression kinetics demonstrated significant upregulation of these genes in the lon mutant ST, suggesting that the mutation may result in an influx of iron ions. These observations corroborate the previously described phenotype of increased OH· generation by the lon mutant (Fenton reaction). This suggests lon is involved in iron-dependent oxidative stress response pathways in Salmonella. Interestingly, a fepA deletion in the lon mutant ST (JOL2472) and the ST wild type strain (JOL2473) completely abolished the expression of fur, sitA, and dps in both strains, possibly due to “lon-mediated” and “fepA-mediated” iron uptake is controlled by two independent pathways or these genes may activate upon a certain threshold of iron in the cytoplasm that is never reached in the absence of fepA. In addition to iron, the magnesium ion transporters^43,44 mgtA, mgtB, and corA also demonstrated an increasing trend in the lon mutant. Herein, corA expression is particularly important due to its involvement in iron transport along with magnesium. Therefore, it can be concluded that the lon mutant naturally experiences an increased level of oxidative stress that is generated by its cellular machinery.

When examining the virulence gene expression profile of the lon mutant and the wild type strain, it is possible to propose that Lon protease activity is linked to SPI-1 rather than SPI-2,⁴⁵ as indicated by a sharp increase in the SPI-1 candidate invF and hilC genes but not the SPI-2 sifA and sseJ candidate genes. This allows us to hypothesize that Lon activity is essential during the early phases of infection to maintain the virulence phenotype of Salmonella until the cells have entered and established infection in the host. This was also evident in the lon mutant whereby it lost virulence regulation, which is a key failure for a successful intracellular pathogen. In turn, such an enhanced virulence phenotype may elicit enhanced cytokine response from the host perspective, which was confirmed by assaying cytokine profiles in phagocytic (RAW 264.7) and non-phagocytic (HeLa and HepG2) cells.^46,47 Results confirmed that the lon mutant is hyper immunogenic relative to the wild type strain, possibly due to the aforementioned overexpressed virulence genes. However, according to our hypothesis, the lon mutant should be unable to mitigate stress-induced damage and must be rapidly eliminated from the host cell compared to the wild type strain due to weakened virulence homeostasis. To further investigate the idea of early clearance, we exposed Salmonella mutant and wild type strains to all the tested stress conditions for a predetermined time (5 h) and verified their survival by plate counting. Besides, in vivo inoculation studies were also conducted to evaluate bacterial stress survival using a mouse model. As expected, the mutant resulted in rapidly dwindling numbers in both in vitro and in vivo conditions, suggesting that the activity of Lon protease is indispensable for Salmonella systemic infection and phagocytic survival during dissemination stages in the host.^48–50

In the present study, the expression of lon was studied under normal laboratory conditions where the Salmonella was allowed to grow aerobically. Oxygen deprivation is yet another virulence determinant of wild type Salmonella. Being a facultative anaerobe, Salmonella can effectively survive under partial or complete anaerobic conditions in the gut. Low oxygen tension in the anaerobic environment upregulates the SPI-1 genes which in turn enhances the invasiveness of the bacteria.⁵¹ On the other hand, such a reducing environment causes downregulation of lon expression. Absence of oxygen hampers the synthesis of ATP thereby decreasing the activity of Lon protease (AAA+ ATPase). Once the Salmonella maneuvers through the gut and reaches an oxidizing environment, the Lon protease activity increases. Our observation of a reduction in lon expression under anaerobic condition corroborates with the earlier studies in Escherichia coli. Lon activity is controlled by a redox switch which adjusts the size of the enzyme’s exit pore depending on the presence or absence of oxygen⁵² and thereby regulates the lon-dependent pathways in Salmonella. Additionally, the exposure to the anaerobic environment helps Salmonella to survive in the extra-intestinal oxygen-rich conditions.⁵³ Further, Salmonella virulence is influenced by multiple genes and pathways other than lon.^54,55

Summarizing the results presented here, we conclude that Lon protease is essential for Salmonella oxidative stress regulation followed by acid stress, both of which are important aspects of Salmonella survival in macrophages and the hostile gastric tract. The role of Lon protease goes beyond its cellular homeostasis and quality control role; rather it is a key element in Salmonella pathogenesis and is indispensable for systemic infection. We propose that Lon protease is a mediator of virulence genes hosted in the SPI-1 genomic region. This study adds mechanistic insight to the fate of Salmonella in the absence of Lon protease and highlights its essentiality as a virulence mediator on top of its already known function as a housekeeping protease.

Materials and methods

Bacterial strains, plasmids, and primers

The bacterial strains, plasmids, and primers used in this study are listed in Tables 1 and 2. Salmonella strains were routinely grown in Luria broth (LB) at 37°C with constant agitation or on LB agar. For all stress-induced experiments, Salmonella Typhimurium (ST) grown to log phase was utilized unless otherwise indicated.

Table 1. List of bacterial strains and plasmids used in this study.

Bacteria/Plasmid	Genotypic characteristics	Reference
S.Typhimurium
JOL 401	Salmonella Typhimurium wild type, SPI-1 invAE^+ hilA^+ avr^+; SPI-2, amino acid permease; SPI-3, mgtC^+; SPI4, ABC transporter; SPI5, pipB^+	Lab stock
JOL 909	JOL 401 ∆lon	Lab stock
JOL 909::lon	JOL 909 carrying pWSK29+ lon gene	This study
JOL 2472	JOL 401 ∆fepA:: cat	This study
JOL 2473	JOL 909 ∆fepA:: cat	This study
E. coli
DH5α	E.coli F^−Φ80dlacZΔM15Δ (lacZYA-argF)U169 recA1 endA1 hsdR17(rk^−, mk^+) phoA supE44 thi-1 gyr A96 relA1λ^−	Invitrogen
Plasmid
pWSK29	Low copy cloning vector, Ap^R	^Citation42
pKD46	oriR101-repA101ts; encodes lambda red genes (exo, bet, gam); native terminator (tL3); arabinose-inducible promoter for expression (ParaB); bla	^Citation57
pKD3	oriR6Kgamma, bla (ampR), rgnB (Ter), catR, FRT	^Citation57
pHLJ 65:: GFP	asd+, pBR ori,-lactamase signal sequence-based periplasmic secretion plasmid, 6 His tag containing GFP	Lab stock

Table 2. List of primers used in this study.

Gene	Sequence (5ʹ-3ʹ)	Reference
cat^R	F:ACTTCCCGTTGCCAGCGCCTCTTTCAT TATAACCCTGTGTTTATTATGAAGTGTAGG CTGGAGCTGCTTC R:AGCATCGTTTTTGCCAATTCCCTCCC CGAATGAGGGAGGGAAGGTTGCCAATG GGAATTAGCCATGGTCC	This study
FepA-F^1	F:ACTTCCCGTTGCCAGCGCCTCTTTC R:AGCATCGTTTTTGCCAATTCCCTCC	This study
FepA-I^2	F:GACGGTAAACCCGTCACCAG R:ACTCCCGGTACGTTCAGACT	This study
invF	F:GCAGCAAATTATTACGCCTTC R: AGTCTTCTCCCAGCATTCTC	This study
hilC	F:GCAATAATCACCTTCAACAGCC R:GCCCTTCTATTTCGCTCATAC	This study
sifA	F:ATTACCACCACCGCATACC R:GCTTCGACTATCTCAGCTAAC	This study
sseJ	F:TCAATAAATCACATCCCAAGCC R:CCCTTCAGCAAAATTAAGCATC	This study
ompD	F:GCAACCGTACTGAAAGCCAGGG R:GCCAAAGAAGTCAGTGTTACGGT	^Citation56
ompF	F:CGTGCTGGCGGTTTGTTGAC R:TTGCTGTACGCTGCGGTGAC	^Citation56
fepA	F:TTTTGCTCCGCTTCAGTC R: GTTACGCCATTATTCCAGCC	This study
csgB	F:AGGAAATAATCGGGCGAAAG R:GCGAATAGCCATATGCGAC	This study
sitA	F:AAAGCACCATATTCCCGCC R:AATCGACATACAGCACGCC	This study
dps	F:ACCACGCAAGTTATCAACAG R:CGAACATCATTAGCGACAACAG	This study
Fur regulon	F:CAGGAACCAGATAACCATCAC R:CCGCAATCAAGGCAGATAAG	This study
corA	F:GCATATCCACTCCTTCTTCTTC R:TTTCATCCGCCAACTGTTC	This study
mgtB	F:ACTGCTTTTTGTCGCCGCC R:TACCGTTGTTGCCGTGTTCC	This study
mgtA	F:AACAAATTCAAAAGCCGCAG R:CAGTAATCCCACCACAAACC	This study
lon	F:CGCTGAAACTGGCGGATAAA R:TAGTACTCGCGCTGGGATTT	This study
Lon-C^3	F:attGAATTCTTGACCATTACGAAACTTGC R:tacAAGCTTGTTCGGAATAAAAGCGTT	This study
rrsG	F:GTTACCCGCAGAAGAAGCAC R:CACATCCGACTTGACAGACC	^Citation56
IL-2	F:TTCAAGCCCCACTTCAAGCC R:TGAGTCAAATCCAGAACATGCC	This study
IL-4	F:ACGGATGCGACAAAAATCAC R:ACCTTGGAAGCCCTACAGAC	This study
IFN-γ	F:AGACAATGAACGCTACACAC R:TCTTTTCTTCCACATCTATGCC	This study
IL-10	F:GACAACATACTGCTAACCGAC R:ATCACTCTTCACCTGCTCC	This study
GAPDH	F:AGAGGAGAAAGTGGGGAAAAG R:AAATCCGTTCACACCGACC	This study

¹-primers for the flanking region of fepA;²- primers for the internal gene region of fepA;³-primers used to amplify lon for complementation experiment

Construction of a fepA Salmonella mutant and a lon complemented strain

The mutant strain was constructed as previously described.⁵⁷ Briefly, prior to the target gene deletion from the host strain, JOL 401 and JOL 909 were electroporated with the helper plasmid pKD46, which provides the inducible red lambda components required for recombination. The target gene fepA was replaced with the cat^R gene contained on a linear PCR product amplified from the pkD3 plasmid. Recombinant clones were selected by plating on LB agar containing 25 µg/mL chloramphenicol. Successful fepA gene deletion was confirmed using flanking (fepA-F) and inner (fepA-I) PCR primers (Table 2). The lon gene was amplified for complementation experiments using the Lon-C primers (Table 2). The general PCR steps used for amplification consisted of a 3-min denaturation step at 95°C, 30 cycles of denaturation for 45 s at 90°C, annealing for 45 s at 55°C, an extension for the 60 s at 73°C, followed by a final elongation of 10 min at 73°C. The PCR product was cloned into the pWSK29 low copy number plasmid using EcoRI and HindIII as restriction enzymes.

Cell culture

The human epithelial cancer cell lines HeLa, HepG2, and Caco-2 were procured from the American Type Culture Collection (Manassas, VA, USA) and were maintained in Dulbecco’s modified Eagle medium (DMEM) (Thermo Fisher Scientific, MA, USA) containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin, and L-glutamine (0.3 mg/mL). When the cells reached 60–80% confluence, cells were harvested and seeded in 96-well tissue culture plates at a rate of 2.5x10⁵/mL in 10% DMEM. Plates were incubated at 37°C in a 5% CO₂ atmosphere.

Growth curve analysis

Growth of the ST wild type strain (JOL 401), ST lon mutant (JOL 909), and ST lon complemented strain (JOL 909::lon) was analyzed by sub-culturing overnight grown cultures at a 1:100 ratio in LB broth. Cultures were incubated at 37°C on a shaking platform (225 rpm). Aliquots of 1 mL were collected every hour for a period of 12 h and OD₆₀₀ was determined.

Bacterial growth in stress conditions

Log phase bacterial cultures were subjected to various stress conditions with slight modifications.⁵⁸ Briefly, bacterial cells were grown to mid-log phase (0.4 OD₆₀₀) and cells were collected by centrifugation at 3,500 x g for 15 min. Bacterial cells were re-suspended in pre-warmed LB medium at pH 3.0 (adjusted with hydrochloric acid) for acid stress, into pre-chilled LB medium at 4°C for cold stress, into pre-warmed LB with 5 mM H₂O₂ for oxidative stress, and LB with 5% NaCl (wt/v) for osmotic stress. Bacteria were exposed to respective stress conditions for 5 h.

Adhesion and invasion assays

Adhesion and invasion assays were conducted using HeLa, HepG2, and Caco-2 cell lines according to a previously described procedure.⁵⁹ Confluent cell monolayers were infected with Salmonella strains JOL401, JOL909, and the JOL909::lon complemented strain at 20 multiplicity of infection (MOI) in 24-well tissue culture plates. After 45 min of incubation, cells were washed three times with PBS, and adhered bacterial cells were enumerated by lysis of monolayers with 0.1% Triton X-100 for 10 min. For invasion assays, bacterial interaction with cell monolayers was conducted for 2 h and subsequently treated with gentamycin (100 μg/mL) for 1.5 h to eliminate any extracellular bacteria. Then, the cells were lysed with Triton X-100 for 10 min. Invaded cells were enumerated by plating on LB agar using decimal dilutions. The data from three independent experiments are presented as CFU/mL with standard deviation.

Intracellular survival in phagocytic cells

To assess the intracellular survival of the lon mutant in phagocytic cells, the Salmonella strains were labeled using a CFSE Cell Division Tracking Kit (BioLegend, CA, USA). Then, the labeled Salmonella strains were injected into the mouse macrophage cell line RAW and incubated at 10⁵ cells/mL for 24 h in 24-well plates. Cells were observed under fluorescent microscopy (Leica, Wetzlar, Germany) at 400 x magnification. Further, actual CFU counts were taken at 1, 3, 24 and 48 h post-inoculation to confirm the reduction in cell numbers in phagocytic cells.⁴⁵

Phalloidin staining of actin

HeLa, HepG2 and Caco-2 cells growing on a coverslip were infected with stress exposed JOL 401, JOL 909, or JOL 909::lon cells at 20 MOI. Uninfected cells at t = 0 were considered as the control. Cells were incubated for 20 min and washed three times with PBS. Slides containing infected cells were fixed with 4% paraformaldehyde for 30 min at room temperature and washed three times with 1X PBS. Cells were permeabilized with 0.01% Triton X-100 in PBS for 10 min at room temperature. After washing, cells were blocked with 5% bovine serum albumin. To stain actin filaments, 100 μL of 1X phalloidin (Abnova, Taiwan) was prepared in 1% BSA and incubated with cells for 1 h in the dark. Nuclear staining was carried out using DAPI (Sigma, CA, USA) and observed under fluorescence microscopy (Leica). The extent of actin polymerization by each ST strain was compared in three independent experiments.

Hydroxyl radical (OH·) measurement

Production of OH· was measured following the manufacturer’s protocol (Abcam, Cambridge, UK). After exposure to stress treatments, Salmonella strains were plated in 96-well plates at a concentration of 1 × 10⁴ cells/90 µL per well. Plates were centrifuged at 800 rpm for 2 min. The supernatant was removed, and wells were supplemented with 100 µL/well OH580 Stain Working Solution and incubated at 37ºC for 1 h. Then, cells were washed three times with DPBS followed by the addition of 100 µl assay buffer to each well. Fluorescence was recorded at Ex/Em = 540/590 nm. As a positive control, HeLa and HepG2 cells treated with 10 µM CuCl₂ and 100 µM H₂O₂ (Fenton reaction) at 37ºC for 1 h. Cells without any treatment were considered as a negative control.

In vitro bacterial survival

To evaluate the ability of bacterial cells to withstand stress treatment, cells were collected by vacuum filtration using a 47-mm filter membrane with 0.45-µm pores (Thermo Fisher Scientific) after 5 h. Filtered cells were resuspended in pre-warmed LB broth by vortexing. The viable bacterial counts were performed through serial dilutions and plating on BGA agar plates.

In vitro bacterial virulence gene expression

The expression of virulence genes in Salmonella strains after stress treatment was investigated at the mRNA level at 5 hpi under each stress condition. The total RNA was isolated, and cDNA was synthesized (Toyobo, Osaka, Japan). Virulence gene expression was quantitatively analyzed using qRT-PCR (Table 2). The PCR conditions were 95°C for 5 min, 40 cycles at 95°C for 30 s, 56°C for 30 s, 72°C for 1 min, and final annealing and extension of 55°C for 5 s and 95°C for 30 s. The expression results were normalized against the housekeeping gene rrsG of Salmonella and data are presented in the 2^−ΔΔCT method.⁶⁰

In vitro cytokine gene expression

The cytokine gene expression in HeLa, HepG2, and RAW cells in response to bacterial infection (oxidative stress treated) were carried out for IL-2, IFN- γ, IL-4, and IL-10 (Table 2). Herein, infected HeLa and HepG2 cells were cultured for 12, 24, and 48 h. The cells were harvested at each time point and the total RNA was isolated and cDNA was synthesized as described above. The expression of cytokine genes was normalized against GAPDH and non-infected control. The changes in the relative expression of cytokine genes were determined using the 2^−ΔΔCT method.⁶⁰

Survival and growth in epithelial cells

HeLa and HepG2 cells were grown in DMEM complete medium. A total of 4 × 10⁵ cells were seeded per well in 24-well plates. When the cells reached 80% confluence, monolayers were infected with JOL401, JOL909, and JOL909::lon at 10 MOI for 2 h. After incubation, cells were washed three times with PBS and the extracellular bacteria were removed by gentamycin treatment (100 μg/mL) for 2 h. Then, cells were washed three times with PBS and replenished with complete medium. The bacteria-induced cytotoxicity was assessed using IncuCyte (Thermo Fisher Scientific) live-cell imaging system using Cytotox Green reagent at 2 ng/mL.

Assessment of virulence in mice

All animal experimental procedures were approved (CBNU2015-00085) by the Jeonbuk National University Animal Ethics Committee in accordance with the guidelines of the Korean Council on Animal Care and Korean Animal Protection Law, 2007; Article 13 (experiments with animals). Specific pathogen-free, female BALB/c inbred mice (4–5 weeks average age) were purchased from Koatech (Pyeongtaek, Gyeonggi-do, Korea). The animals were provided with water and antibiotic-free food ad libitum. Animals were monitored twice daily for behavioral and physiological signs.

To assess in vivo virulence of Salmonella strains, mice (n = 10) were orally inoculated at 1 × 10⁹ CFU/mice/100 μl of PBS with JOL401, JOL909, or JOL909::lon. Mice were monitored for Salmonella induced clinical signs including ruffled fur, weight loss, reduced feed intake, and mortality. Mice were sacrificed on days 1, 6, 9, and 12 post-infection, and enlargement of the spleen was recorded. Further, the level of bacterial colonization in the spleen was assessed by plating tissue homogenate (1 g) on BGA agar after serial dilutions.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism 7 (GraphPad Software, CA, USA) and IBM SPSS software. Student’s t-test and analysis of variance (ANOVA) followed by Turkey’s multiple comparison test was used to compare means among the treatment groups and to compute the p-value. The difference was considered significant if p was ≤0.05.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Data availability statement

The fluorescent microscopy imaging conducted for Salmonella mediated actin cytoskeleton rearrangement compiled and has been deposited in the Mendeley Data repository and can be accessed using the following link. Lee, John Hwa (2020), “Salmonella mediated actin condensation upon entry into HeLa and HepG2 cells”, Mendeley Data, V1, doi: 10.17632/czng5n4fdk.1.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry (IPET) through Animal Disease Management Technology Development Program, funded by Ministry of Agriculture, Food and Rural Affairs (MAFRA) [grant number. 119058-02].