DnaK-Mediated Protein Deamidation: a Potential Mechanism for Virulence and Stress Adaptation in Cronobacter sakazakii

ABSTRACT

Cronobacter sakazakii is a Gram-negative bacterium that causes infections in individuals of all ages, with neonates being the most vulnerable group. The objective of this study was to explore the function of the dnaK gene in C. sakazakii and to elucidate the impact of alterations in the protein composition regulated by dnaK on virulence and stress adaptation. Our research demonstrates the critical role of the dnaK gene in various key virulence factors, including adhesion, invasion, and acid resistance in C. sakazakii. Through the use of proteomic analysis, we discovered that deletion of the dnaK gene in C. sakazakii leads to an upregulation of protein abundance and increased levels of deamidated posttranscriptional modifications, suggesting that DnaK may play a role in maintaining proper protein activity by reducing protein deamidation in bacteria. These findings indicate that DnaK-mediated protein deamidation may be a novel mechanism for virulence and stress adaptation in C. sakazakii. These findings suggest that targeting DnaK could be a promising strategy for developing drugs to treat C. sakazakii infections.

IMPORTANCE Cronobacter sakazakii can cause disease in individuals of all ages, with infections in premature infants being particularly deadly and resulting in bacterial meningitis and sepsis with a high mortality rate. Our study demonstrates that dnaK in Cronobacter sakazakii plays a critical role in virulence, adhesion, invasion, and acid resistance. Using proteomic analysis to compare protein changes in response to dnaK knockout, we found that dnaK knockout significantly upregulates the abundance of some proteins but also results in the deamidation of many proteins. Our research has identified a connection between molecular chaperones and protein deamidation, which suggests a potential future drug development strategy of targeting DnaK as a drug target.

INTRODUCTION

Cronobacter sakazakii, formerly known as Enterobacter sakazakii prior to 2007, is an opportunistic pathogen that typically enters the body through food transmission (1, 2). Its remarkable ability to withstand dry conditions allows it to survive and persist in dehydrated foods with low moisture content (3). This pathogen has been isolated from various food sources, including infant formula, baby food, and cereal products (4). C. sakazakii infection affects not only infants under 6 months of age but also older individuals, although the severity of infection is typically lower. However, in premature infants and those under 6 months old, infections caused by C. sakazakii can be highly fatal, leading to bacterial meningitis and sepsis with mortality rates ranging from 40% to 80% (5, 6). Epidemiological investigations suggest that infant infections with C. sakazakii are often related to contamination of infant formula during factory production or preparation of bottles (7). As a result, many countries have adopted a zero-tolerance policy for C. sakazakii in infant formula powder (8).

The pathogen possesses a range of virulence factors that contribute to tissue adhesion, invasion, and host cell damage (9). Outer membrane protein A (OmpA) has been identified as a potential pathogenic marker for C. sakazakii (10). Virulence-related plasmids and their associated genes, such as those responsible for iron acquisition and hemagglutinin, have been identified in different strains of C. sakazakii (11). Additionally, environmental tolerance-related growth factors, including NlpD and Yrt2a, have been found to contribute to the acid tolerance and pathogenic potential of this pathogen (12, 13). The ptsH gene regulates carbon metabolism, stress response, and virulence and is related to oxidative and osmotic stress in C. sakazakii (14). Finally, the PmrA/PmrB two-component system can modify lipopolysaccharide (LPS) structure to change cell membrane permeability and hydrophobicity, enhance bacterial tolerance to cationic antimicrobial peptides, and play a critical role in responding to macrophage killing (15). Additionally, the traversal of human brain microvascular endothelial cells (HBMECs) by C. sakazakii via transcytosis plays a critical role in the development of neonatal meningitis (16). Previous investigations have highlighted several crucial mechanisms employed by C. sakazakii within HBMECs, contributing to the pathogen’s invasion of host tissues and the subsequent onset of infection (17, 18). Primarily, the bacterium triggers intracellular bacterial endocytosis, enabling its internalization into HBMECs. Additionally, C. sakazakii stimulates inflammation and apoptosis, of which both likely contribute to the pathogenesis of neonatal meningitis. Lastly, C. sakazakii disrupts the tight junctions of the cell monolayer, further facilitating bacterial translocation.

DnaK is a ubiquitous molecular chaperone protein in bacteria, which has been implicated in conferring bacterial acid tolerance, desiccation tolerance, and pathogenicity (19). However, the precise role of DnaK in the environmental resistance and virulence of C. sakazakii remains elusive. In this study, we aim to examine the effects of dnaK knockout on both environmental tolerance and virulence and to investigate how dnaK affects intracellular proteins via proteomic and posttranslational modification analyses.

RESULTS

Construction and growth characteristics of the dnaK mutant in C. sakazakii.

To assess the role of the dnaK gene in the environmental tolerance and pathogenesis of C. sakazakii BAA-894, the generation of a dnaK knockout mutant was carried out using the pCVD442 suicide plasmid technique (Fig. 1A). The validity of the dnaK gene loss was confirmed through sequencing. The absence of any polar effect on the downstream dnaJ gene was verified through transcription-quantitative PCR (qRT-PCR) analysis (results not shown). The growth rate of the mutant was observed to be slightly slower than that of the wild-type strain when both were cultivated in Luria-Bertani (LB) medium at 37°C. The growth deficit of the mutant was restored by the introduction of a dnaK complementing plasmid (Fig. 1B). These results suggest that the removal of the dnaK gene in C. sakazakii leads to a reduction in the growth rate.

FIG 1.

Construction and growth characteristics of the dnaK mutant in C. sakazakii. (A) Construction of a pCVD442 suicide plasmid for dnaK deletion in C. sakazakii. (B) The growth of wild-type (WT) and ΔdnaK C. sakazakii strains, as well as the complementation of dnaK in the ΔdnaK strain (ΔdnaK-C), was monitored in LB medium at pH 7.0. The optical density at 600 nm (OD₆₀₀) was used to measure the change in bacterial optical density over time, while CFUs were used to determine the viable cell count at different time points. Each data point represents the average value and standard deviation derived from three independent biological replicates.

dnaK knockout reduces bacterial virulence.

Rats were infected with wild-type (WT), ΔdnaK, and ΔdnaK-complemented strains of bacteria, and their survival was periodically monitored. A significant difference was observed in the survival curves between rats infected with ΔdnaK bacteria and those infected with the WT strain, with higher survival rates being observed in the group infected with ΔdnaK bacteria at 5 days postinfection (Fig. 2A). The impact of dnaK on the invasion of C. sakazakii was also investigated in rats. The bacterial load in the blood, liver, and spleen of rats infected with bacteria for 24 h was determined by homogenizing, spreading on plates, and counting. ΔdnaK exhibited a significantly lower bacterial load in the blood, liver, and spleen than the wild-type strain (Fig. 2B, C, and D). The defect of the ΔdnaK strain was partially complemented by the presence of the dnaK plasmid.

FIG 2.

Impact of dnaK knockout on bacterial virulence. (A) Survival curves of rat pups after oral infection with wild-type, dnaK knockout, and complemented strains. Combined data from three independent experiments are presented (*, P < 0.05; ***, P < 0.001, using log-rank test for comparison between the survival curves of wild-type and dnaK knockout strains). (B to D) Quantification of CFUs in the blood, liver, and spleen of infected rats at 24 h postinfection, determined by organ homogenizing and plating.

dnaK affects acid resistance, adhesion, and invasion.

Acid tolerance is crucial for the survival of various foodborne pathogens in the stomach and macrophages. The role of dnaK in bacterial acid tolerance was tested, and the results showed that the dnaK knockout strain exhibited significant inactivation in pH 3.0 to pH 4.0, indicating the crucial role of dnaK in the acid tolerance of C. sakazakii (Fig. 3A). To further investigate the role of the dnaK gene in the virulence of C. sakazakii BAA-894, an adhesion and invasion assay was conducted using HCT-8 cells. The results revealed that the adhesion (Fig. 3B) and invasion (Fig. 3C) rates of the ΔdnaK mutant were significantly lower than those of the wild-type strain. The defect in adhesion and invasion was significantly restored upon complementation of the mutant.

FIG 3.

Impact of dnaK on acid sensitivity, cell adhesion, and invasion of C. sakazakii. (A) Growth curves of the WT, ΔdnaK mutant strain, and complemented bacteria in LB medium at different pH values were measured. (B) The relative adhesion of HCT-8 cells to the WT, ΔdnaK mutant strain, and complemented strain was assessed. (C) The relative invasion of HCT-8 cells by the WT, ΔnlpD mutant strain, and complemented strain was measured. In the adhesion and invasion experiments, Salmonella enterica was used as a positive control and Escherichia coli DH5α was used as a negative control. All experiment was repeated 3 times; *, indicates significant differences.

dnaK does not affect biofilm biomass and desiccation resistance.

The formation of biofilms is crucial for the environmental tolerance of bacteria, and to examine the contribution of the dnaK gene to the biofilm biomass of C. sakazakii ATCC BAA-894, a crystal violet (CV) staining assay was conducted. The results showed that the ΔdnaK mutant and the complemented strain did not exhibit a significant alteration in biofilm biomass compared with the wild-type (WT) strain (Fig. 4A). Additionally, the ability of C. sakazakii to survive in powdered formula is dependent on its tolerance to drying. However, experiments indicated that the dnaK gene has no significant impact on the desiccation tolerance of C. sakazakii (Fig. 4B and C). These results suggest that there is no association between the dnaK gene and biofilm biomass and desiccation resistance in C. sakazakii.

FIG 4.

Lack of an effect of dnaK on biofilm biomass and desiccation resistance. (A) Biofilm formation was assessed using a crystal violet staining assay. The biofilm was stained with 1% crystal violet, and the absorbance was measured at 570 nm. The assays were biologically repeated three times. (B) Changes in the number of surviving bacteria after 7, 14, and 21 days of treatment in a dry environment were determined by serially diluting and plating the bacteria. The assays were biologically repeated three times. ns, no significant difference. (C) Changes in the number of surviving bacteria after 7, 14, and 21 days of treatment in infant formula powder were determined by serial dilution and plating of the bacteria. The assays were biologically repeated three times.

GO and KEGG pathway analysis.

Proteomic analysis revealed that the absence of DnaK resulted in the significant upregulation of 147 proteins and significant downregulation of 8 proteins (Fig. 5A). The downregulated proteins did not exhibit any significant enrichment in Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). In contrast, the GO analysis showed significant enrichment of the upregulated proteins in biological processes related to cell division, protein folding, and the peptidoglycan biosynthetic process; significant enrichment in the cellular component of cytoplasm; and significant enrichment in molecular functions, such as ATPase activity, manganese ion binding, and ATP binding (Fig. 5B). KEGG analysis demonstrated that DnaK knockout significantly upregulated metabolic pathways, biosynthesis of amino acids, biosynthesis of secondary metabolites, biosynthesis of antibiotics, lysine biosynthesis, and arginine biosynthesis (Fig. 5C). These results suggest that C. sakazakii may respond to the protein stress caused by dnaK knockout through upregulation of protein expression.

FIG 5.

Protein profiling changes caused by dnaK gene knockout in the proteomic analysis. (A) Volcano plot showing the impact of dnaK knockout on protein abundance. The criteria for significant changes are defined as a fold change greater than 2 and a P value of less than 0.05. Red dots represent significantly upregulated proteins, green dots represent significantly downregulated proteins, and gray dots represent proteins without significant differences. The number of proteins in each category is labeled in the figure. See also Table S1 in the supplemental material. (B) Gene Ontology (GO) analysis of the significantly upregulated proteins caused by dnaK knockout in the biological process (BP), cellular component (CC), and molecular function (MF) categories. The size of the shape represents the number of proteins classified in that category, while the color represents the ratio of the number of proteins classified in that category to the total number of proteins in that category. See also Table S2 in the supplemental material. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the significantly upregulated proteins caused by dnaK knockout in metabolic pathways. No significant enrichment was observed in the KEGG analysis for the downregulated proteins. The size of the shape represents the number of proteins classified in that category, while the color represents the ratio of the number of proteins classified in that category to the total number of proteins in that category. See also Table S2.

dnaK deletion cause protein deamidation.

In our previous study, we found that treatment with p-coumaric acid led to deamidation and growth disadvantage in C. sakazakii (20, 21), indicating that protein deamidation may have an impact on bacterial phenotypes. To investigate potential protein stress resulting from dnaK knockout, proteomics was employed to detect changes in protein modifications. We observed that 77 proteins exhibited deamidation modifications, and among them, 34 proteins (44%) showed a significant increase in abundance upon dnaK knockout (Fig. 6A). GO and KEGG analyses of the 34 proteins with significantly increased deamidation caused by dnaK knockout showed no significant enrichment in any KEGG pathways. However, GO analysis revealed significant enrichment in molecular function-related rRNA binding and cellular component-related cytoplasm (Fig. 6B), suggesting that dnaK knockout may affect ribosome function. This finding is consistent with our observation that the dnaK knockout strain exhibits slower growth than the wild-type strain.

FIG 6.

Protein deamidation changes caused by dnaK gene knockout in proteomic analysis. (A) The volcano plot illustrates the effect of dnaK knockout on the abundance of deamidated proteins. Significance was determined based on fold change >2 and a P value of <0.05. Proteins that were significantly upregulated are shown in red, while those that were significantly downregulated are shown in green, and those with no significant difference are shown in gray. The number of proteins in each category is indicated in the figure. See also Table S3 in the supplemental material. (B) Gene Ontology (GO) analysis was performed to investigate the enrichment of significantly upregulated deamidated proteins caused by dnaK knockout in biological processes (BP), cellular components (CC), and molecular functions (MF). The size of each shape represents the number of deamidated proteins in that category, while the color represents the proportion of deamidated proteins in that category relative to the total number of proteins in that category. See also Table S4 in the supplemental material.

dnaK deletion result in Chaperonin grpEL transcriptional activation and increased deamidation.

Venn diagram analysis was conducted on the proteins that undergo changes in abundance and deamidation levels caused by dnaK deletion (Fig. 7A), revealing that only one protein, GrpEL, had both changes in abundance and deamidation levels. Specifically, knocking out dnaK led to a significant increase in both the abundance of GrpEL and the deamidation level in the N(326), N(433), N(467), and N(487) position of GrpEL (Fig. 7B and C). This result is interesting because GrpEL and DnaK are both protein molecular chaperones with similar functions. Further transcriptional analysis revealed that dnaK knockout led to the transcriptional activation of grpEL (Fig. 7D), suggesting that C. sakazakii upregulates the expression of GrpEL to compensate for the lack of protein molecular chaperones caused by DnaK knockout. To further investigate the complementary functions of DnaK and GrpEL, attempts were made to delete the grpEL gene in a dnaK knockout strain. However, despite several attempts, no double gene knockout strains were obtained, suggesting that the simultaneous knockout of grpEL and dnaK was likely to be lethal. To verify the lethal effect of knocking out both grpEL and dnaK, grpEL was first introduced into the dnaK knockout strain using a temperature-sensitive plasmid and then the genomic grpEL was knocked out using a suicide plasmid. This approach successfully deleted the genomic grpEL in the dnaK mutant. The growth curve test indicated that the knockout of the single gene grpEL did not affect the bacterial growth rate at 37°C (Fig. 7E and F), but the double-knockout strain showed a significant growth disadvantage. These results suggest that the absence of DnaK results in the deamidation of GrpEL, but the bacterium upregulates the expression of grpEL to maintain bacterial growth.

FIG 7.

Impact of dnaK Deletion on GrpE. (A) Venn diagram depicting the proteins that exhibited altered abundance (blue) or deamidation modification (yellow) upon dnaK deletion. The corresponding data for each protein are presented in the diagram. See also Table S5 in the supplemental material. (B) Amino acid sequence of GrpE protein with deamidation modification sites (in red) identified by proteomics in the dnaK deletion strain. The deamidation modification sites that showed a significant increase in abundance upon dnaK deletion compared with that in the wild type are shaded in gray. (C) Abundance of GrpE protein with deamidation modification at the N(326), N(433), N(467), and N(487) sites upon dnaK deletion. *, indicates a significant difference. (D) Transcriptional levels of the grpEL gene in wild-type, dnaK deletion, and dnaK deletion complementation strains. The experiments were repeated biologically three times. (E and F) Growth curves of wild-type, dnaK deletion, grpEL deletion, and dnaK-grpEL double-deletion strains at 30°C (E) and 37°C (F). The experiments were repeated biologically three times. Ns, indicates no significant difference.

DISCUSSION

Molecular chaperones are proteins that play various roles in microbial cells, including assisting in the folding of newly synthesized proteins, aiding in peptide secretion during translation, and repairing proteins that are broken or misfolded due to stress (22, 23). Numerous studies have demonstrated that molecular chaperones can impact the growth, metabolism, pathogenicity, and virulence of bacterial pathogens (24–26). In this study, we demonstrate that the molecular chaperone protein DnaK is essential for bacterial acid tolerance, adhesion, and invasion in C. sakazakii. Additionally, we observed that deletion of dnaK weakened bacterial dissemination and lethality in mice, indicating that DnaK is a virulence factor of Cronobacter sakazakii. Other molecular chaperones, including DnaJ, GrpE, and the RNA chaperone Hfq, have also shown to play significant roles in bacterial tolerance and pathogenicity (1, 27, 28), highlighting the functional conservation and importance of molecular chaperones.

Although the crucial role of molecular chaperones in maintaining proper protein conformation has been well-established, their impact on the bacterial proteomic profile remains unclear. In this study, label-free protein mass spectrometry was employed to investigate the proteomic changes induced by dnaK knockout. The proteomic analysis revealed that the loss of DnaK triggered a widespread activation response within the bacterium. Gene Ontology (GO) analysis demonstrated a significant enrichment of upregulated proteins across 7 different categories, including biological processes (BP), cellular components (CC), and molecular functions (MF). Furthermore, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis identified significant enrichment of upregulated proteins in 6 pathways, suggesting that the increased production of these proteins may represent a compensatory response to the reduced protein activity caused by the loss of DnaK. However, despite these broad metabolic activations, the dnaK deletion strain still exhibited growth rate and acid tolerance disadvantages, indicating that the activated metabolism was insufficient to fully compensate for the deleterious effects of DnaK deletion and underscoring the critical role of DnaK. In fact, the conservation of the dnaK gene sequence across microbes, plants, and animals suggests the essentiality of DnaK in maintaining biological competitiveness (29).

In this study, we found that DnaK plays a role in the deamidation of proteins, a modification that can lead to decreased protein activity (30). DnaK knockout resulted in a significant upregulation of the deamidation level of 44% of proteins, which were significantly enriched in ATP binding function. This result is similar to our previous finding that p-coumaric acid treatment caused deamidation and growth disadvantage in C. sakazakii, suggesting that DnaK knockout leads to fitness pressure and activates protein synthesis in bacteria (20, 21). However, the absence of the molecular chaperone DnaK may result in the misfolding of nascent peptides, and degradation of misfolded proteins may be delayed, leading to the accumulation of inactive proteins in the cell (31). The accumulation of inactive proteins has been found in various organisms and has been shown to be harmful. These results suggest that DnaK may play a role in maintaining proper protein activity by reducing protein deamidation in bacteria. Our study has revealed that the deletion of the dnaK gene leads to increased expression of the molecular chaperone grpEL and increased deamidation modification at multiple sites. Moreover, the double knockout of dnaK and grpEL is lethal for bacteria, indicating that grpEL is a compensatory molecular chaperone for dnaK. In eukaryotes, it has been demonstrated that GrpE assists DnaK in releasing folded proteins, thereby enhancing the efficiency of DnaK (32). However, further investigation is required to understand the interaction between GrpEL and DnaK in prokaryotes.

In conclusion, our study demonstrates that dnaK is essential for the growth, acid tolerance, adhesion, invasion, and pathogenicity in C. sakazakii. Moreover, we found that dnaK may enhance the activity of key proteins related to growth and pathogenicity by reducing the deamidation of proteins. These findings suggest the potential of dnaK as a target for antimicrobial therapy.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

The strains and plasmids used in this study are listed in Table 1, with the primers provided in Table 2. The bacterial strains were preserved in glycerol preservation tubes (Kirgen, China) at −80°C. Prior to experiments, the strains were revived in LB broth (Oxoid, UK) and cultured overnight. The construction of plasmids followed standard protocols with a simplified method of recombination using a commercial seamless cloning and assembly kit (Vazyme, China) replacing the traditional enzymatic cutting and circularization of DNA. The selection of transformants was done using 100 μg/mL ampicillin (Sangon, China) or 50 μg/mL kanamycin (Sangon). The generation of in-frame deletion mutants of targeted genes was achieved using the pCVD442 suicide vector (Miaolingbio, China), which was amplified using Escherichia coli S17 lambda pir (Weidi, China). In-frame deletion mutants were generated using the pCVD442 suicide vector method described previously (33). Briefly, plasmid construction was performed according to standard protocols with the minor modification that cloning the PCR fragment into the linearized vector was accomplished using a commercial seamless cloning and assembly kit (Vazyme). pCVD442 was linearized by PCR using the primer pair pCVD442-fwd and pCVD442-rev, and the upstream and downstream fragments of genes to be knocked out were amplified by PCR using the primers listed in Table 2. The pACYC184 vector was used for complementation of deleted strains. The transformation and selection of C. sakazakii were performed using the method described previously (12).

TABLE 1.

Bacterial strains and plasmids used in this study

Strain and plasmid	Description	Reference or source
Cronobacter sakazakii
WT	Wild-type Cronobacter sakazakii BAA-894	14
ΔdnaK	Markerless deletion mutant ΔdnaK	This study
ΔdnaK-C	dnaK complementation in ΔdnaK	This study
ΔgrpEL	Markerless deletion mutant ΔgrpEL	This study
ΔgrpEL-C(T)	grpEL temp sensitive complementation in ΔgrpEL	This study
ΔdnaK-ΔgrpEL-C(T)	dnaK knockout in ΔgrpEL-C(T)	This study
Escherichia coli
S17 lambda pir	Strain for construction harbouring lambda pir	20
S17 lambda pir-ΔdnaK	S17 lambda pir harbouring pCVD442-ΔdnaK	This study
S17 lambda pir-ΔgrpEL	S17 lambda pir harbouring pCVD442-ΔgrpEL	This study
DH5α	Strain for construction	39
DH5α-grpEL-C(T)	DH5α harbouring pKC1139-grpEL	This study
Plasmids
pACYC184	Low-copy plasmid	40
pACYC184-dnaK	dnaK complementation vector	This study
pCVD442	Suicide plasmid for markerless deletion	36
pCVD442-ΔdnaK	dnaK deletion plasmid	This study
pKC1139	Temp sensitive carrier	41
pKC1139-grpEL	grpEL temp sensitive complementation vector	This study

TABLE 2.

Primers used in this study

Primer name by use	Sequence (5′–3′)
For construction
pCVD442-fwd	GGCTGTCAGACCAAGTTTACTCATATATACTTTAGATTG
pCVD442-rev	GCAGATACTCTTCCTTTTTCAATATTATTGAAGCATTTATCAGGGTTATTG
ΔdnaK-A	GAAAAAGGAAGAGTATCTGCGGTACGGTACGGCCATACTGTTCG
ΔdnaK-B	GCGATTAACCCATCTAAACGTCTCCACTAAAAAATCGTCATC
ΔdnaK-C	CGTTTAGATGGGTTAATCGCCCTGATGCAGGGTAGATAAC
ΔdnaK-D	GATTAATTGTCAAGGCTAGCGTGCTGTTTCACCTGCACCTGAAC
dnaK-comp-fwd	GGTCTAGAGCCGTCTATGATCCCGGAGG
dnaK-comp-rev	GGTCTAGATACCCTGCATCAGGGCGATTA
ΔgrpEL-A	GAAAAAGGAAGAGTATACCTGATGCTGATGCAGCAG
ΔgrpEL-B	TACGTTCTTTAGCTGCCATATCGTTATTTC
ΔgrpEL-C	GATATGGCAGCTAAAGACGTAGGTATGATGTAATTTAATCTTCATAC
ΔgrpEL-D	GATTAATTGTCAAGGCTGCGTTAAAATCGTAGGGCGGGT
pACYC184-fwd	GAATTCGTAATCATGTCATAGCTGT
pACYC184-rev	GGATCCTCTAGAGTCGACCTGCA
grpEL-comp-fwd	TGCAGGTCGACTCTAGAGGATCCGAAATTGAATCTAAATCTGCTG
grpEL-comp-rev	ACAGCTATGACATGATTACGAATTCCCTGAAGTATGAAGATTAAATTCACAT
For sequencing  confirmation
ΔdnaK-E	GTGGCACGACGTCGGGTAAATC
ΔdnaK-F	GTGGCCGCCATCGCAAAGTTGATC
ΔgrpEL-E	CTTTATTCAGATGGCGCATG
ΔgrpEL-E	CCTGTAGGGCGGGTAAGCGAAGC

Construction of plasmid-based conditional complementation strains.

Plasmid-based conditional complementation strains were generated using a temperature-sensitive replicating plasmid, following the method described previously (34, 35). Specifically, in the first step, the complementation gene was ligated into a temperature-sensitive pSC101 plasmid (Miaolingbio, China), and recombinants were selected using tetracycline (Sigma, USA). In the second step, the constructed plasmid was amplified and transformed into C. sakazakii, and then transformants were selected using tetracycline. Finally, the target gene on the chromosome was deleted using the method described previously with the pCVD442 suicide plasmid (20, 36). All cultures were grown at 30°C.

RT-PCR.

TRIzol (Thermo, USA) was used to extract RNA, which was then transcribed into cDNA using a reverse transcription kit (Thermo). Quantification was performed using the Bio-Rad CFXConnect system (Bio-Rad, USA) and SYBR green mix (New England BioLabs [NEB], USA). The internal reference for all samples was the 16S RNA level.

Rat survival assay.

The rats were administered using the method described previously (12, 20). The survival status of the infected rats was regularly monitored. The bacterial load of the organs in the rats at 24 h postinfection was determined by homogenizing, dilution plating, and colony counting. Specifically, each organ was individually collected in a thick-walled centrifuge tube (Jingxin, China) containing steel beads with a diameter of 3 mm. To standardize the organ weight, the organs were rinsed with saline to remove external blood and then weighed (weight, 1 g ± 0.1 g). The centrifuge tubes containing the organs were stored at −20°C for 1 h, followed by the addition of 0.5 mL of cold (4°C) phosphate-buffered saline (PBS) and immediate transfer to a tissue homogenizer (Jingxin). Homogenization of the organs was achieved by running two sequences of 15 s each at 6,000 rpm, followed by a 10-minute storage at −20°C and an additional two sequences of 20 s each at 5,000 rpm. This process resulted in a nearly fully homogenized liquid sample. Each sample was subjected to a 10-fold serial dilution in PBS, and 50 μL of each dilution was plated onto LB agar plates. The plates were incubated at 37°C for 24 h, and the CFUs were counted.

Adhesion and invasion assay.

Bacterial adhesion of C. sakazakii to HCT-8 cells was determined by mixing bacteria with the adherent HCT-8 cell monolayer at a multiplicity of infection (MOI) of 100, incubating them together at 37°C for 45 min, and then washing away nonadherent bacteria with PBS. Adherent bacteria were lysed from the cells using 1% Triton X-100 (Abcone, China) and quantified by serial dilution and plating on LB agar. The bacterial invasion assay was similar to the adhesion assay, but with the use of a gentamicin protection assay to kill extracellular bacteria after adhesion, followed by lysis of intracellular bacteria for quantification by plating.

Desiccation resistance assays.

The strains stored at −80°C were transferred to LB medium and incubated at 37°C in a shaking incubator for 16 h. A 1/100 dilution of the culture was then inoculated into 10 mL of LB liquid culture medium and incubated until an optical density of 0.6 to 0.8 was reached to produce a bacterial suspension. A total of 200 mL of the bacterial suspension was transferred into a sterile 96-well plate and dried at 37°C in a desiccator for 7 days, 14 days, and 21 days. To record the initial cell count, 100 mL of each bacterial suspension was serially diluted in PBS and plated onto agar. To determine the survival rate after drying, the bacterial cells were suspended in PBS or infant formula powder by pipetting and scraping to remove cells from the surface of the 96-well plate. The suspended cells were serially diluted in physiological saline and plated onto agar for counting. The bacterial colony count on the plates was recorded after incubating overnight at 37°C. Each sample was repeated three times.

Biofilm biomass.

The formation of biofilm was evaluated using a crystal violet staining assay as described previously (14). C. sakazakii was inoculated into 10 mL of Luria-Bertani (LB) broth and incubated at 37°C with shaking at 200 rpm until the cell density reached 10⁷ CFU/mL. Subsequently, 100 μL of bacterial suspensions from different strains was added to individual wells in a 96-well plate and incubated for 48 h in a 37°C incubator (80% of relative humidity). The plates were then washed with sterile water, air dried and fixed with 200 μL of 99% methanol for 15 min. The biofilm was stained with 200 μL of 1% crystal violet (CV) for 30 min. The plates were washed three times with sterile water and air dried, and then 200 μL of 95% ethanol was added. The CV bound to the plate was fully dissolved by shaking, and the absorbance was measured at 570 nm using a Sunrise Basic spectrophotometer.

Protein extraction and digestion.

After overnight culture in LB medium, the cells of the wild-type and knockout strains were pelleted by centrifugation. The cells were then washed twice with PBS buffer to remove any residual medium and mixed with a protease inhibitor cocktail (Beyond, China) and PBS. The cells were lysed by exposing them to ultrasonic waves (200 W) in an ice bath until the bacterial suspension became clear. The resulting lysate was purified using a microporous filter with a pore size of 0.22 μm. The protein concentration was determined using a bicinchoninic acid (BCA) assay kit (Thermo, USA). A total of 50 μg of protein was mixed with 1 μL of 200 mM tris(2-carboxyethyl)phosphine (Aladdin, China) and 1 μL of 500 mM iodoacetamide (Aladdin) and brought to a total volume of 100 μL with 6 M guanidine hydrochloride (Aladdin). The mixture was incubated in the dark at room temperature for 40 min. The protein was then digested overnight at 37°C using 2 μg of recombinant trypsin (from Promega, USA) in a 15-kDa ultrafiltration tube. The resulting filtrate was collected by centrifugation and desalted using a desalting column (from Millipore, USA). The desalted sample was then analyzed by mass spectrometry (MS) after the addition of formic acid.

Liquid chromatography-tandem MS (LC-MS/MS) analysis.

The protein samples after trypsin digestion were analyzed by a Q Exactive Plus mass spectrometer. The mass spectra were compared to the UniProt database using Maxquant software. The analysis parameters for Maxquant included carbamidomethylation of cysteine as a fixed modification, oxidation of methionine as a variable modification, and trypsin cleavage specificity with a maximum of two missed cleavages allowed for each protein. The protein and peptide identifications were based on a maximum false discovery rate (FDR) of 1.0%, and protein quantification was normalized by the median.

Statistical analysis.

The experiments were all independently conducted three times to ensure accuracy and reliability. The results of the bioinformatics analysis of proteomics, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, were obtained from the online database DAVID (https://david.ncifcrf.gov), and the statistical analysis was performed using GraphPad Prism software (version 8.3). The significance of differences was determined using the two-tailed unpaired Student’s t test, with a P-value less than 0.05 considered statistically significant.

Ethics statement.

The study was approved by the ethics committee of our department, and written informed consent was obtained from all participants before the study.

Data availability.

All data sets generated for this study are included in the manuscript and/or the supplemental files. All of the MS proteomics data have been deposited to iProX (https://www.iprox.cn/) and can be accessed with the accession IPX0005932000 (37, 38).