The host immune responses to Staphylococcus epidermidis, a frequent cause of nosocomial infections, are not well understood. We have established a bath immersion model of this infection in zebrafish (Danio rerio) larvae. Macrophages play a primary role in the host immune response and are involved in clearance of infection in the larvae. S. epidermidis infection results in upregulation of tlr-2. There is marked inflammation characterized by heightened NF-κB signaling and elevation of several proinflammatory cytokines. There is rapid upregulation of il-1b and tnf-a transcripts, whereas an increase in il-6 levels is relatively more delayed. The IL-6 signaling pathway is further amplified by elevation of IL-6 signal transducer (il-6st) levels, which negatively correlates with miRNA dre-miR-142a-5p. Enhanced IL-6 signaling is protective to the host in this model as inhibition of the signaling pathway resulted in increased mortality upon S. epidermidis infection. Our study describes the host immune responses to S. epidermidis infection, establishes the importance of IL-6 signaling, and identifies a potential role of miR-142-5p–il-6st interaction in this infection model.

Initially thought to be innocuous commensals, Staphylococci are increasingly being recognized as pathogens that are responsible for the etiology of a number of opportunistic human and animal diseases. S. aureus and S. epidermidis infections rank at the top of causative agents of nosocomial infections (1). Staphylococci have also been observed to rapidly gain antibiotic resistance (2). Hence it is important to understand the virulence mechanisms and to decipher the host immune responses.

Research on Staphylococci has mostly focused on S. aureus, which is a leading cause of bacteremia worldwide. S. aureus is a coagulase-positive bacterium which has an extensive arsenal of virulence factors comprising both structural and secreted products (3). S. epidermidis, a coagulase-negative bacterium, has remained relatively less-studied, plausibly because S. epidermidis infection is usually not life-threatening. Although S. epidermidis lacks aggressive virulence determinants, the genome possesses a wide range of genes that equip the organism to survive in harsh conditions to establish a commensal relation with its host (4). The bacterium possesses an extensive arsenal of osmoprotectants and produces exopolymers that are critical for its success as a commensal. Under pathogenic conditions, these determinants aid the bacterium in immune evasion (4). S. epidermidis shows high diversity, with 74 identified sequence types. Ironically, there is no genomic data available for the isolate that has most commonly been observed in infections and is thought to be potentially the most invasive (4). Multidrug resistant strains of S. epidermidis are quite widespread. Nearly 76% of clinical samples obtained from patients with catheter-related bacteremia were multidrug resistant (5). Although the prevalence of drug resistance is low in North America, it is as high as 40% in Asian and African populations (6, 7).

S. epidermidis is the most frequent pathogen causing infections in indwelling medical devices and early-onset neonatal sepsis (4, 8). The ability of this bacterium to form biofilm further protects it from host immunity and antimicrobial treatment regimens (9). S. epidermidis infections often lead to chronic infections, and occasionally to acute sepsis, the latter especially in neonates (10, 11). The high prevalence of multidrug resistant strains, particularly in the clinical setting, and the increasing incidence of such infections have made S. epidermidis infections a serious burden for health care worldwide (4, 6, 9).

Mechanisms of S. epidermidis biofilm formation and immune responses to the infection have been investigated in cell cultures and ex vivo models. Characterization of host immune responses to this pathogen in vivo, however, has been scarce (6). Here we report the establishment of an in vivo bath immersion model of S. epidermidis infection in zebrafish. We demonstrate an NF-κB mediated proinflammatory response that upregulates the IL-6 signaling pathway. IL-6 is a multifunctional cytokine which plays an important role in inflammation and immunity (12). IL-6 binds to IL-6 receptor α (IL-6Rα) which triggers the recruitment of the IL-6 signal transducer (IL-6st) that subsequently initiates the signal transduction. IL-6st signaling progresses via two major signaling pathways, namely the JAK/STAT-3 and SHP2/Gab/MAPK signaling cascades (12). Activation of the IL-6 pathway has been reported to elicit distinct and often contradictory signals. Evidences suggest that the IL-6 classic signaling leads to the regenerative and anti-inflammatory effects of IL-6, whereas IL-6 trans signaling is responsible for proinflammatory phenotypes (13). Given the importance of IL-6 signaling in host defense, the different genes in this pathway are regulated at multiple levels, including miRNA-mediated modulation (14). miR-142 is one of the major miRNAs that has been shown to have important functions in infections and inflammation (15). Here, we report a plausible role of miR-142-5p in IL-6 signal amplification.

Zebrafish were maintained at 28°C with a 14 h light/10 h dark cycle. Embryos were collected by natural spawning and maintained at 28.5°C in E3 medium. For imaging experiments, phenylthiourea (0.003%) was added to 1 d postfertilization (dpf) embryos to avoid melanization. Wild-type (Tü) strain and three transgenic reporters were used: Tg(8xHs.NF-κB:GFP,Luciferase) (hereafter known as Tg(NFB:GFP, Luc)), Tg(pu.1:Gal4.UAS RFP) (subsequently referred to as Tg(pu.1:RFP)), and Tg(mpeg1:Gal4-VP16/UAS:Kaede) (hereafter referred to as Tg(mpeg1:Kaede)) in this study. All the experimental procedures have been approved by the Institutional Biosafety and Animal Ethics committees.

Staphylococcus epidermidis (strain O-47) (16) was grown in tryptic soy broth supplemented with 1% NaCl. It was streaked onto HiCrome Staph Selective Agar (HiMedia) and tryptic soy agar plates. The S. epidermidis colonies were blue in color in the Staph Selective media.

An overnight culture of S. epidermidis was grown and 1:50 inoculum made on the day of infection. The bacteria were grown at 28.5°C till O.D600 1.0 and pelleted down by centrifugation at 4000 rpm for 10 min. The pellet was resuspended in sterile PBS and washed twice, and resuspended in E3 medium. In experiments using heat-killed bacteria, the final S. epidermidis suspension was incubated at 95°C for 5 min.

Zebrafish larvae (4 dpf) were infected by exposure to S. epidermidis in a bath immersion. Each well of a 24-well plate had 15–20 larvae in 2 ml of E3 medium. The zebrafish larvae were exposed to varying concentrations (5 × 108 CFU/ml, 1 × 109 CFU/ml, and 1 × 1010 CFU/ml) of S. epidermidis for either 6 h or 24 h. Subsequently, the larvae were removed from the infected E3 medium, washed twice in E3 medium and transferred to fresh medium for further study. The infection dose was validated by CFU plating. The time of addition of the bacterial inoculum was considered as time zero and time points were calculated as hours postinfection (hpi) thereafter. For survival studies, triplicate wells of each infection dose were monitored till 80 hpi.

For inhibition of IL-6 signaling, two STAT-3 inhibitors (STAT-3 inhibitor III [WP-1066] and STAT-3 inhibitor XIII [C-188-9]) were tested (Merck, Sigma). WP-1066 and C-188-9 were used for infection studies at final concentrations of 12 μM and 10 μM, respectively. The inhibitors were added to E3 medium 30 min prior to infection and the larvae were maintained in that medium during the entire period of the mortality study. Equivalent volumes of DMSO were added to wells with larvae serving as vehicle control.

From each experimental condition, three larvae were randomly selected at different time points postinfection. An individual larva was anesthetized with a lethal dose of tricaine, homogenized mechanically in 1% Triton X-100 in PBS and multiple dilutions of the lysate plated in triplicate. The plates were incubated at 37°C for 16–20 h, the colonies counted, and the CFU calculated accordingly.

For establishment of hypoxia, 4-dpf larvae were exposed to hypoxia (5% oxygen). The larvae were maintained in the hypoxic chamber (NuAire) set at 28.5°C for either 6 h or 24 h, then taken out and homogenized in TRIzol immediately.

Total RNA was extracted from a pool of 10–15 embryos at different time points from each experimental condition using TRIzol (Thermo Fisher Scientific) following the manufacturer’s protocol. cDNA was prepared by reverse transcription using SuperScript Vilo (Thermo Fisher Scientific). RNA was subjected to DNase I treatment prior to cDNA synthesis for real-time PCR for tagrfp. Gene-specific primers were designed using Universal Probe Library (Roche) and real-time PCR performed using GoTaq qPCR Master Mix (Promega) on ABI QuantStudio 7 Flex Real-Time PCR System. b-actin was used as a housekeeping gene for normalization. For miRNA quantitative PCR (qPCR), cDNA was prepared using miRNA first strand cDNA synthesis kit (Agilent). Mature miRNA sequence and Universal reverse primer (Agilent) were used as forward and reverse primers, respectively. The Tm of the qPCR was adjusted to 59°C (17). dre-miR-451 was used for normalization. The primer sequences are tabulated in Table I.

Table I.

Sequences of forward and reverse primers (5′-3′) used for qPCR

Gene namePrimer sequence (5′-3′)
b-actin Forward: GATCTTCACTCCCCTTGTTCA
Reverse: GGCAGCGATTTCCTCATC 
il-1b Forward: TGAAGTCACCATAGCTCCAAAAA
Reverse: GCATGTCGCATCTGTAGCTC 
tnf-a Forward: AGGCAATTTCACTTCCAAGG
Reverse: AGGTCTTTGATTCAGAGTTGTATCC 
il-6 Forward: AAGGGGTCAGGATCAGCAC
Reverse: GCTGTAGATTCGCGTTAGACATC 
il-6st Forward: TTCATTCCAAGATCAAATGACG
Reverse: TCCATCATGAACCCACCAC 
il-12b Forward: CGCTGTAGGAAACGCAAAA
Reverse: GGAGACTTTGTGTGCGGTAAG 
tlr-2 Forward: AGACACAATAATGGCAGTCAGG
Reverse: TACATGTCTGGGAGCACTCG 
tlr-4ba Forward: CATGCATGGGAAGAAACCTC
Reverse: CCAGAGTTTGAACCGAGGAA 
tlr-8a Forward: TGCCTTCATCACCTACGACA
Reverse: GATCGGGAGGAAGAGTTCG 
tlr-8b Forward: CGAGTGTCTTGCAACTGGAC
Reverse: TTGAGGCAAGTGATGTTCTCC 
hif-1a Forward: CATGAGGAAGCTGCTCAATTC
Reverse: GCTGACTTTCCAGCTCATTCTC 
mpeg1 Forward: TTCCGTAACAACACGAGTTCA
Reverse: TCAATAGCATCTGCGAGTTCC 
tagRFP Forward: CTGACCGCTACCCAGGAC
Reverse: CACCCCTCTGATCTTGACG 
dre-miR-451 Forward: CCGTTACCATTACTGAGTT 
dre-miR-142a-5p Forward: ACATAAAGTAGAAAGCACTACT 
dre-miR-142a-3p Forward: GTAGTGTTTCCTACTTTATGGA 
Gene namePrimer sequence (5′-3′)
b-actin Forward: GATCTTCACTCCCCTTGTTCA
Reverse: GGCAGCGATTTCCTCATC 
il-1b Forward: TGAAGTCACCATAGCTCCAAAAA
Reverse: GCATGTCGCATCTGTAGCTC 
tnf-a Forward: AGGCAATTTCACTTCCAAGG
Reverse: AGGTCTTTGATTCAGAGTTGTATCC 
il-6 Forward: AAGGGGTCAGGATCAGCAC
Reverse: GCTGTAGATTCGCGTTAGACATC 
il-6st Forward: TTCATTCCAAGATCAAATGACG
Reverse: TCCATCATGAACCCACCAC 
il-12b Forward: CGCTGTAGGAAACGCAAAA
Reverse: GGAGACTTTGTGTGCGGTAAG 
tlr-2 Forward: AGACACAATAATGGCAGTCAGG
Reverse: TACATGTCTGGGAGCACTCG 
tlr-4ba Forward: CATGCATGGGAAGAAACCTC
Reverse: CCAGAGTTTGAACCGAGGAA 
tlr-8a Forward: TGCCTTCATCACCTACGACA
Reverse: GATCGGGAGGAAGAGTTCG 
tlr-8b Forward: CGAGTGTCTTGCAACTGGAC
Reverse: TTGAGGCAAGTGATGTTCTCC 
hif-1a Forward: CATGAGGAAGCTGCTCAATTC
Reverse: GCTGACTTTCCAGCTCATTCTC 
mpeg1 Forward: TTCCGTAACAACACGAGTTCA
Reverse: TCAATAGCATCTGCGAGTTCC 
tagRFP Forward: CTGACCGCTACCCAGGAC
Reverse: CACCCCTCTGATCTTGACG 
dre-miR-451 Forward: CCGTTACCATTACTGAGTT 
dre-miR-142a-5p Forward: ACATAAAGTAGAAAGCACTACT 
dre-miR-142a-3p Forward: GTAGTGTTTCCTACTTTATGGA 

Zebrafish larvae were anesthetized with 0.01% tricaine, embedded in 1.2% low-melting-point agarose, and imaged using Nikon TiEB (20×) and Leica DMi8 (20× and 40×). A total of 20–30 z-stacks of 2.5-μm thickness was captured. Image analysis was done using Fiji software.

NF-κB activity of infected and uninfected Tg(NFB:GFP, Luc) larvae was visualized by imaging similar regions of uninfected controls and infected larvae at different time points postinfection. In order to enumerate leukocytes, multiple fields of uninfected and infected Tg(pu.1:RFP) larvae, spanning the trunk and tail regions, were imaged at different time points postinfection. The total number of red fluorescent protein (RFP)-positive cells in identical image volumes (full field, z-thickness of 50 μm in trunk and 25 μm in tail) of uninfected and infected larvae were counted manually and the average number of leukocytes was calculated. Double-positive larvae, obtained from mating Tg(pu.1:RFP) with Tg(mpeg1:Kaede), were used to count number of macrophages (RFP and Kaede [green] positive) and neutrophils (RFP positive) (n = 3–7 larvae). Image analysis was done using Fiji software (18).

Tg(NFB:GFP, Luc) larvae were used to quantify NF-κB activity by assaying luciferase activity as reported (19). Briefly, an individual larva was transferred to each well of a 96-well optical bottom plate (Nunc) in 50 μl of E3 medium, without methylene blue (n = 9–13 larvae). The medium was supplemented with 1 mM beetle luciferin potassium salt solution (Promega) and bioluminescence assayed at room temperature using Varioskan Flash (Thermo Fisher).

Data were plotted using GraphPad PRISM. Multiple t tests and two-way ANOVA were performed for statistical analysis. A p value < 0.05 at a confidence interval of 95% was considered significant. Given the limitations of the p value in understanding the biological significance of the differences (20), effect size was calculated for all the statistically significant differences observed. Cohen’s d (for same sample size) or Hedge’s g (for different sample sizes) was calculated and effect sizes were classified as small (d = 0.2), medium (d = 0.5), and large (d ≥ 0.8) (21).

As a first step toward establishment of this host–pathogen model, we wished to ascertain the bacterial doses to which zebrafish are susceptible, and the infection model is amenable to experimentation. In order to achieve this, 4-dpf larvae were subjected to three doses of bacteria, 5 × 108 (low), 1 × 109 (medium), and 1 × 1010 (high) CFU/ml of S. epidermidis, for an exposure period of either 6 h or 24 h. Survival was then monitored till 80 hpi, 0 h being the beginning of the bacterial exposure (Fig. 1A). The larvae incubated with the highest dose of bacteria for 6 h (Inf-high-6h) showed severe mortality (open inverted triangles). As early as 8 hpi, survival dropped to 65% (p = 0.002), and by 16 hpi less than 48% survived (p = 0.0001; d = 6.93). Longer (24 h) exposure at this dose caused death of most of the animals and hence was not analyzed. For the doses 5 × 108 CFU/ml (open squares, Inf-low-6h) and 1 × 109 CFU/ml (open circles, Inf-med-6h), a 6 h exposure resulted in 95% and 83% larvae surviving, respectively. The mortalities in these cases were not significantly different with respect to the control population. However, when the exposure window was increased to 24 h, both these infection doses (Inf-low-24h and Inf-med-24h) caused significant mortalities. At 16 hpi, larvae exposed to Inf-med-24h (closed circles) had 75% survivors (p = 0.048). The death steadily increased till 48 hpi when there were only 40% survivors (p = 0.0001; d = 3.92). Thereafter, no further significant mortality was observed. For larvae exposed to Inf-low-24h (closed squares), increasing larval death was apparent from 12 hpi, but the mortality reached values significantly higher than controls at 48 hpi (p = 0.02). Thereafter it stabilized and at the end of the study, at 80 hpi, 73% of the larvae had survived (p = 0.004; d = 1.3). Heat-killed bacteria (5 × 108 CFU/ml, gray squares, heat-killed-low-24h; 1 × 109 CFU/ml, gray circles, heat-killed-med-24h) did not cause any mortality, indicating that mortalities observed were due to S. epidermidis infection.

FIGURE 1.

Zebrafish larvae infected with S. epidermidis show bacterial colonization followed by clearance. (A) Larvae, 4 dpf, were exposed to varying doses (low: 5 × 108 CFU/ml, medium: 1 × 109 CFU/ml, and high: 1 × 1010 CFU/ml) of S. epidermidis for 6 or 24 h and mortality monitored. Each well had 15 larvae. Larvae exposed to low or medium doses of bacteria for 6 h (Inf-low-6h and Inf-med-6h) did not show significant mortality. Significant mortalities were observed in larvae exposed to high dose for 6 h (Inf-high-6h) and medium and low doses for 24 h (Inf-med-24h and Inf-low-24h). Heat-killed bacteria had no effect on zebrafish survival. (B) S. epidermidis colonization and clearance was monitored by CFU plating of uninfected, Inf-low-24h, and Inf-med-24h larvae (15 larvae per well). Compared to control larvae, the CFU counts of infected larvae remained significantly high till 24 hpi. Inf-low-24h showed clearance post-24 hpi, but CFU counts in Inf-med-24h larvae remained significantly elevated till 72 hpi. Inf-low-24h had significantly lower CFU counts than Inf-med-24h between 16 and 48 hpi. (C) Bacterial clearance was monitored by infecting 4-dpf larvae with Inf-low-6h (15 larvae per well). Infected larvae had significantly higher CFU counts at 2 hpi compared with controls. At later time points there was no difference in the CFU counts between uninfected and Inf-low-6h larvae. (*p < 0.05, **p < 0.01, ****p < 0.0001, #p < 0.05, ##p < 0.01). The data are representative of three independent experiments.

FIGURE 1.

Zebrafish larvae infected with S. epidermidis show bacterial colonization followed by clearance. (A) Larvae, 4 dpf, were exposed to varying doses (low: 5 × 108 CFU/ml, medium: 1 × 109 CFU/ml, and high: 1 × 1010 CFU/ml) of S. epidermidis for 6 or 24 h and mortality monitored. Each well had 15 larvae. Larvae exposed to low or medium doses of bacteria for 6 h (Inf-low-6h and Inf-med-6h) did not show significant mortality. Significant mortalities were observed in larvae exposed to high dose for 6 h (Inf-high-6h) and medium and low doses for 24 h (Inf-med-24h and Inf-low-24h). Heat-killed bacteria had no effect on zebrafish survival. (B) S. epidermidis colonization and clearance was monitored by CFU plating of uninfected, Inf-low-24h, and Inf-med-24h larvae (15 larvae per well). Compared to control larvae, the CFU counts of infected larvae remained significantly high till 24 hpi. Inf-low-24h showed clearance post-24 hpi, but CFU counts in Inf-med-24h larvae remained significantly elevated till 72 hpi. Inf-low-24h had significantly lower CFU counts than Inf-med-24h between 16 and 48 hpi. (C) Bacterial clearance was monitored by infecting 4-dpf larvae with Inf-low-6h (15 larvae per well). Infected larvae had significantly higher CFU counts at 2 hpi compared with controls. At later time points there was no difference in the CFU counts between uninfected and Inf-low-6h larvae. (*p < 0.05, **p < 0.01, ****p < 0.0001, #p < 0.05, ##p < 0.01). The data are representative of three independent experiments.

Close modal

These results demonstrated that within the range of dose and exposure tested, mortality and severity of the disease were dependent on the dose and duration of exposure to the bacteria. Based on these results Inf-low (5 × 108 CFU/ml) and Inf-med (1 × 109 CFU/ml) were chosen for further analyses in this study as these doses produced phenotypes that were consistently measurable. Given the high lethality that was observed, Inf-high-6h was not used for further studies.

We then followed the infection progress and looked at initial colonization of S. epidermidis followed by its clearance over time (Fig. 1B). Because larvae exposed to the low and medium doses of bacteria for 6 h did not show significant mortality, the longer exposure period (24 h) was chosen for this purpose. CFUs were counted at different time points till 72 hpi from uninfected control, Inf-low-24h, and Inf-med-24h larvae. The CFU counts from the uninfected larvae remained low throughout the experiment. The infected larvae of both the doses showed significantly higher CFU counts than controls from 2 hpi till 24 hpi (p = 0.013 for Inf-low-24h; p = 0.047 at 2 and 16 hpi, p = 0.008 at 24 hpi for Inf-med-24h). These differences corresponded to large effect sizes: d = 2.25, 1.03, and 3.54 for Inf-low-24h and d = 4.32, 2.90, and 10.5 for Inf-med-24h at 2, 16, and 24 hpi, respectively. At 2 hpi, there was no difference between the two infected groups indicating that the initial colonization efficiency was not influenced by the dose. From the next time point that we checked, 16 hpi onward, although both the populations showed progressively decreasing CFU yields, the larvae infected with Inf-med-24h consistently showed significantly higher CFU counts than those infected with Inf-low-24h (p = 0.047, d = 1.21; p = 0.0055, d = 2.37). At 48 hpi, the CFU counts of Inf-low-24h larvae had decreased to values similar to the control population. The CFU counts of Inf-med-24h larvae also showed progressive decrease, but remained significantly elevated compared with the controls till the end of the experiment, 72 hpi (p = 0.011, d = 4.84). In both the doses, viability of S. epidermidis in the infection medium did not reduce significantly over the 24 h period of incubation (CFU counts of 3.63 × 108 ± 3.1 × 106 CFU/ml and 0.8 × 109 ± 0.6 × 107 CFU/ml as opposed to initial densities of 5 × 108 ± 4.2 × 106 and 1 × 109 ± 1.8 × 107, respectively).

Although the CFU counts decreased over time after exposure to Inf-low-24h, the CFU counts remained elevated and bacteria were not completely cleared till 72 hpi when exposed to Inf-med-24h. The reduction in CFU counts coincides with significant larval death. Hence it can be argued that the reduction in CFU is due to the severely infected larvae dying out, leaving behind survivors with lower levels of infection, and thus is not indicative of bacterial clearance per se. This argument, however, does not explain the difference in the CFU counts that are observed between the two infection doses. The larvae infected with low dose showed lower mortalities but were able to clear bacteria faster than the larvae infected with medium dose, as evidenced by the CFU counts over time. Additionally, larvae of Inf-med-24h show reduction of bacterial counts post-24 hpi without any significant accompanying death. This can be seen in the significant decrease in CFU counts between 24 hpi and 72 hpi of this population (CFU counts: p = 0.0017, d = 7.70; larval mortality: p = 0.87). This strongly suggests that the zebrafish are able to clear the bacteria after the initial colonization phase at these multiplicities of infection (MOIs), and this phenomenon contributes, at least in part, to the reduction in bacterial counts in the surviving larvae.

To investigate further, we tested bacterial colonization and clearance in larvae infected with Inf-low-6h, a dose that showed minimal mortality and was indistinguishable from the uninfected controls (Fig. 1C). At 2 hpi, Inf-low-6h larvae yielded significantly higher CFU counts than their uninfected counterparts (p = 5.17 × 10−8, d = 1.70) indicating that infection does occur at this dose. However, by 8 hpi, the CFU counts from the infected population had fallen to values similar to the uninfected controls. This demonstrates that zebrafish are able to clear the infecting bacteria likely by mounting an immune response against the infectious agent. Exposure to higher density or duration of the pathogen has deleterious effects likely by overwhelming host clearance mechanisms, and thus leading to colonization.

We next aimed to identify the cellular lineage(s) that might be involved in mediating the immune response to S. epidermidis infection. Both macrophages and neutrophils have been shown to be involved in host immune responses to staphylococcal infections (9, 22, 23). Given that the leukocytic population of 4-dpf larvae is primarily made up of macrophages and neutrophils, the cells of myeloid lineage, we used Tg(pu.1:RFP), a transgenic reporter fish line to investigate the role of circulating leukocytes in the response to S. epidermidis infection. The pu.1 promoter, a myeloid lineage–specific transcription factor (24), drives the expression of TagRFP in these fish. Hence, the number of pu.1-positive cells is indicative of the total number of leukocytes in the larvae. To determine whether S. epidermidis infection resulted in a change in the migration of leukocytes, Tg(pu.1:RFP) larvae were infected with the Inf-med-24h dose, and tail and trunk regions of uninfected and infected larvae were imaged at different time points postinfection. A higher number of pu.1-positive cells was seen in the infected larvae compared with the controls. This indicated that there was a greater number of leukocytes closer to the larval surface upon infection. In order to quantitate this, the total numbers of pu.1-positive cells in the same segment and identical image volume of multiple infected and control larvae were counted and the average values compared. The infected larvae showed a significant increase in the number of pu.1-positive cells visible in the imaged trunk region as early as 2 hpi (Fig. 2A, 2B). The infected larvae had an average of 24.6 ± 7.0 pu.1-positive cells at this time point compared with controls (11.6 ± 5.1) (p = 0.0003, d = 2.24) (Fig. 2E). The number of leukocytes that were mobilized to the larval surface in the infected population remained elevated even at 24 hpi, showing a nearly 3-fold increase over the uninfected control population (Fig. 2F). At 24 hpi, the average number of pu.1-positive cells in Inf-med-24h larvae was 19.3 ± 4.5 as opposed to 6.3 ± 4.0 in controls (p = 0.0008, d = 2.55) (Fig. 2C–E). Although the data presented here reflect the number of pu.1-positive cells in the trunk region (Fig. 2A′–D′), we also counted the number of leukocytes in the tail region of the larvae (Supplemental Fig. 1). In the tail region also there was a significant increase in the number of pu.1-positve cells visible (Supplemental Fig. 1B, 1D, 1E), upon S. epidermidis infection.

FIGURE 2.

S. epidermidis infection results in an increase in leukocyte migration. Tg(pu.1:RFP), 4 dpf, were infected with Inf-med-24h dose, and number of pu.1-positive cells counted in similar segments of control (A and C) and infected (B and D) larvae. (AD) Brightfield images of the representative regions. (E) The average number of pu.1 cells in Inf-med-24h and uninfected controls were plotted at 2 hpi and 24 hpi (n = 6–7 larvae). (F) The Inf-med-24h larvae showed significant fold increase at 2 and 24 hpi. (**p < 0.01, ***p < 0.001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

FIGURE 2.

S. epidermidis infection results in an increase in leukocyte migration. Tg(pu.1:RFP), 4 dpf, were infected with Inf-med-24h dose, and number of pu.1-positive cells counted in similar segments of control (A and C) and infected (B and D) larvae. (AD) Brightfield images of the representative regions. (E) The average number of pu.1 cells in Inf-med-24h and uninfected controls were plotted at 2 hpi and 24 hpi (n = 6–7 larvae). (F) The Inf-med-24h larvae showed significant fold increase at 2 and 24 hpi. (**p < 0.01, ***p < 0.001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

Close modal

If leukocyte mobilization plays a critical role in limiting and clearing the bacterial infection, then the larvae infected with Inf-low-6h, the dose that clears S. epidermidis asymptomatically, should also exhibit elevation in the number of pu.1-positive cells in surface regions. Inf-low-6h larvae show an initial increase in the number of pu.1-positive cells followed by a gradual decrease in the imaged trunk regions (Fig. 3). At 2 hpi, more than 2-fold increase in the number of pu.1-positive cells was seen in Inf-low-6h larvae (22.6 ± 14) compared with uninfected controls (9.9 ± 5.4) (p = 0.0012, d = 1.1) (Fig. 3A, 3B, 3I, 3J). This was comparable with the increase observed in the Inf-med-24h case. At 6 hpi, the infected larvae still showed 1.9-fold increase (21.8 ± 11.7 pu.1-positive cells) (p = 0.0069, d = 1.11) (Fig. 3C, 3D, 3I, 3J), which fell to 1.7-fold by 18 hpi (20.6 ± 7.9 pu.1-positive cells) (p = 0.015, d = 1.40) (Fig. 3E, 3F, 3I, 3J). The number of pu.1-positive cells in uninfected larvae at these time points was 11.0 ± 4.6 and 11.7 ± 3.7, respectively. By 24 hpi, the number of pu.1-positive cells in the infected larvae (13.2 ± 4.0) was similar to the uninfected larvae (12.4 ± 3.6) (Fig. 3G–J). The data suggested that leukocyte mobilization is a critical component of the immune response mounted by infected larvae in our model. The S. epidermidis infection–induced increase in peripheral leukocytes appears to be concomitant with bacterial colonization. The resolution phase of this phenomenon, however, tails off in time beyond the infection clearance window.

FIGURE 3.

Leukocytes are critical in mediating bacterial clearance. Tg(pu.1:RFP), 4 dpf, were infected with Inf-low–6h, and number of pu.1-positive cells counted in similar segments of control and infected larvae at 2 hpi (A and B), 6 hpi (C and D), 18 hpi (E and F), and 24 hpi (G and H), respectively. (AH) Brightfield images of the representative regions. (I) The average number of pu.1 cells in Inf-low-6h and uninfected controls were plotted at each time point. Compared with 24 hpi, Inf-low-6h larvae showed significant fold increase in pu.1 cells at 2 hpi, 6 hpi, and 18 hpi. (J) The fold increase in pu.1-positive cells was significant till 18 hpi. Compared with 24 hpi, Inf-low-6h larvae showed significant fold increase in pu.1 cells at 2 hpi and 6 hpi. (*p < 0.05, **p < 0.01, #p < 0.05, ##p < 0.01). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

FIGURE 3.

Leukocytes are critical in mediating bacterial clearance. Tg(pu.1:RFP), 4 dpf, were infected with Inf-low–6h, and number of pu.1-positive cells counted in similar segments of control and infected larvae at 2 hpi (A and B), 6 hpi (C and D), 18 hpi (E and F), and 24 hpi (G and H), respectively. (AH) Brightfield images of the representative regions. (I) The average number of pu.1 cells in Inf-low-6h and uninfected controls were plotted at each time point. Compared with 24 hpi, Inf-low-6h larvae showed significant fold increase in pu.1 cells at 2 hpi, 6 hpi, and 18 hpi. (J) The fold increase in pu.1-positive cells was significant till 18 hpi. Compared with 24 hpi, Inf-low-6h larvae showed significant fold increase in pu.1 cells at 2 hpi and 6 hpi. (*p < 0.05, **p < 0.01, #p < 0.05, ##p < 0.01). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

Close modal

Given that both macrophages and neutrophils have been implicated in host immune responses to S. epidermidis (9), we next sought to analyze the relative contributions of the two populations to the elevated peripheral numbers observed in our model. The pu.1-expressing responder cell population could be macrophages, or neutrophils, or both. In order to differentiate between the two lineages, larvae obtained from mating Tg(pu.1:RFP) with Tg(mpeg1:Kaede) were used for infection and imaging experiments. Only macrophages are labeled by Tg(mpeg1:Kaede) (25). These larvae were imaged under conditions where Kaede behaves as a GFP. As can be seen from (Fig. 4, the number of neutrophils in the areas imaged remained unaffected by S. epidermidis infection (Fig. 4E), whereas there was a significant increase in the number of macrophages at both 2 hpi (Fig. 4A″, 4B″, 4E) and 24 hpi (p = 0.0001; g = 1.56 and 2.92 for 2 and 24 hpi, respectively) (Fig. 4C″, 4D″, 4E). The average number of macrophages in Inf-med-24h larvae at 2 and 24 hpi were 22.4 ± 7.7 and 23.6 ± 3.0, respectively, whereas the macrophage numbers in controls at the corresponding time points were 11.9 ± 2.8 and 13.2 ± 3.6, respectively. In order to corroborate the finding, we counted the number of macrophages and neutrophils in the tail region of the larvae (Supplemental Fig. 2). Infection resulted in increased macrophage counts in the imaged regions whereas the neutrophil numbers remained unchanged (p = 0.003) (Supplemental Fig. 2). The increase in leukocyte numbers close to the larval surface thus mostly owes to increased number of peripheral macrophages. Hence it is likely that macrophages are the predominant cellular responders to S. epidermidis infection.

FIGURE 4.

Macrophages are central responders to S. epidermidis infection. Double-positive larvae obtained from Tg(pu.1:RFP) × Tg(mpeg:Kaede) mating were infected with Inf-med-24h and number of macrophages and neutrophils enumerated at 2 and 24 hpi at similar segments of controls (AA and CC) and infected cohorts (BB and DD). There was a significant increase in the macrophage numbers in the infected population at both 2 (A versus B) and 24 hpi (C versus D). (E, open and closed circles) The average number of macrophages at 2 hpi and 24 hpi were significantly elevated in infected larvae in comparison with controls. The neutrophil numbers were comparable between the two experimental groups, at both the time points (A″ versus B″ at 2 hpi; C″ versus D″ at 24 hpi; E, open and closed triangles) (n = 9–13 larvae). Apparent increase in macrophage number is unlikely due to increased differentiation (F and G). mRNA levels of tagrfp (F) and mpeg1 (G) in Inf-med-24h larvae remained unchanged at 2 and 24 hpi. (****p < 0.0001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

FIGURE 4.

Macrophages are central responders to S. epidermidis infection. Double-positive larvae obtained from Tg(pu.1:RFP) × Tg(mpeg:Kaede) mating were infected with Inf-med-24h and number of macrophages and neutrophils enumerated at 2 and 24 hpi at similar segments of controls (AA and CC) and infected cohorts (BB and DD). There was a significant increase in the macrophage numbers in the infected population at both 2 (A versus B) and 24 hpi (C versus D). (E, open and closed circles) The average number of macrophages at 2 hpi and 24 hpi were significantly elevated in infected larvae in comparison with controls. The neutrophil numbers were comparable between the two experimental groups, at both the time points (A″ versus B″ at 2 hpi; C″ versus D″ at 24 hpi; E, open and closed triangles) (n = 9–13 larvae). Apparent increase in macrophage number is unlikely due to increased differentiation (F and G). mRNA levels of tagrfp (F) and mpeg1 (G) in Inf-med-24h larvae remained unchanged at 2 and 24 hpi. (****p < 0.0001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 25 μm.

Close modal

Zebrafish larvae, which still lack an adaptive immunity, have been reported to induce emergency granulopoiesis (26). Although the imaging data did not point toward increase in neutrophil counts, this phenomenon cannot be ruled out, given that the imaging captures limited volumes close to the larval surface. To address this point, mRNA levels of tagrfp were ascertained, as a readout of pu.1 transcript levels, between uninfected and Inf-med-24h larvae of the Tg(pu.1:RFP) line (Fig. 4F). There was no significant difference in the mRNA levels of tagRFP between infected and control populations (Fig. 4F). Given that the imaging experiments showed an apparent increase in macrophage numbers, we checked the levels of mpeg1 transcripts from these larvae and found those also to be similar between the two groups (Fig. 4G). Taken together, the data indicate that the apparent increase in macrophage numbers upon S. epidermidis infection most likely stems from an increase in the number of peripheral macrophages that come into circulation and not from differentiation of hematopoietic stem cells.

We were interested in understanding the signaling pathways that are active in the host response to S. epidermidis infection. Tlrs act at the forefront of innate host defenses, and given the importance of Tlr-2 signaling in driving host immune response to Gram-positive bacteria (27), we measured the transcript levels of tlr-2. Real-time qPCR experiments showed that tlr-2 mRNA levels were significantly upregulated upon infection in our model (Fig. 5A). As early as 2 hpi, the Inf-med-24h larvae, compared with controls, had over 1.5-fold increase in tlr-2 mRNA levels (p = 0.007, d = 3.45). The levels increased further at 16 hpi (over 5.5-fold, p = 0.0007, d = 5.0) and remained high even at 24 hpi (over 3-fold, p = 0.0013, d = 6.04). Apart from Tlr-2, involvement of Tlr-4 and Tlr-8 have been reported in murine and human cell lines, respectively, in response to S. aureus infection (28, 29). When we looked at tlr-4b levels, we noticed a modest but steady rise of this transcript in infected larvae over the uninfected controls (Fig. 5B). At 2 hpi, there was over 1.5-fold increase in tlr-4b levels. The tlr-4b levels rose to 2.5-fold at 16 hpi and remained elevated till 24 hpi (2.9-fold, p = 0.001, d = 4.6). The levels of tlr-8a and tlr-8b remained unaffected (Supplemental Fig. 3).

FIGURE 5.

S. epidermidis infection upregulates tlr-2 and activates NF-κB pathway. Transcriptional upregulation of tlr-2 (A) and tlr-4b (B) was tested by real-time qPCR. (A) tlr-2 showed significant upregulation in Inf-med-24h larvae as early as 2 hpi and the levels remained elevated till 24 hpi. (B) tlr-4b mRNA expression was also upregulated upon infection. Data representative of three independent experiments. (CI) Tg(NFB:GFP, Luc) larvae were infected with Inf-med-24h and NF-κB activity was followed by GFP expression (CH) and luciferase activity (I). At 6 hpi (C and D), there was no discernible increase in NF-κB activity between the two experimental conditions. Infected larvae exhibited higher GFP expression than the control population at 20 hpi (E and F) and 42 hpi (G and H). (CH) Brightfield images of the representative regions. (I) The luciferase levels in infected larvae were significantly elevated 20 hpi onward (9–13 larvae). (*p < 0.05, **p < 0.01, ***p < 0.001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 10 μm.

FIGURE 5.

S. epidermidis infection upregulates tlr-2 and activates NF-κB pathway. Transcriptional upregulation of tlr-2 (A) and tlr-4b (B) was tested by real-time qPCR. (A) tlr-2 showed significant upregulation in Inf-med-24h larvae as early as 2 hpi and the levels remained elevated till 24 hpi. (B) tlr-4b mRNA expression was also upregulated upon infection. Data representative of three independent experiments. (CI) Tg(NFB:GFP, Luc) larvae were infected with Inf-med-24h and NF-κB activity was followed by GFP expression (CH) and luciferase activity (I). At 6 hpi (C and D), there was no discernible increase in NF-κB activity between the two experimental conditions. Infected larvae exhibited higher GFP expression than the control population at 20 hpi (E and F) and 42 hpi (G and H). (CH) Brightfield images of the representative regions. (I) The luciferase levels in infected larvae were significantly elevated 20 hpi onward (9–13 larvae). (*p < 0.05, **p < 0.01, ***p < 0.001). Box represents the area imaged. Images are representative of three independent experiments. Quantification reflects combined data of three independent experiments. Scale bar, 10 μm.

Close modal

Activation of the NF-κB pathway has been shown to occur as a consequence of activation of macrophages as well as upregulation of Tlr signaling pathways (30, 31). We looked into the activation status of the NF-κB pathway in our model using the Tg(NFB:GFP, Luc) transgenic reporter line, where an NF-κB responsive element drives the expression of GFP and luciferase. GFP intensities from similar regions of interest of uninfected and Inf-med-24h larvae were imaged at different time points postinfection (Fig. 5C–H). At 6 hpi, there was no discernible difference in GFP intensities between the control population and the infected population (Fig. 5C, 5D). However, at 20 hpi, Inf-med-24h larvae had significantly higher GFP signals compared with the controls, indicating activation of the NF-κB pathway (Fig. 5E, 5F). The infected larvae showed elevated GFP signal till 42 hpi (Fig. 5G, 5H). We further corroborated these findings by quantitating the luciferase activity. As can be seen from (Fig. 5I, there was no detectable increase in NF-κB levels at 6 hpi. By 20 hpi, Inf-med-24h larvae showed elevated luciferase levels, indicative of increased NF-κB activity (p = 0.046, d = 1.85). Consistent with the GFP data, luciferase levels were significantly higher in Inf-med-24h larvae at 30 and 42 hpi (p = 0.029 and p = 0.0008; d = 3.37 and 1.70, respectively) (Fig. 5I). Both the GFP and luciferase data indicated that NF-κB signaling remained elevated till 42 hpi.

NF-κB is a cardinal mediator of inflammation. Given the elevated levels of NF-κB in our infection model, our next aim was to characterize the inflammatory profile of the larvae infected with S. epidermidis. Toward this objective, we analyzed the mRNA levels of a select set of NF-κB–responsive proinflammatory cytokines at different times postinfection by real-time qPCR (Fig. 6). The mRNA levels of il-1b (Fig. 6A), tnf-a (Fig. 6B), il-6 (Fig. 6C) and il-12b (Fig. 6D) were normalized against the housekeeping gene b-actin. As early as 2 hpi, mRNA levels of il-1b were significantly elevated (2.9-fold) in infected larvae. The il-1b levels in infected larvae peaked at 16 hpi, (16.5-fold rise) and then abated. At 24 hpi, the il-1b levels were still significantly high (6.5-fold) (p = 0.0003; p = 4.9 × 10−5; d = 9.9, 6.0, and 13.4 at 2, 16, and 24 hpi, respectively) (Fig. 6A). tnf-a mRNA levels showed a similar pattern: a rapid rise at 2 hpi (over 3-fold), with further increase at 16 hpi (over 11-fold), followed by slight abatement at 24 hpi (4-fold) (Fig. 6B) (p = 0.047; p = 0.0008; d = 2.2, 9.2, and 7.9 at 2, 16, and 24 hpi, respectively). We saw upregulation of il-1b and tnf-a mRNAs as early as 2 hpi (Fig. 6A, 6B), even though NF-κB activation was evident by 20 hpi (Fig. 5I). A likely reason for this apparent delay is the difference in the nature of the respective measurements. NF-κB activation was measured as a function of mature reporter proteins (GFP and luciferase), whereas NF-κB–responsive genes were detected at their mRNA levels. However, it is worth noting, that upregulation of the cytokines closely matches the increase in macrophage numbers (Fig. 4), indicating a precise and swift host immune response.

FIGURE 6.

Zebrafish host response has a proinflammatory signature. mRNA levels of various proinflammatory cytokines were investigated in control and Inf-med-24h larvae by real-time qPCR. il-1b (A), tnf-a (B), and il-6 (C) had similar expression profiles and showed peak expression at 16 hpi. il-12b (D) levels between the two experimental groups were similar till 16 hpi. At 24 hpi, the infected larvae had a modest but significant increase in il-12b levels. (E) Induction of hypoxia owing to static immersion was tested by monitoring mRNA levels of hif-1a. Uninfected larvae maintained under conditions of hypoxia (5% oxygen) for 6 or 24 h served as positive controls. Inf-low-6h and Inf-med-6h did not show hif-1a elevation. Compared with controls, larvae of all the other groups (Hypoxia-6h, Hypoxia-24h, Inf-low-24h, and Inf-med-24h) had significant elevation of hif-1a. (F) Hypoxia-24h larvae showed mild elevation of il-1b, tnf-a, il-6, and il-12b mRNA. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). Data shown are representative of three independent experiments.

FIGURE 6.

Zebrafish host response has a proinflammatory signature. mRNA levels of various proinflammatory cytokines were investigated in control and Inf-med-24h larvae by real-time qPCR. il-1b (A), tnf-a (B), and il-6 (C) had similar expression profiles and showed peak expression at 16 hpi. il-12b (D) levels between the two experimental groups were similar till 16 hpi. At 24 hpi, the infected larvae had a modest but significant increase in il-12b levels. (E) Induction of hypoxia owing to static immersion was tested by monitoring mRNA levels of hif-1a. Uninfected larvae maintained under conditions of hypoxia (5% oxygen) for 6 or 24 h served as positive controls. Inf-low-6h and Inf-med-6h did not show hif-1a elevation. Compared with controls, larvae of all the other groups (Hypoxia-6h, Hypoxia-24h, Inf-low-24h, and Inf-med-24h) had significant elevation of hif-1a. (F) Hypoxia-24h larvae showed mild elevation of il-1b, tnf-a, il-6, and il-12b mRNA. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001). Data shown are representative of three independent experiments.

Close modal

In contrast to il-1b and tnf-a, mRNA levels of both il-6 and il-12b showed a delayed response in infected larvae (Fig. 6C, 6D). The levels of il-6 transcripts were similar to that of uninfected controls at 2 hpi. But by 16 hpi there was significant transcriptional upregulation of il-6 (8-fold increase) and it remained elevated till 24 hpi (7.5-fold increase) (p = 0.0001; g = 8.5 and 4.1 at 16 and 24 hpi, respectively) (Fig. 6C). mRNA levels of il-12b remained unaffected till 16 hpi, and showed a mild, albeit significant elevation at 24 hpi (1.5-fold, p = 0.009, d = 2.1 at 24 hpi) (Fig. 6D). In terms of effect size, all the significant differences were large (d ≥ 2.0).

Pseudomonas aeruginosa infection of zebrafish by static immersion has been shown to enrich the activation of the hypoxia signaling pathway (32). Hence we sought to address the contribution of hypoxia, if any, toward the immune response that was observed in our immersion model in all the infection conditions. HIF-1α plays an integral role in hypoxia signaling and is upregulated during hypoxia (33). Inf-low-6h and Inf-med-6h did not show elevation of hif-1a levels at the end of the 6 h exposure window (∼0.8-fold and 0.92-fold, respectively). Larvae exposed to hypoxia (5% oxygen) for 6 h showed ∼1.4-fold increase in hif-1a (p = 0.005; g = 3.05). Larvae exposed for 24 h to either low or medium dose showed a similar increase in hif-1a levels (∼1.5-fold with respect to controls, p = 0.003, g = 0.98) (Fig. 6E). Larvae that were exposed to hypoxia for 24 h showed ∼2-fold elevation (p = 0.0002, g = 4.7) (Fig. 6E).

Hypoxia signaling has major implications in various infection models and can also modulate immune signaling (34). Hence it is plausible that elevation of the proinflammatory cytokines reported here is mainly due to hypoxia. In order to investigate this, we looked at the effect of hypoxia on the transcriptional regulation of il-1b, tnf-a, il-6, and il-12b (Fig. 6F). Compared with controls (normoxic larvae) all the four cytokines were elevated upon induction of hypoxia (p = 0.04; p = 0.004 and 0.005, respectively). These significances corresponded to large effect sizes, with d values of 2.13 (il-1b), 4.82 (tnf-a), 1.86 (il-6), and 4.0 (il-12b). However, when the elevation of transcript levels of il-1b, tnf-a, and il-6 in hypoxic larvae was compared with that of the infected larvae, the latter showed significantly greater elevation (comparing (Fig. 6F with (Fig. 6A–C). Hypoxia resulted in ∼1.5- to 2-fold increase in il-1b, tnf-a, and il-6 levels. The infected larvae, however, showed 6.5-fold, 4-fold, and 7.5-fold elevation of il-1b, tnf-a, and il-6 levels, respectively at 24 hpi (for il-1b: p = 0.001, d = 6.87; for tnf-a: p = 0.003, d = 4.23; for il-6: p = 0.002, d = 5.09). The il-12b transcript levels, however, showed a similar increase in larvae that were hypoxic and those infected with S. epidermidis (Fig. 6F, 6D). It is thus possible that il-12b elevation is an artifact of hypoxic stress and is not a component of the host immune response to S. epidermidis infection per se. The data indicate that although prolonged incubation in the bacterial medium results in mild hypoxia, it has a limited role in shaping the host immune response that we have observed in this model of S. epidermidis infection. Collectively the data indicated that S. epidermidis infection resulted in the activation of the NF-κB pathway that led to a robust proinflammatory profile with hypoxia playing a minor role in the host immune response.

We next wanted to elucidate the significance of the enhanced IL-6 signaling in zebrafish host response to S. epidermidis. IL-6 signals through the JAK/STAT-3 pathway and hence we inhibited the IL-6 signaling pathway with the help of two well-known STAT-3 inhibitors (WP-1066 and C-188-9) (35, 36). WP-1066 and C-188-9 were used for infection experiments at final concentrations of 12 μM and 10 μM, respectively (Fig. 7A and Supplemental Fig. 4B), after identification of the highest doses that caused no mortality on treatment for 80 h (Supplemental Fig. 4A). Uninfected larvae maintained in E3 medium with WP-1066 (inhibitor) or DMSO (vehicle control) showed no significant morbidities or mortalities. Infection progression was compared between Inf-med-24h larvae maintained in the presence of DMSO or WP-1066 (DMSO–Inf-med-24h and Inh W–Inf-med-24h, respectively) (Fig. 7A, open and closed circles). From 20 hpi, Inf-med-24h larvae in WP-1066 showed significant mortality when compared with its uninfected counterpart (p = 9 × 10−4). At this time point, the survival values in the two infected populations were comparable (80% for DMSO–Inf-med-24h versus 79% for Inh W–Inf-med-24h). However, by 54 hpi, Inh W–Inf-med-24h group exhibited significantly lower survival values (51%) compared with its untreated, DMSO–Inf-med-24h, counterparts (71%) (p = 0.0008). At the end of the study, 57% of larvae maintained in DMSO (DMSO–Inf-med-24h) survived as opposed to 25% of Inh W–Inf-med-24h (p = 1 × 10−5) (Fig. 7A, open and closed circles), corresponding to a large effect size (d = 3.9). This indicated that perturbation of IL-6 signaling rendered the larvae more susceptible to infection.

FIGURE 7.

IL-6 signaling protects the host against infection. (A) Inhibition of IL-6 signaling rendered larvae more susceptible to S. epidermidis infection. Larvae (20 per well) infected with DMSO–Inf-low-6h were not affected by S. epidermidis. Inh W–Inf-low-6h larvae, however, showed significantly higher mortalities. The same trend was observed when comparing mortality profiles of DMSO–Inf-med-24h and Inh W–Inf-med-24h. (B) Inh W–Inf-med-24h did not show rapid elevation of il-1b levels, unlike their DMSO counterparts (DMSO–Inf-med-24h). (C and D) qPCR showed that although dre-miR-142a-3p (C) was unaffected by S. epidermidis infection, levels of dre-miR-142a-5p (D) were drastically downregulated. Inf-med-24h larvae showed significant downregulation both at 16 hpi and 24 hpi. (E) Levels of il-6st, a target of dre-miR-142a-5p, was significantly upregulated in Inf-med-24h at 16 and 24 hpi. (*p < 0.05, **p < 0.01, ****p < 0.0001, #p < 0.01). Data shown are representative of three independent experiments. (F) The 3'-UTR of il-6st is predicted to have two dre-miR-142a-5p binding sites: between 3188–3194 and 5611–5617.

FIGURE 7.

IL-6 signaling protects the host against infection. (A) Inhibition of IL-6 signaling rendered larvae more susceptible to S. epidermidis infection. Larvae (20 per well) infected with DMSO–Inf-low-6h were not affected by S. epidermidis. Inh W–Inf-low-6h larvae, however, showed significantly higher mortalities. The same trend was observed when comparing mortality profiles of DMSO–Inf-med-24h and Inh W–Inf-med-24h. (B) Inh W–Inf-med-24h did not show rapid elevation of il-1b levels, unlike their DMSO counterparts (DMSO–Inf-med-24h). (C and D) qPCR showed that although dre-miR-142a-3p (C) was unaffected by S. epidermidis infection, levels of dre-miR-142a-5p (D) were drastically downregulated. Inf-med-24h larvae showed significant downregulation both at 16 hpi and 24 hpi. (E) Levels of il-6st, a target of dre-miR-142a-5p, was significantly upregulated in Inf-med-24h at 16 and 24 hpi. (*p < 0.05, **p < 0.01, ****p < 0.0001, #p < 0.01). Data shown are representative of three independent experiments. (F) The 3'-UTR of il-6st is predicted to have two dre-miR-142a-5p binding sites: between 3188–3194 and 5611–5617.

Close modal

In order to substantiate this observation, we used the lowest, asymptomatic dose of infection, Inf-low-6h. As expected, larvae exposed to DMSO–Inf-low-6h did not show any significant mortality compared with the uninfected controls (DMSO–Control). However, when IL-6 signaling was inhibited (Inh W–Inf-low-6h), the same infection condition resulted in significant mortalities in larvae 54 hpi onward compared with their corresponding control population (Inh W–Control) (p = 0.02). At 80 hpi, Inh W–Inf-low-6h showed 65% survival (Fig. 7A, closed squares) as opposed to 92% and 98% in DMSO–Inf-low-6h and Inh W–Control, respectively (p = 3 × 10−6, d = 4.3) (Fig. 7A, open squares and closed triangles). The data showed that inhibition of IL-6 signaling caused an asymptomatic infection dose to become symptomatic in nature. Larvae maintained in C-188-9 corroborated the phenotype that was observed with WP-1066 (Supplemental Fig. 4B). Inhibition with C-188-9 resulted in higher mortalities with both the infection doses (Inf-med-24h and Inf-low-6h). This data demonstrated that IL-6 signaling is an important component of the mechanism of resistance of zebrafish larvae against S. epidermidis infection.

Enhanced IL-6 signaling during S. aureus infection has been shown to have a protective effect via upregulation of inflammatory cytokines such as IL-1α and TNF-α (37). Given the similar protective phenotype that we found in our model, we looked at the transcript levels of il-1b in the Inf-med-24h population that was maintained in WP-1066 (Inh W–Inf-med-24h) (Fig. 7B). Similar to our earlier observation, DMSO–Inf-med-24h larvae showed rapid and significant elevation of il-1b from 2 hpi (∼2-fold) onwards reaching a peak at 16 hpi (∼12-fold), and continuing to remain elevated at 24 hpi (∼3.5-fold) (p = 0.002; d = 4.5, 5.4, and 4.2 for 2, 16, and 24 hpi, respectively). The WP-1066–treated larvae, upon S. epidermidis infection (Inh W–Inf-med-24h), however, did not show any significant elevation in their il-1b levels at the 2-hpi and 16-hpi time points (Fig. 7B, closed circles). At 24 hpi, there was modest elevation of il-1b mRNA (∼2.5-fold) (p = 0.008, d = 5.4).

Given the importance of IL-6 signaling in the host immune response to S. epidermidis infection, we investigated the mechanism(s) of IL-6 regulation by miRNAs. miR-142 is one of the several miRNAs that has been shown to directly target various components of the IL-6 signaling pathway in humans and mouse (3840). Additionally, it has also been shown to impede clearance of S. aureus at wound sites, albeit by modulating small GTPases involved in neutrophil motility, and thereby altering neutrophil chemotaxis (41). Both the isoforms of miR-142 (dre-miR-142a-5p and dre-miR-142a-3p) are present in zebrafish and the latter has been shown to be necessary for neutrophil development (42). However, there is no evidence for their interaction with any gene of the IL-6 signaling pathway in zebrafish. To investigate if dre-miR-142 was involved in the host response to S. epidermidis infection, we studied the miRNA levels of both dre-miR-142a-3p (Fig. 7C) and dre-miR-142a-5p (Fig. 7D). There was no significant change in the levels of dre-miR-142a-3p at both 16 and 24 hpi, suggesting that this miRNA was not involved in regulating il-6 in this infection model. Levels of dre-miR-142a-5p, however, were significantly downregulated in Inf-med-24h larvae. At 16 hpi, it was 0.032-fold (∼−30-fold) in infected larvae compared with controls, whereas at 24 hpi, the levels further decreased to 0.012-fold (∼−83-fold) (p = 0.002; g = 6.89 and 3.11 at 16 and 24 hpi, respectively) (Fig. 7D). Given that there is no experimental or computational evidence of il-6 being directly targeted by miR-142-5p, we looked at other genes of the IL-6 signaling pathway, namely IL-6Rα and IL-6st. Although no predictions of a direct interaction were found between hsa-miR-142-5p and il-6ra in humans, a direct interaction between hsa-miR-142-5p and il-6st was predicted by the miRNA prediction software miRDB (43). Additionally, TargetScanFish (44) predicted two dre-miR-142a-5p target sites in the zebrafish il-6st sequence (Fig. 7F). Hence, we decided to look at the mRNA levels of il-6st (Fig. 7E). The il-6st levels of Inf-med-24h larvae showed a very tight inverse correlation with dre-miR-142a-5p levels at both the time points we checked. Compared with uninfected controls, at 16 hpi, il-6st showed ∼1.9-fold elevation (p = 0.044, g = 1.52) that increased to ∼2.6-fold by 24 hpi (p = 0.005, g = 4.76). Although in vitro assay showing direct binding between dre-miR-142a-5p and il-6st is required for confirmation, our data suggest that the upregulation of the IL-6 signaling pathway in this infection model could involve dre-miR-142a-5p. Under uninfected condition, dre-miR-142a-5p potentially interacts with and downregulates il-6st, keeping the pathway under check. S. epidermidis infection causes severe downregulation of dre-miR-142a-5p, thus enabling upregulation of the IL-6 signaling pathway.

S. epidermidis has recently emerged as an important opportunistic pathogen that causes frequent nosocomial infections. Owing to the natural occurrence of the bacteria as commensals on human skin, it is a common cause of acute sepsis in premature neonates and immunocompromised individuals (6). It is the single largest cause of infections in implants and indwelling medical devices (4, 45). Various implant models have been established to study infection as well as clearance of this bacterium (46, 47). Although these studies have demonstrated the conditions that promote infections, host immune responses to S. epidermidis have remained less well studied. Here the host immune response to a systemic infection of S. epidermidis has been analyzed using a zebrafish model.

Zebrafish has emerged as an effective model for studying host–pathogen infections in vivo (48, 49) and zebrafish models of S. aureus infection have shed light on various virulence mechanisms (50, 51). Veneman et al. (52) have recently established a high-throughput screening system where zebrafish embryos were injected with S. epidermidis. We have developed an S. epidermidis bath immersion infection model for zebrafish to elucidate the host immune responses. S. epidermidis recognition and bacterial clearance has been shown to be mediated by Tlr-2 in a mouse model (53). The zebrafish host response also showed elevation of tlr-2 transcription and NF-κB activation. There was marked inflammation with upregulation of various proinflammatory cytokines such as il-1b, tnf-a, and il-6. The infected larvae exhibited a rapid cellular response that was driven by macrophages. There was over 2-fold increase in macrophage numbers in the periphery within the first 2 h of infection, and this continued throughout the infection period and beyond indicating a role in clearance of infection. We further show the key role of IL-6 signaling in modulating inflammation and thus infection prognosis, and discover dre-miR-142a-5p as a potential candidate that may regulate this pathway.

The new noninvasive model of a bath immersion that we have established is experimentally highly amenable. As expected, compared with experiments that inject the bacterium into the embryo (52), a relatively higher inoculum was required to obtain reproducible infection phenotypes in this model. For most of our experiments we used a dose that induced marked phenotypes of infection and host immune responses (Inf-med-24h) to facilitate the analyses. However, it is worth noting that even the lowest dose (Inf-low-6h) that caused little morbidity and no mortality resulted in bacterial colonization with a concomitant measurable host immune response as indicated by mobilization of leukocytes. These suggest that S. epidermidis does infect the larvae at lower doses and a robust host immune response efficiently clears the infection. Hence this model is usable at lower MOIs for experiments that do not require overt morbidity and mortality phenotypes. The mechanisms of host immune response to nonpathogenic low doses could potentially differ from the inflammatory mechanisms that we have uncovered in our pathogenic model and may be more relevant to explaining the host–pathogen interface in asymptomatic commensal relationships. This aspect would be interesting to explore in further studies.

Zebrafish infection models are mainly established by static immersion or by injection. A recent study used global proteomic profiling to compare these two modes of infection and reported that static immersion resulted in enrichment of genes involved in the hypoxic pathway (32). Our data show mild induction of hypoxia in larval populations exposed to S. epidermidis for 24 h. Interestingly, the period of bacterial exposure (24 h as opposed to 6 h) was the determinant in inducing hypoxia, whereas bacterial inoculum had little effect (5 × 108 versus 1 × 109). Thus, shorter infection periods can be used to further limit the effect of hypoxia from this infection model. The signature of the proinflammatory cytokines in hypoxic larvae was, however, markedly different from the infected population. This indicated that in our model S. epidermidis infection was the main driver of the zebrafish immune response.

CFU counts from larvae treated with both Inf-low-24h and Inf-med-24h doses of bacteria were higher at 2 hpi, compared with later time points, although the animals continued to remain in the infection medium till 24 hpi. This indicates that after the initial colonization, rapid immune responses can efficiently act as a bulwark against further accretion of infections. Macrophages have been shown to play an important role in combatting S. aureus infection in murine, human, and zebrafish hosts (22, 50, 54). Similar to S. aureus infection, S. epidermidis infection in zebrafish resulted in the involvement of myeloid populations, predominantly the macrophages. Within 2 h postinfection we were able to measure a significant increase in the number of peripheral macrophages in both the doses we analyzed. In our experiments, neutrophils do not appear to play a major role in the innate response and macrophages act as the primary innate responder cells. The initial cellular response to the presence of S. epidermidis is found to be comparable in magnitude across all the doses that we have tried: a 2-fold increase in macrophage numbers in 2 h, in the surface volumes we measured. Macrophages probably have an important role in clearance and repair as their numbers thereafter lower gradually. This function is also suggested by the nearly 16 h that macrophages take after the bacterial clearance to return to control levels in the fish infected with Inf-low-6h. S. epidermidis strains that are unable to form biofilm are more susceptible to phagocytosis (55), whereas S. aureus and S. epidermidis biofilms can inhibit the phagocytic activity of macrophages (56, 57). Given that the macrophage numbers remain elevated while systemic bacterial counts are decreasing, it would be interesting to see if the bacteria on prolonged colonization activate any of the biofilm pathways.

Macrophage phagosomes recruit Tlr-2 that gets activated by Gram-positive bacteria and not Gram-negative (58). Tlr-2 is essential for neutrophil and macrophage recruitment post-S. aureus infection (59). In the context of S. epidermidis infection, Tlr-2 has been shown to be involved in recognition and clearance of the bacteria (53). S. epidermidis infection in this model also resulted in the upregulation of tlr-2. We also observed moderate elevation of tlr-4b levels. Although the role of tlr-4 in S. epidermidis infection is not known, in S. aureus infection of murine brain, whereas tlr-2 is central for activating downstream immune responses, tlr-4 is also important for leukocyte apoptosis and cytokine production (28). Degradation of S. aureus within phagosomes can oftentimes result in activation of Tlr-8 signaling in the endosomal compartment. In such a scenario, the IRF5 pathway is activated that eventually results in induction of IFN-β, a classical antiviral cytokine. Tlr-2 antagonizes the Tlr-8-IRF5 signaling axis (29). Given that tlr-2 was elevated and the levels of tlr-8a and tlr-8b were unaffected in the infected larvae, it is likely that S. epidermidis infection did not activate the Tlr-8-IRF5 signaling cascade.

The data further showed activation of the NF-κB signaling pathway in this infection model. Although convergence of Tlr signaling into NF-κB activation is well documented (60), NF-κB activation has also been reported to upregulate the Tlr-2 pathway (61). Thus our data cannot definitively say whether S. epidermidis activates NF-κB through Tlr-2 signaling or through another receptor or pathway. S. epidermidis lysates have been shown to activate NF-κB signaling in human and mouse epithelial cells (62). Staphylococcal infections lead to proinflammatory immune response in a myriad of host models (63, 64). In our model, all the NF-κB–responsive proinflammatory cytokines that we tested were elevated. The mRNA levels of il-1b, tnf-a, and il-6 showed stark increase. S. aureus infection has been shown to induce il-1b in murine brain abscess models (65, 66). Spiliopoulou et al. (57) showed that human cell lines infected with S. epidermidis induced IL-1β, and we report elevation of the same in an in vivo model. They further showed that induction of IL-1β, TNF-α, and IL-6 were unaffected by biofilm formation; but IL-12 elevation was dampened when S. epidermidis formed biofilms (57). In light of this observation, it will be interesting to investigate whether the profile of S. epidermidis infection in our model resembles biofilm or planktonic cultures. TNF-α is another cytokine that is upregulated upon S. aureus infection in mouse models (66, 67). S. epidermidis infection also induces this cytokine in human cell culture (57) and mouse and zebrafish models (52, 53). Our bath immersion model recapitulates this feature of infection. S. aureus and S. epidermidis infections have been shown to result in upregulation of IL-6 levels in mouse (53, 67). Neonatal sepsis owing to S. epidermidis infection show elevated levels of IL-6 (68). In zebrafish, S. epidermidis infection resulted in heightened il-6 levels, and we sought to look at its relevance and the mechanism(s) of its regulation.

IL-6 is a well-known pleiotropic cytokine, with pro- and anti-inflammatory effects based on trans or cis signaling cascades, respectively (12). Tlr-2 upregulation has been shown to induce IL-6 trans signaling in monocytes (69). IL-6 trans signaling has also been implicated in mediating inflammatory responses in various models of infection and endotoxin challenge (70). Although IL-6 signaling promoted a Tlr-4 dependent proinflammatory response in a murine endotoxin model (71), IL-6 signaling in Toxoplasma gondii had an anti-inflammatory effect (72), both of which were detrimental to animal survival. In our model, however, inhibition of IL-6 signaling using STAT-3 inhibitors rendered the zebrafish more susceptible to S. epidermidis infection–induced mortality with markedly reduced elevation of il-1b levels. This is reminiscent of S. aureus infection reports which showed that elevated IL-6 signaling led to higher survival of mice and augmentation of various cytokines such as IL-1α, IL-4, and TNF-α (37). Conversely, lack of IL-6 signaling in mice resulted in greater infection loads of S. aureus (73). It will be interesting to delve further into how IL-6 signaling might play a protective role against S. epidermidis and S. aureus infections. In this regard, it must be mentioned that although inhibition of STAT-3 has been shown to affect the IL-6 signaling pathway (74), given the pleiotropic nature of STAT-3 signaling, the effect of this inhibition on other cellular pathways cannot be completely excluded (75). Although upregulation of IL-6 during S. epidermidis infection has been shown ex vivo and in vivo, to the best of our knowledge, this is the first report of a protective role of IL-6 signaling in this infection.

IL-6 signaling is known to be regulated by miRNA-mediated modulation of different genes of this pathway, miR-142 being one of them (14). miR-142-3p has been reported to target il-6 in models of endotoxin challenge and age-related studies (38, 39). Regulation of il-6 by miR-142-3p in an infection setting has not been previously reported. However, miR-142-3p has been implicated in affecting phagocytosis upon S. aureus as well as mycobacterial infections (41, 76, 77). In our model dre-miR-142a-3p levels remained unaffected, indicating this miRNA is unlikely to be involved in regulating the immune response to S. epidermidis infection. The role of miR-142-5p, in contrast, has remained relatively obscure. It has been shown to directly target il-6st in THP-1 cells under conditions of temperature stress (40) and during adaptive hypertrophy in mice (78). Nothing has been shown about the role of this miRNA or the role of its interaction with il-6st in infections or in immune responses. Here, we show for the first time a likely involvement of miR-142-5p in regulating the IL-6 signaling pathway upon an infection. We see an inverse relationship between levels of dre-miR-142a-5p and of il-6st. Thus, it is possible that, similar to what has been reported in humans, in zebrafish also dre-miR-142a-5p regulates il-6st by directly targeting it. miRNA profiling in bovine models gives conflicting reports regarding the effect of S. aureus infection on bta-miR-142-5p expression (7981). Here we show drastic downregulation of dre-miR-142a-5p with concomitant upregulation of its target gene il-6st. Thus we propose that the host response to S. epidermidis infection is characterized by a NF-κB–dependent proinflammatory response and an upregulation of IL-6 signaling with the likely involvement of dre-miR-142a-5p.

In conclusion, we have established a bath immersion model of S. epidermidis in zebrafish and have elucidated several key features of the host immune response. The model is versatile in that it allows experimentation with low as well as high MOIs and can serve as a platform to address various aspects of host–pathogen interactions and drug screens. Given the ease of establishment of this model, host immune response to various features of S. epidermidis infection, such as biofilm formation or antibiotic resistance, can be investigated. S. epidermidis infection results in increased migration of leukocytes to the surface along with activation of NF-κB. The infection causes a strong proinflammatory response and sustained upregulation of theh IL-6 signaling pathway that has a protective effect. We also identify a possible role for dre-miR-142-5p in this infection model.

We thank Maria Leptin, Francesca Peri, and Graham Lieschke for the generous gifts of Tg(NFB:GFP, Luc), Tg(pu.1:RFP), and Tg(mpeg1:Kaede) reporter lines, respectively, Harapriya Mohapatra and Martina Rembold for critical comments on the manuscript, Ramanujam Srinivasan and Renjith Mathew for insightful discussions throughout the project, and the School of Biological Sciences, National Institute of Science Education and Research for departmental support.

This work was supported by Ramanujan Fellowship Grant SB/S2/RJN-174/2014, Department of Science and Technology Early Career Research Award Grant 2017-000528, intramural funds from the National Institute of Science Education and Research (an autonomous institute of the Department of Atomic Energy) (to S.B.), intramural funds from Institute of Life Sciences (an autonomous institute of the Department of Biotechnology), and Science and Engineering Research Board Grant EMR/2016/003780 (to R.K.S.).

P.T.D. carried out most of the experiments and analyzed the data. P.K.K. performed experiments and data analysis. R.K.S. collaborated on the project. S.B designed the project, analyzed the data, and wrote the manuscript.

The online version of this article contains supplemental material.

Abbreviations used in this article

dpf

day postfertilization

hpi

hour postinfection

MOI

multiplicity of infection

qPCR

quantitative PCR

RFP

red fluorescent protein

1.
Chessa
D.
,
G.
Ganau
,
V.
Mazzarello
.
2015
.
An overview of Staphylococcus epidermidis and Staphylococcus aureus with a focus on developing countries.
J. Infect. Dev. Ctries.
9
:
547
550
.
2.
Marsilio
F.
,
C. E.
Di Francesco
,
B.
Di Martino
.
2018
.
Coagulase-positive and coagulase-negative Staphylococci animal diseases.
In
Pet-To-Man Travelling Staphylococci.
V.
Savini
.
Academic Press
, p.
43
50
.
3.
Gordon
R. J.
,
F. D.
Lowy
.
2008
.
Pathogenesis of methicillin-resistant Staphylococcus aureus infection.
Clin. Infect. Dis.
46
(
S5
,
Suppl 5
)
S350
S359
.
4.
Otto
M
.
2009
.
Staphylococcus epidermidis--the ‘accidental’ pathogen.
Nat. Rev. Microbiol.
7
:
555
567
.
5.
Cherifi
S.
,
B.
Byl
,
A.
Deplano
,
C.
Nonhoff
,
O.
Denis
,
M.
Hallin
.
2013
.
Comparative epidemiology of Staphylococcus epidermidis isolates from patients with catheter-related bacteremia and from healthy volunteers.
J. Clin. Microbiol.
51
:
1541
1547
.
6.
Sabaté Brescó
M.
,
L. G.
Harris
,
K.
Thompson
,
B.
Stanic
,
M.
Morgenstern
,
L.
O’Mahony
,
R. G.
Richards
,
T. F.
Moriarty
.
2017
.
Pathogenic mechanisms and host interactions in Staphylococcus epidermidis device-related infection.
Front. Microbiol.
8
:
1401
.
7.
Morgenstern
M.
,
C.
Erichsen
,
S.
Hackl
,
J.
Mily
,
M.
Militz
,
J.
Friederichs
,
S.
Hungerer
,
V.
Bühren
,
T. F.
Moriarty
,
V.
Post
, et al
2016
.
Antibiotic resistance of commensal Staphylococcus aureus and coagulase-negative Staphylococci in an international cohort of surgeons: a prospective point-prevalence study.
PLoS One
11
:
e0148437
.
8.
Widerström
M
.
2016
.
Significance of Staphylococcus epidermidis in health care-associated infections, from contaminant to clinically relevant pathogen: this is a wake-up call!
J. Clin. Microbiol.
54
:
1679
1681
.
9.
Le
K. Y.
,
M. D.
Park
,
M.
Otto
.
2018
.
Immune evasion mechanisms of Staphylococcus epidermidis biofilm infection.
Front. Microbiol.
9
:
359
.
10.
Nguyen
T. H.
,
M. D.
Park
,
M.
Otto
.
2017
.
Host response to Staphylococcus epidermidis colonization and infections.
Front. Cell. Infect. Microbiol.
7
:
90
.
11.
Cheung
G. Y.
,
M.
Otto
.
2010
.
Understanding the significance of Staphylococcus epidermidis bacteremia in babies and children.
Curr. Opin. Infect. Dis.
23
:
208
216
.
12.
Su
H.
,
C. T.
Lei
,
C.
Zhang
.
2017
.
Interleukin-6 signaling pathway and its role in kidney disease: an update.
Front. Immunol.
8
:
405
.
13.
Chalaris
A.
,
C.
Garbers
,
B.
Rabe
,
S.
Rose-John
,
J.
Scheller
.
2011
.
The soluble interleukin 6 receptor: generation and role in inflammation and cancer.
Eur. J. Cell Biol.
90
:
484
494
.
14.
Servais
F. A.
,
M.
Kirchmeyer
,
M.
Hamdorf
,
N. W. E.
Minoungou
,
S.
Rose-John
,
S.
Kreis
,
C.
Haan
,
I.
Behrmann
.
2019
.
Modulation of the IL-6-signaling pathway in liver cells by miRNAs targeting gp130, JAK1, and/or STAT3.
Mol. Ther. Nucleic Acids
16
:
419
433
.
15.
Shrestha
A.
,
R. T.
Mukhametshina
,
S.
Taghizadeh
,
E.
Vásquez-Pacheco
,
H.
Cabrera-Fuentes
,
A.
Rizvanov
,
B.
Mari
,
G.
Carraro
,
S.
Bellusci
.
2017
.
MicroRNA-142 is a multifaceted regulator in organogenesis, homeostasis, and disease.
Dev. Dyn.
246
:
285
290
.
16.
Heilmann
C.
,
C.
Gerke
,
F.
Perdreau-Remington
,
F.
Götz
.
1996
.
Characterization of Tn917 insertion mutants of Staphylococcus epidermidis affected in biofilm formation.
Infect. Immun.
64
:
277
282
.
17.
Busk
P. K
.
2014
.
A tool for design of primers for microRNA-specific quantitative RT-qPCR.
BMC Bioinformatics
15
:
29
.
18.
Schindelin
J.
,
I.
Arganda-Carreras
,
E.
Frise
,
V.
Kaynig
,
M.
Longair
,
T.
Pietzsch
,
S.
Preibisch
,
C.
Rueden
,
S.
Saalfeld
,
B.
Schmid
, et al
2012
.
Fiji: an open-source platform for biological-image analysis.
Nat. Methods
9
:
676
682
.
19.
Kuri
P.
,
K.
Ellwanger
,
T. A.
Kufer
,
M.
Leptin
,
B.
Bajoghli
.
2017
.
A high-sensitivity bi-directional reporter to monitor NF-κB activity in cell culture and zebrafish in real time.
J. Cell Sci.
130
:
648
657
.
20.
Wasserstein
R. L.
,
N. A.
Lazar
.
2016
.
The ASA statement on p values: context, process, and purpose.
Am. Stat.
70
:
129
133
.
21.
Sullivan
G. M.
,
R.
Feinn
.
2012
.
Using effect size-or why the P value is not enough.
J. Grad. Med. Educ.
4
:
279
282
.
22.
Verdrengh
M.
,
A.
Tarkowski
.
2000
.
Role of macrophages in Staphylococcus aureus-induced arthritis and sepsis.
Arthritis Rheum.
43
:
2276
2282
.
23.
Rigby
K. M.
,
F. R.
DeLeo
.
2012
.
Neutrophils in innate host defense against Staphylococcus aureus infections.
Semin. Immunopathol.
34
:
237
259
.
24.
Rhodes
J.
,
A.
Hagen
,
K.
Hsu
,
M.
Deng
,
T. X.
Liu
,
A. T.
Look
,
J. P.
Kanki
.
2005
.
Interplay of pu.1 and gata1 determines myelo-erythroid progenitor cell fate in zebrafish.
Dev. Cell
8
:
97
108
.
25.
Ellett
F.
,
L.
Pase
,
J. W.
Hayman
,
A.
Andrianopoulos
,
G. J.
Lieschke
.
2011
.
mpeg1 promoter transgenes direct macrophage-lineage expression in zebrafish.
Blood
117
:
e49
e56
.
26.
Willis
A. R.
,
V.
Torraca
,
M. C.
Gomes
,
J.
Shelley
,
M.
Mazon-Moya
,
A.
Filloux
,
C.
Lo Celso
,
S.
Mostowy
.
2018
.
Shigella-induced emergency granulopoiesis protects zebrafish larvae from secondary infection.
MBio
9
:
e00933-18
.
27.
Takeuchi
O.
,
K.
Hoshino
,
T.
Kawai
,
H.
Sanjo
,
H.
Takada
,
T.
Ogawa
,
K.
Takeda
,
S.
Akira
.
1999
.
Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cell wall components.
Immunity
11
:
443
451
.
28.
Stenzel
W.
,
S.
Soltek
,
M.
Sanchez-Ruiz
,
S.
Akira
,
H.
Miletic
,
D.
Schlüter
,
M.
Deckert
.
2008
.
Both TLR2 and TLR4 are required for the effective immune response in Staphylococcus aureus-induced experimental murine brain abscess.
Am. J. Pathol.
172
:
132
145
.
29.
Bergstrøm
B.
,
M. H.
Aune
,
J. A.
Awuh
,
J. F.
Kojen
,
K. J.
Blix
,
L.
Ryan
,
T. H.
Flo
,
T. E.
Mollnes
,
T.
Espevik
,
J.
Stenvik
.
2015
.
TLR8 senses Staphylococcus aureus RNA in human primary monocytes and macrophages and induces IFN-β production via a TAK1-IKKβ-IRF5 signaling pathway.
J. Immunol.
195
:
1100
1111
.
30.
Ivashkiv
L. B
.
2011
.
Inflammatory signaling in macrophages: transitions from acute to tolerant and alternative activation states.
Eur. J. Immunol.
41
:
2477
2481
.
31.
Dorrington
M. G.
,
I. D. C.
Fraser
.
2019
.
NF-κB signaling in macrophages: dynamics, crosstalk, and signal integration.
Front. Immunol.
10
:
705
.
32.
Díaz-Pascual
F.
,
J.
Ortíz-Severín
,
M. A.
Varas
,
M. L.
Allende
,
F. P.
Chávez
.
2017
.
In vivo host-pathogen interaction as revealed by global proteomic profiling of zebrafish larvae.
Front. Cell. Infect. Microbiol.
7
:
334
.
33.
Ton
C.
,
D.
Stamatiou
,
C. C.
Liew
.
2003
.
Gene expression profile of zebrafish exposed to hypoxia during development.
Physiol. Genomics
13
:
97
106
.
34.
Elks
P. M.
,
S. A.
Renshaw
,
A. H.
Meijer
,
S. R.
Walmsley
,
F. J.
van Eeden
.
2015
.
Exploring the HIFs, buts and maybes of hypoxia signalling in disease: lessons from zebrafish models.
Dis. Model. Mech.
8
:
1349
1360
.
35.
Horiguchi
A.
,
T.
Asano
,
K.
Kuroda
,
A.
Sato
,
J.
Asakuma
,
K.
Ito
,
M.
Hayakawa
,
M.
Sumitomo
,
T.
Asano
.
2010
.
STAT3 inhibitor WP1066 as a novel therapeutic agent for renal cell carcinoma.
Br. J. Cancer
102
:
1592
1599
.
36.
Jung
K. H.
,
W.
Yoo
,
H. L.
Stevenson
,
D.
Deshpande
,
H.
Shen
,
M.
Gagea
,
S. Y.
Yoo
,
J.
Wang
,
T. K.
Eckols
,
U.
Bharadwaj
, et al
2017
.
Multifunctional effects of a small-molecule STAT3 inhibitor on NASH and hepatocellular carcinoma in mice.
Clin. Cancer Res.
23
:
5537
5546
.
37.
Onogawa
T
.
2005
.
Local delivery of soluble interleukin-6 receptors to improve the outcome of alpha-toxin producing Staphylococcus aureus infection in mice.
Immunobiology
209
:
651
660
.
38.
Liu
Y.
,
X.
Song
,
S.
Meng
,
M.
Jiang
.
2016
.
Downregulated expression of miR-142-3p in macrophages contributes to increased IL-6 levels in aged mice.
Mol. Immunol.
80
:
11
16
.
39.
Sun
Y.
,
S.
Varambally
,
C. A.
Maher
,
Q.
Cao
,
P.
Chockley
,
T.
Toubai
,
C.
Malter
,
E.
Nieves
,
I.
Tawara
,
Y.
Wang
, et al
2011
.
Targeting of microRNA-142-3p in dendritic cells regulates endotoxin-induced mortality.
Blood
117
:
6172
6183
.
40.
Wong
J. J.
,
A. Y.
Au
,
D.
Gao
,
N.
Pinello
,
C. T.
Kwok
,
A.
Thoeng
,
K. A.
Lau
,
J. E.
Gordon
,
U.
Schmitz
,
Y.
Feng
, et al
2016
.
RBM3 regulates temperature sensitive miR-142-5p and miR-143 (thermomiRs), which target immune genes and control fever.
Nucleic Acids Res.
44
:
2888
2897
.
41.
Tanaka
K.
,
S. E.
Kim
,
H.
Yano
,
G.
Matsumoto
,
R.
Ohuchida
,
Y.
Ishikura
,
M.
Araki
,
K.
Araki
,
S.
Park
,
T.
Komatsu
, et al
2017
.
MiR-142 is required for Staphylococcus aureus clearance at skin wound sites via small GTPase-mediated regulation of the neutrophil actin cytoskeleton.
J. Invest. Dermatol.
137
:
931
940
.
42.
Fan
H. B.
,
Y. J.
Liu
,
L.
Wang
,
T. T.
Du
,
M.
Dong
,
L.
Gao
,
Z. Z.
Meng
,
Y.
Jin
,
Y.
Chen
,
M.
Deng
, et al
2014
.
miR-142-3p acts as an essential modulator of neutrophil development in zebrafish.
Blood
124
:
1320
1330
.
43.
Chen
Y.
,
X.
Wang
.
2020
.
miRDB: an online database for prediction of functional microRNA targets.
Nucleic Acids Res.
48
(
D1
):
D127
D131
.
44.
Ulitsky
I.
,
A.
Shkumatava
,
C. H.
Jan
,
A. O.
Subtelny
,
D.
Koppstein
,
G. W.
Bell
,
H.
Sive
,
D. P.
Bartel
.
2012
.
Extensive alternative polyadenylation during zebrafish development.
Genome Res.
22
:
2054
2066
.
45.
Uçkay
I.
,
D.
Pittet
,
P.
Vaudaux
,
H.
Sax
,
D.
Lew
,
F.
Waldvogel
.
2009
.
Foreign body infections due to Staphylococcus epidermidis.
Ann. Med.
41
:
109
119
.
46.
Walker
J. N.
,
L. H.
Poppler
,
C. L.
Pinkner
,
S. J.
Hultgren
,
T. M.
Myckatyn
.
2020
.
Establishment and characterization of bacterial infection of breast implants in a murine model.
Aesthet. Surg. J.
40
:
516
528
.
47.
Sabaté Brescó
M.
,
L.
O’Mahony
,
S.
Zeiter
,
K.
Kluge
,
M.
Ziegler
,
C.
Berset
,
D.
Nehrbass
,
R. G.
Richards
,
T. F.
Moriarty
.
2017
.
Influence of fracture stability on Staphylococcus epidermidis and Staphylococcus aureus infection in a murine femoral fracture model.
Eur. Cell. Mater.
34
:
321
340
.
48.
Gomes
M. C.
,
S.
Mostowy
.
2020
.
The case for modeling human infection in zebrafish.
Trends Microbiol.
28
:
10
18
.
49.
Renshaw
S. A.
,
N. S.
Trede
.
2012
.
A model 450 million years in the making: zebrafish and vertebrate immunity.
Dis. Model. Mech.
5
:
38
47
.
50.
Prajsnar
T. K.
,
V. T.
Cunliffe
,
S. J.
Foster
,
S. A.
Renshaw
.
2008
.
A novel vertebrate model of Staphylococcus aureus infection reveals phagocyte-dependent resistance of zebrafish to non-host specialized pathogens.
Cell. Microbiol.
10
:
2312
2325
.
51.
Prajsnar
T. K.
,
R.
Hamilton
,
J.
Garcia-Lara
,
G.
McVicker
,
A.
Williams
,
M.
Boots
,
S. J.
Foster
,
S. A.
Renshaw
.
2012
.
A privileged intraphagocyte niche is responsible for disseminated infection of Staphylococcus aureus in a zebrafish model.
Cell. Microbiol.
14
:
1600
1619
.
52.
Veneman
W. J.
,
O. W.
Stockhammer
,
L.
de Boer
,
S. A.
Zaat
,
A. H.
Meijer
,
H. P.
Spaink
.
2013
.
A zebrafish high throughput screening system used for Staphylococcus epidermidis infection marker discovery.
BMC Genomics
14
:
255
.
53.
Strunk
T.
,
M. R.
Power Coombs
,
A. J.
Currie
,
P.
Richmond
,
D. T.
Golenbock
,
L.
Stoler-Barak
,
L. C.
Gallington
,
M.
Otto
,
D.
Burgner
,
O.
Levy
.
2010
.
TLR2 mediates recognition of live Staphylococcus epidermidis and clearance of bacteremia.
PLoS One
5
:
e10111
.
54.
Kubica
M.
,
K.
Guzik
,
J.
Koziel
,
M.
Zarebski
,
W.
Richter
,
B.
Gajkowska
,
A.
Golda
,
A.
Maciag-Gudowska
,
K.
Brix
,
L.
Shaw
, et al
2008
.
A potential new pathway for Staphylococcus aureus dissemination: the silent survival of S. aureus phagocytosed by human monocyte-derived macrophages.
PLoS One
3
:
e1409
.
55.
Shiro
H.
,
E.
Muller
,
N.
Gutierrez
,
S.
Boisot
,
M.
Grout
,
T. D.
Tosteson
,
D.
Goldmann
,
G. B.
Pier
.
1994
.
Transposon mutants of Staphylococcus epidermidis deficient in elaboration of capsular polysaccharide/adhesin and slime are avirulent in a rabbit model of endocarditis.
J. Infect. Dis.
169
:
1042
1049
.
56.
Thurlow
L. R.
,
M. L.
Hanke
,
T.
Fritz
,
A.
Angle
,
A.
Aldrich
,
S. H.
Williams
,
I. L.
Engebretsen
,
K. W.
Bayles
,
A. R.
Horswill
,
T.
Kielian
.
2011
.
Staphylococcus aureus biofilms prevent macrophage phagocytosis and attenuate inflammation in vivo.
J. Immunol.
186
:
6585
6596
.
57.
Spiliopoulou
A. I.
,
F.
Kolonitsiou
,
M. I.
Krevvata
,
M.
Leontsinidis
,
T. S.
Wilkinson
,
D.
Mack
,
E. D.
Anastassiou
.
2012
.
Bacterial adhesion, intracellular survival and cytokine induction upon stimulation of mononuclear cells with planktonic or biofilm phase Staphylococcus epidermidis.
FEMS Microbiol. Lett.
330
:
56
65
.
58.
Underhill
D. M.
,
A.
Ozinsky
,
A. M.
Hajjar
,
A.
Stevens
,
C. B.
Wilson
,
M.
Bassetti
,
A.
Aderem
.
1999
.
The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens.
Nature
401
:
811
815
.
59.
Fournier
B
.
2013
.
The function of TLR2 during staphylococcal diseases.
Front. Cell. Infect. Microbiol.
2
:
167
.
60.
Kawai
T.
,
S.
Akira
.
2007
.
Signaling to NF-kappaB by Toll-like receptors.
Trends Mol. Med.
13
:
460
469
.
61.
Wang
T.
,
W. P.
Lafuse
,
B. S.
Zwilling
.
2001
.
NFkappaB and Sp1 elements are necessary for maximal transcription of Toll-like receptor 2 induced by Mycobacterium avium.
J. Immunol.
167
:
6924
6932
.
62.
Val
S.
,
H.
Mubeen
,
A.
Tomney
,
S.
Chen
,
D.
Preciado
.
2015
.
Impact of Staphylococcus epidermidis lysates on middle ear epithelial proinflammatory and mucogenic response.
J. Investig. Med.
63
:
258
266
.
63.
Fournier
B.
,
D. J.
Philpott
.
2005
.
Recognition of Staphylococcus aureus by the innate immune system.
Clin. Microbiol. Rev.
18
:
521
540
.
64.
Askarian
F.
,
T.
Wagner
,
M.
Johannessen
,
V.
Nizet
.
2018
.
Staphylococcus aureus modulation of innate immune responses through Toll-like (TLR), (NOD)-like (NLR) and C-type lectin (CLR) receptors.
FEMS Microbiol. Rev.
42
:
656
671
.
65.
Cho
J. S.
,
Y.
Guo
,
R. I.
Ramos
,
F.
Hebroni
,
S. B.
Plaisier
,
C.
Xuan
,
J. L.
Granick
,
H.
Matsushima
,
A.
Takashima
,
Y.
Iwakura
, et al
2012
.
Neutrophil-derived IL-1β is sufficient for abscess formation in immunity against Staphylococcus aureus in mice.
PLoS Pathog.
8
:
e1003047
.
66.
Kielian
T.
,
E. D.
Bearden
,
A. C.
Baldwin
,
N.
Esen
.
2004
.
IL-1 and TNF-alpha play a pivotal role in the host immune response in a mouse model of Staphylococcus aureus-induced experimental brain abscess.
J. Neuropathol. Exp. Neurol.
63
:
381
396
.
67.
Yimin
M.
,
M.
Kohanawa
,
S.
Zhao
,
M.
Ozaki
,
S.
Haga
,
G.
Nan
,
Y.
Kuge
,
N.
Tamaki
.
2013
.
Contribution of Toll-like receptor 2 to the innate response against Staphylococcus aureus infection in mice.
PLoS One
8
:
e74287
.
68.
Tröger
B.
,
M.
Heidemann
,
I.
Osthues
,
D.
Knaack
,
W.
Göpel
,
E.
Herting
,
J. K.
Knobloch
,
C.
Härtel
.
2020
.
Modulation of S. epidermidis-induced innate immune responses in neonatal whole blood.
J. Microbiol. Immunol. Infect.
53
:
240
249
.
69.
Flynn
C. M.
,
Y.
Garbers
,
J.
Lokau
,
D.
Wesch
,
D. M.
Schulte
,
M.
Laudes
,
W.
Lieb
,
S.
Aparicio-Siegmund
,
C.
Garbers
.
2019
.
Activation of Toll-like receptor 2 (TLR2) induces interleukin-6 trans-signaling.
Sci. Rep.
9
:
7306
.
70.
Hunter
C. A.
,
S. A.
Jones
.
2015
.
IL-6 as a keystone cytokine in health and disease. [Published erratum appears in 2017 Nat. Immunol. 18: 1271.].
Nat. Immunol.
16
:
448
457
.
71.
Greenhill
C. J.
,
S.
Rose-John
,
R.
Lissilaa
,
W.
Ferlin
,
M.
Ernst
,
P. J.
Hertzog
,
A.
Mansell
,
B. J.
Jenkins
.
2011
.
IL-6 trans-signaling modulates TLR4-dependent inflammatory responses via STAT3.
J. Immunol.
186
:
1199
1208
.
72.
Silver
J. S.
,
J. S.
Stumhofer
,
S.
Passos
,
M.
Ernst
,
C. A.
Hunter
.
2011
.
IL-6 mediates the susceptibility of glycoprotein 130 hypermorphs to Toxoplasma gondii.
J. Immunol.
187
:
350
360
.
73.
Hume
E. B.
,
N.
Cole
,
L. L.
Garthwaite
,
S.
Khan
,
M. D.
Willcox
.
2006
.
A protective role for IL-6 in staphylococcal microbial keratitis.
Invest. Ophthalmol. Vis. Sci.
47
:
4926
4930
.
74.
Mori
T.
,
T.
Miyamoto
,
H.
Yoshida
,
M.
Asakawa
,
M.
Kawasumi
,
T.
Kobayashi
,
H.
Morioka
,
K.
Chiba
,
Y.
Toyama
,
A.
Yoshimura
.
2011
.
IL-1β and TNFα-initiated IL-6-STAT3 pathway is critical in mediating inflammatory cytokines and RANKL expression in inflammatory arthritis.
Int. Immunol.
23
:
701
712
.
75.
Hillmer
E. J.
,
H.
Zhang
,
H. S.
Li
,
S. S.
Watowich
.
2016
.
STAT3 signaling in immunity.
Cytokine Growth Factor Rev.
31
:
1
15
.
76.
Bettencourt
P.
,
S.
Marion
,
D.
Pires
,
L. F.
Santos
,
C.
Lastrucci
,
N.
Carmo
,
J.
Blake
,
V.
Benes
,
G.
Griffiths
,
O.
Neyrolles
, et al
2013
.
Actin-binding protein regulation by microRNAs as a novel microbial strategy to modulate phagocytosis by host cells: the case of N-Wasp and miR-142-3p.
Front. Cell. Infect. Microbiol.
3
:
19
.
77.
Xu
G.
,
Z.
Zhang
,
J.
Wei
,
Y.
Zhang
,
Y.
Zhang
,
L.
Guo
,
X.
Liu
.
2013
.
microR-142-3p down-regulates IRAK-1 in response to Mycobacterium bovis BCG infection in macrophages.
Tuberculosis (Edinb.)
93
:
606
611
.
78.
Sharma
S.
,
J.
Liu
,
J.
Wei
,
H.
Yuan
,
T.
Zhang
,
N. H.
Bishopric
.
2012
.
Repression of miR-142 by p300 and MAPK is required for survival signalling via gp130 during adaptive hypertrophy.
EMBO Mol. Med.
4
:
617
632
.
79.
Sun
J.
,
K.
Aswath
,
S. G.
Schroeder
,
J. D.
Lippolis
,
T. A.
Reinhardt
,
T. S.
Sonstegard
.
2015
.
MicroRNA expression profiles of bovine milk exosomes in response to Staphylococcus aureus infection.
BMC Genomics
16
:
806
.
80.
Luoreng
Z. M.
,
X. P.
Wang
,
C. G.
Mei
,
L. S.
Zan
.
2018
.
Comparison of microRNA profiles between bovine mammary glands infected with Staphylococcus aureus and Escherichia coli.
Int. J. Biol. Sci.
14
:
87
99
.
81.
Jin
W.
,
E. M.
Ibeagha-Awemu
,
G.
Liang
,
F.
Beaudoin
,
X.
Zhao
,
L.
Guan
.
2014
.
Transcriptome microRNA profiling of bovine mammary epithelial cells challenged with Escherichia coli or Staphylococcus aureus bacteria reveals pathogen directed microRNA expression profiles.
BMC Genomics
15
:
181
.

The authors have no financial conflicts of interest.

Supplementary data