Neutrophil-Associated Responses to Vibrio cholerae Infection in a Natural Host Model

ABSTRACT

Vibrio cholerae, the cause of human cholera, is an aquatic bacterium found in association with a variety of animals in the environment, including many teleost fish species. V. cholerae infection induces a proinflammatory response followed by a noninflammatory convalescent phase. Neutrophils are integral to this early immune response. However, the relationship between the neutrophil-associated protein calprotectin and V. cholerae has not been investigated, nor have the effects of limiting transition metals on V. cholerae growth. Zebrafish are useful as a natural V. cholerae model as the entire infectious cycle can be recapitulated in the presence of an intact intestinal microbiome and mature immune responses. Here, we demonstrate that zebrafish produce a significant neutrophil, interleukin 8 (IL-8), and calprotectin response following V. cholerae infection. Bacterial growth was completely inhibited by purified calprotectin protein or the chemical chelator N,N,N′,N′-tetrakis(2-pyridinylmethyl)-1,2-ethanediamine (TPEN), but growth was recovered by the addition of the transition metals zinc and manganese. The expression of downstream calprotectin targets was also significantly increased in the zebrafish. These findings illuminate the role of host calprotectin in combating V. cholerae infection. Inhibition of V. cholerae growth through metal limitation may provide new approaches in the development of anti-V. cholerae therapeutics. This study also establishes a major role for calprotectin in combating infectious diseases in zebrafish.

INTRODUCTION

Cholera is a life-threatening severe diarrheal disease caused by the Gram-negative bacterium Vibrio cholerae. V. cholerae is common in coastal regions and fresh, brackish, or salt water, and cholera is endemic in many developing countries (1). These areas often have poor sanitation and infrastructure, aiding in the transmission of the bacteria via the ingestion of contaminated food or water through the fecal-oral route. While there are hundreds of V. cholerae serogroups in the aquatic environment, only the O1 and O139 serogroups cause pandemic cholera (2). The O1 serogroup can be further divided into the classical and El Tor biotypes. While classical strains caused the first six pandemics, El Tor strains have caused the current pandemic, ongoing since 1961 (3).

Due to the extracellular nature of V. cholerae infection, cholera is generally considered to be a noninflammatory disease, although in reality, some inflammation does occur. During early infection, acute cholera results in a proinflammatory response characterized by increases in the levels of inflammatory cytokines, receptors, and white blood cells (4, 5). Following this acute response is a noninflammatory convalescent phase, characterized by the suppression of inflammatory markers and increases in vibriocidal IgM and mucosal antibody IgA (6–8). Innate immunity has been shown to be upregulated during V. cholerae infection, with neutrophils being documented as being an essential cell type. In a neonatal mouse model, neutrophils were shown to be recruited to the site of infection and to enter the lumen (9), while infection in neutropenic mice led to bacterial spread to extraintestinal organs and decreased survival, suggesting that the role of neutrophils in V. cholerae infection may be limiting the infection to the intestine (10). One major factor in neutrophil recruitment is the cytokine interleukin 8 (IL-8) (11). Known as neutrophil chemotactic factor, IL-8 functions to induce the chemotaxis of neutrophils to the site of infection and stimulates phagocytosis. During cholera infection, intestinal epithelial cells have been shown to secrete IL-8 in response to V. cholerae flagellins as well as the outer membrane protein OmpU (11, 12). In a zebrafish injury model, IL-8 was upregulated in response to acute inflammatory stimuli and was crucial for normal neutrophil recruitment to the site of injury and subsequent inflammation resolution (13).

One antimicrobial protein released by neutrophils is the heterodimer of EF hand proteins S100A8 and S100A9 termed calprotectin (CP). Released in neutrophil extracellular traps (NETs) as well as during cellular death, CP has been shown to make up 60% of the soluble protein in the cytosol of neutrophils (14). S100 proteins function as intracellular Ca²⁺ sensors, but when exported from the cytosol, CP functions to sequester transition metals, including zinc, manganese, and copper (15–17). Thus, CP has antimicrobial activity by limiting access to these transition metals, a mechanism known as nutritional immunity (18). During Candida albicans infection, CP has been shown to be the major antifungal component of NETs (19) and can directly inhibit the growth of several Gram-negative and Gram-positive bacteria (20). Increases in intestinal CP are not only seen in infectious diseases, as this protein is abundant in several inflammatory intestinal conditions, such as celiac disease (21), food allergies (22), and inflammatory bowel disease (IBD) (23). Many cases of these disorders can be diagnosed by detecting increases in fecal calprotectin levels. CP can also have downstream effects leading to the activation of a positive-feedback loop of inflammation. CP can act as a cytokine and chemokine, can activate anti- and proinflammatory responses, and can also bind to receptors such as Toll-like receptor 4 (TLR4) to activate a signaling cascade of inflammation through the transcription factor NF-κB (24, 25).

Zebrafish have previously been shown to be an effective model system for the study of V. cholerae infection, having many advantages over more commonly used mammalian animal models (3, 26–28). Among these advantages is that teleosts are a natural reservoir for V. cholerae species (29), and the model is relatively inexpensive, is easy to use, and recapitulates the characteristic diarrheal symptoms and transmission dynamics seen in humans (28). Colonization in the fish model also allows the assessment of the interaction between V. cholerae and the mature intact microbiota of the zebrafish (26). Furthermore, the zebrafish model remains relevant in studies examining the immune response due to the evolutionarily conserved nature of the immune system, allowing direct parallels between the immune responses of fish and humans (30).

While the zebrafish model has proven useful in the study of V. cholerae, no research has been published to date describing the zebrafish immune response to cholera infection, and our knowledge of the immune response of zebrafish and teleosts is still rudimentary and evolving (30). Moreover, our understanding of the immune response to V. cholerae continues to progress but remains far from complete. It will likely require innovative thinking and techniques to tease out the intricacies of this relationship. The link between the nutritional immunity protein CP and V. cholerae was not established until recently. To date, only one study examining adult cholera patients identified a 3.6-fold increase in S100A8 in lamina propria cells during acute-stage cholera; no other work has been published that further investigates this relationship (31).

In this study, using zebrafish as a natural V. cholerae host model, we investigated the importance of neutrophils and the cytokine IL-8 during V. cholerae infection and for the first time explore the relationship between the nutritional immunity protein CP and V. cholerae. Our results suggest that cholera induces a significant CP response and that this protein and its ability to limit transition metals to inhibit bacterial growth may be an area of interest for the development of novel anti-V. cholerae therapeutics.

RESULTS

In vivo neutrophil response.

Due to the extracellular nature of V. cholerae infection, we hypothesized that the innate immune system would respond early, with neutrophil recruitment into the lumen of the gut that peaked at roughly 24 h, as has been previously reported in the intestine (32). To assess the zebrafish immune response to V. cholerae, we inoculated fish via immersion using 2.5 × 10⁷ CFU of either pandemic O1 V. cholerae strain E7946 (labeled “El Tor”) or environmental V. cholerae strain 25493 (labeled “non-O1”). Fish were then incubated and sacrificed at 24 h, 72 h, or 120 h postinfection. After euthanasia, RNA was isolated from fish intestines, and quantitative PCR (qPCR) was used to assess neutrophil (mpx) and IL-8 responses. In the El Tor-infected fish, neutrophil gene expression was significantly increased at 24 h and 72 h, while it was significantly decreased at the 120-h time point (Fig. 1A). In the non-O1 strain-infected fish, only fish at 24 h had significantly increased levels of neutrophils compared to the control (Fig. 1B). The cytokine IL-8, which induces neutrophil chemotaxis and phagocytosis (13), was also significantly increased at all three time points in fish infected with either of the strains (Fig. 1C and D).

FIG 1.

mpx (neutrophil marker) and IL-8 gene expression levels increase during V. cholerae infection in the zebrafish gut. WT zebrafish were infected with the E7946 (El Tor) or 25493 (non-O1) strain of V. cholerae at 2.5 × 10⁷ CFU/mL and then sacrificed at the indicated time points. The zebrafish gut was taken, and mRNA was isolated from the gut. Gene expression levels of the indicated genes were determined by qRT-PCR. Gene expression was normalized against β-actin and expressed as a fold change. (A) mpx expression in El Tor strain E7946. (B) mpx expression in non-O1 strain 25493. (C) IL-8 expression in El Tor strain E7946. (D) IL-8 expression in non-O1 strain 25493. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. The data shown are from three independent experiments. *, P < 0.05 compared to the control; **, P < 0.05 compared to 24-h infection.

Transgenic (mpx:dendra) zebrafish were also used to image neutrophil recruitment into the lumen of the gut at 24-h time points via fluorescence microscopy. A primary polyclonal antibody directed against V. cholerae along with a secondary monoclonal antibody with a fluorescent tag (Alexa Fluor 647) were used. These images show red fluorescent bacteria forming biofilms along villus projections of the intestinal epithelial cells (nuclei stained blue by DAPI [4′,6-diamidino-2-phenylindole]) of the zebrafish, which were previously reported to be the sites of V. cholerae biofilm formation in both mammals and zebrafish (28, 33). Within these areas of red fluorescence, individual green fluorescent neutrophils can be seen infiltrating bacterial biofilms, which can be seen in both El Tor- and non-O1-infected fish (Fig. 2). The locations of infiltrated neutrophils are indicated by white arrows in Fig. 2.

FIG 2.

Fluorescence microscopy of the neutrophil response to V. cholerae infection of the transgenic (mpx:dendra) zebrafish intestinal epithelium. Fish were exposed to V. cholerae for 24 h and then sacrificed, fixed, and prepared for sectioning. Bacteria were visualized using a polyclonal primary antibody against V. cholerae and a secondary antibody carrying a red fluorescent (Alexa Fluor 647) tag. Blue fluorescence (DAPI) represents intestinal epithelial cell nuclei, red fluorescence (Alexa Fluor 647) represents V. cholerae bacteria, and green fluorescence (dendra) represents neutrophils. (A) Uninfected fish; (B) V. cholerae strain 25493-infected fish; (C) V. cholerae E7946-infected fish. White arrows point to green fluorescent neutrophils.

In vivo calprotectin response.

We next wanted to investigate the in vivo levels of neutrophil-associated antimicrobial proteins in the fish intestine. The expression levels of the genes encoding the nutritional immunity protein calprotectin were measured at the designated time points. Gene expression was significantly increased in fish infected with both V. cholerae strains at 24 h and 72 h but returned to basal levels by 120 h (Fig. 3A and B). Each time point was also significantly different from every other. We then wanted to determine if the in vivo calprotectin response was seen during infection with other V. cholerae serogroups and strains. For these experiments, fish were infected with O395 classical, N16961 El Tor, AM-19226 non-O1/O139, and 25493 non-O1/O139 strains and euthanized at 24 h. All the V. cholerae strains induced increased calprotectin gene expression in zebrafish gut extracts (Fig. 3C).

FIG 3.

Gene expression levels of S100A8/S100A9 (calprotectin) are increased during V. cholerae infection. WT zebrafish were infected with the E7946 (El Tor) or 25493 (non-O1/O139) strain of V. cholerae at 2.5 × 10⁷ CFU/mL and then sacrificed at the indicated time points. The zebrafish gut was taken, and mRNA was isolated from the gut. (A to C) Gene expression levels of S100A8/S100A9 were determined by qRT-PCR. Gene expression was normalized against β-actin and expressed as a fold change. (D and E) Calprotectin protein levels in zebrafish gut homogenates were determined by an ELISA. Levels are shown in nanograms per milliliter on the y axis. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. The data shown are from three experiments. *, P < 0.05 compared to the control; **, P < 0.05 compared to 24-h infection.

Next, we determined the intestinal concentrations of the calprotectin protein responses in the fish, as elevated fecal calprotectin levels are used in assays of a variety of disease states as an indicator of intestinal inflammation (21–23). For this experiment, we included a 6-h time point in addition to the significantly increased RNA time points of 24 h and 72 h. At the 6-h time point, CP was not significantly increased, but at the 24-h and 72-h time points, CP was significantly increased in fish infected with either of the V. cholerae strains, paralleling what was seen for mRNA levels (Fig. 3D and E).

In vitro effects of calprotectin protein on bacterial growth.

Although the direct inhibitory effects of calprotectin have been shown for several other bacteria (20), no investigation has been done to date exploring its effects on V. cholerae. Therefore, we added increasing concentrations of purified calprotectin protein to cultures of V. cholerae strain 25493 in Luria-Bertani (LB) broth medium, which proved to have dose-dependent growth inhibition. Complete growth inhibition was observed at 100 μg/mL (Fig. 4A). Besides taking the optical density (OD) of the culture to assess growth inhibition, the actual bacterial number was also measured by plating serial dilutions for each time point indicated. A proportional decrease in the CFU count was observed with decreasing OD values of the culture. This observation indicates that CP has vibriocidal activity.

FIG 4.

Zinc and manganese sequestration inhibits V. cholerae growth. V. cholerae strain 25493 was grown in LB broth medium with aeration at 37°C and measured at an OD₆₀₀ at the indicated time points. Increasing concentrations of WT (A), ΔSI (B), ΔSII (C), WT A8/A9 3H-3N (D), or ΔSI/SII (E) calprotectin were added to LB broth medium and incubated for 1 h prior to the addition of bacteria. The data shown are from three independent experiments.

To further dissect how calprotectin exerts its antimicrobial activity against V. cholerae, CP mutants with altered binding-site activity were used to explore the transition metal requirements for V. cholerae growth. The CP heterodimer of S100A8 and S100A9 has a pair of transition metal binding sites at the dimer interface (34). Unique site 1 (SI) is composed of six histidine (His) residues that have been shown to have high affinities for both Zn and Mn (20, 35). The canonical S100 protein site 2 (SII) is composed of a 3-His, 1-asparagine (Asp) S100 protein that has a high affinity for Zn but not Mn. CP variants used in this study include (i) a knockout of binding by site II (ΔSII), leaving binding site I available to chelate Zn and Mn; (ii) a knockout of binding site I (ΔSI), leaving binding site II available to chelate Zn but not Mn; (iii) a selective Mn knockout (WT [wild-type] A8/A9 3H-3N) that substitutes two key S100A9 His residues for Asn in S2, leaving a tetravalent 4-His site with normal stoichiometry and a high affinity for Zn; and (iv) a site I and site II double knockout (ΔSI/ΔSII) that lacks high affinities for transition metals, which serves as a negative control (20). Note that for accurate comparisons of growth inhibition, the concentrations of the ΔSI and ΔSII mutants were doubled relative to that of wild-type CP because each dimer could bind only one ion instead of two.

The time course and concentration dependence of the inhibition of the growth of V. cholerae by CP was first investigated for wild-type CP. Inhibition of growth was clearly evident at a CP concentration of 25 μg/mL and was complete at 100 μg/mL (Fig. 4A). No growth decreases were observed in the double-binding-site (site I/II) knockout (Fig. 4E), consistent with growth inhibition arising from the sequestration of transition metals. Assays with the selective Mn knockout mutant resulted in a decrease in bacterial growth by roughly one-third at the highest concentration, 100 μg/mL (Fig. 4D), indicating an important role for manganese in V. cholerae growth. This was confirmed in the experiments with CP ΔSI, which also cannot bind Mn, as the level of growth inhibition was about the same at the highest 100-μg/mL concentration (Fig. 4B). Interestingly, CP ΔSII inhibited bacterial growth by roughly two-thirds (Fig. 4C). This mutant retains the ability to bind Zn and Mn and differed noticeably from WT CP only at the three highest concentrations. This difference could arise from an error in the protein concentration or differences in the solubility of the mutant versus the wild type. Given that it can chelate both transition metals, we have been unable to conceive of other explanations for why this mutant is not ultimately as effective as the wild-type protein. Regardless of this apparent anomaly, these results show that the transition metals zinc and manganese both play important roles in the growth dynamics of V. cholerae and are essential for maximum growth potential.

Growth recovery in the presence of calprotectin proteins.

To confirm that the antimicrobial activity of calprotectin was due to the chelation of these transition metal ions, we then grew bacterial cultures in the presence of CP and CP variants, along with increasing concentrations of Zn, Mn, and Zn plus Mn in combination. Subinhibitory concentrations of wild-type CP at 50 μg/mL and 100 μg/mL of CP mutants were used. In the case of Zn addition, complete or nearly complete growth recovery could be seen for all three calprotectin proteins (Fig. 5A), as was the case for the Zn-plus-Mn additions (Fig. 5C). However, the Mn-only addition resulted in full growth recovery in only wild-type CP and ΔSII, but not the ΔSI mutant, as was expected (Fig. 5B). Overall, these results confirm that the antimicrobial activity of calprotectin in limiting V. cholerae growth is through the sequestration and chelation of zinc and manganese.

FIG 5.

Zinc and manganese are essential for the maximum growth potential of V. cholerae. V. cholerae strain 25493 was grown in LB broth medium with aeration at 37°C and measured at an OD₆₀₀ at the 24-h time point. Subinhibitory concentrations of WT (50 μg/mL), ΔSI (100 μg/mL), or ΔSII (100 μg/mL) calprotectin were added to LB broth medium and incubated for 1 h prior to the addition of bacteria. (A) Zn, (B) Mn, or (C) Zn plus Mn were added at increasing concentrations as indicated. The data shown are from three experiments and are presented as a percentage relative to bacterial growth in LB broth medium alone. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. *, P < 0.05 compared to no metal addition.

Is transition metal ion-associated antimicrobial activity specific to calprotectin?

We then asked whether the antimicrobial activity of transition metal chelation was specific to calprotectin or if this growth inhibition could be achieved via other metal-limiting methods. The high-affinity transition metal chemical chelator N,N,N′,N′-tetrakis(2-pyridinylmethyl)-1,2-ethanediamine (TPEN) was added directly to bacterial cultures at increasing concentrations, and the OD at 600 nm (OD₆₀₀) was read at 24 h. This treatment resulted in the complete inhibition of growth at high-enough concentrations (Fig. 6A). As with CP, we next asked whether this TPEN growth inhibition could be overcome by adding the transition metals Zn, Mn, and Zn plus Mn in combination back into bacterial cultures. Using subinhibitory concentrations of TPEN (31 μM), nearly complete growth levels were recovered using all three combinations of metals (Fig. 6B). This confirms that limiting the availability of the transition metals Zn and Mn has a direct growth inhibition effect on V. cholerae.

FIG 6.

TPEN sequestration of zinc and manganese inhibits V. cholerae growth. TPEN was added to LB broth and incubated for 1 h prior to the addition of bacteria. V. cholerae strain 25493 was grown in LB broth medium with aeration at 37°C and measured at an OD₆₀₀ at 24-h time points. (A) Increasing concentrations of TPEN were added to the culture. (B) A subinhibitory concentration of TPEN (31 μM) was added to LB broth medium with increasing concentrations of Zn, Mn, or Zn plus Mn. The data shown are from three experiments and are presented as a percentage relative to bacterial growth in LB broth medium alone. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. *, P < 0.05 compared to no metal addition.

Downstream effects of calprotectin in the zebrafish model.

In addition to having antimicrobial effects, calprotectin is also well known to have immunomodulatory effects. Known as a damage-associated molecular pattern (DAMP), CP, along with other members of the S100 family, can interact with pattern recognition receptors (PRRs) such as Toll-like receptors (TLRs) and RAGE (36). It can also initiate proinflammatory and anti-inflammatory responses and act as a chemokine and cytokine. As an endogenous agonist of TLR4, CP binding initiates a cascade of signaling through the NF-κB pathway, leading to a proinflammatory response and neutrophil recruitment (24, 25). Direct lipopolysaccharide (LPS) binding has also been shown to be a potent activator of inflammation through TLR4 (37). Furthermore, calprotectin can act as a direct chemoattractant for neutrophils, leading to adhesion at the site of infection (38). Ultimately, calprotectin contributes to a positive-feedback loop through proinflammatory signaling cascades, which leads to active inflammatory responses. The zebrafish model was used to see if these components of inflammation are upregulated in response to V. cholerae infection. Both El Tor and environmental strains induced significantly increased levels of TLR4 and NF-κB in the fish at all three time points (Fig. 7A to D). CD45 is a cell marker for all hematopoietic cells (blood cells), except for mature erythrocytes (red blood cells) and platelets. To measure the cellular immune response of zebrafish against V. cholerae infection, we measured CD45 gene expression and observed significant increases in response to infection with either of the V. cholerae strains at all three time points (Fig. 7E and F). This upregulation in CD45 gene expression suggests that V. cholerae induces increased leukocyte production in zebrafish. Although not specific to CP alone, these results indicate that molecules known to be associated with CP are in fact increased in zebrafish, likely in part due to the immunomodulatory effects of calprotectin.

FIG 7.

Gene expression levels of TLR4, NF-κB, and CD45 are increased during V. cholerae infection. WT zebrafish were infected with the E7946 (El Tor) (A, C, E) or 25493 (non-O1) (B, D, F) strain of V. cholerae at 2.5 × 10⁷ CFU/mL and then sacrificed at the indicated time points. mRNA was isolated from the zebrafish gut, and gene expression levels of TLR4 (A, B), NF-kB (C, D) or CD45 (E, F) were determined by qRT-PCR. Gene expression was normalized against β-actin and expressed as a fold change. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. The data shown are from three experiments. *, P < 0.05 compared to the control; **, P < 0.05 compared to 24-h infection.

IL-8 is essential for neutrophil recruitment.

Finally, the importance of IL-8 in neutrophil recruitment and subsequent calprotectin release during V. cholerae infection in zebrafish was assessed. Previous studies have demonstrated the importance of IL-8 in recruiting neutrophils to the site of injury in zebrafish (13). To achieve this, we used the selective CXCR2 (IL-8 receptor [IL-8R]) chemokine receptor antagonist SB 225002 for drug inhibition at a concentration of 5 μM. We also utilized a splice-blocking morpholino (MO) of cxcl8-l1 E1/I1 injected into single-cell-stage embryos as an alternative method (13). In the drug-treated plus infection group, IL-8 and neutrophil RNA levels were significantly reduced compared to those in the untreated plus infection group yet were still significantly increased compared to those in the control untreated no-infection group (Fig. 8D to F). Calprotectin levels in the drug-treated plus infection group were significantly reduced compared to those in the untreated plus infection group while not having a statistically significant difference in levels compared to those in the control untreated no-infection group. The partial decreases in IL-8 and neutrophil RNA levels may be due to an inability of SB 225002 to completely outcompete IL-8 binding to CXCR2. Calprotectin RNA levels returning to basal control levels after treatment are likely a testament to the fact that a large amount of CP released into the gut is neutrophil derived. In cxcl8-l1 E1/I1 MO experiments, IL-8 splice blocking led to significantly decreased levels of IL-8, neutrophils, and calprotectin in the treated plus infection group compared to those in the untreated plus infection group, while levels of neutrophils in the treated plus infection group were also significantly decreased compared to those in the control untreated no-infection group (Fig. 8A to C). These results highlight the importance of IL-8 in recruiting neutrophils to the site of infection and the subsequent release of calprotectin.

FIG 8.

IL-8, neutrophils (mpx), and S100A8/S100A9 (calprotectin) levels are decreased during V. cholerae infection in groups treated with morpholino (MO) (A to C) or the drug (SB 225002) (D to F) (Treated + Infection). WT larval zebrafish were infected with the 25493 (non-O1) strain of V. cholerae at 2.5 × 10⁷ CFU/mL and then sacrificed at the 24-h time point. mRNA levels were determined by qRT-PCR. Gene expression was normalized against β-actin and expressed as a fold change. Error bars indicate standard deviations. Student’s t test was used for statistical analysis. The data shown are from three experiments. *, P < 0.05 compared to the control Untreated − Infection group; **, P < 0.05 compared to the Untreated + Infection group.

DISCUSSION

Here, we describe the innate immune response of zebrafish to V. cholerae infection, with a particular interest in neutrophils, neutrophil-associated molecules, as well as downstream markers of inflammation such as TLR4 and NF-κB. This work aids in further solidifying the already established utility of the zebrafish model in studying V. cholerae infection as well as expanding its effectiveness to immune responses to V. cholerae and the field of infection immunology. Additionally, new relationships between the nutritional immunity protein calprotectin and V. cholerae are established in this work, expanding its potential as a diagnostic marker or therapeutic target of interest.

El Tor V. cholerae strains are associated with pandemic cholera and are the cause of the ongoing 7th pandemic worldwide. Non-O1/O139 strains are more commonly associated with the environment and cannot cause pandemic cholera but are responsible for sporadic outbreaks of diarrhea and bloody diarrhea in humans. Here, the immune responses of zebrafish against both types of V. cholerae were assessed as both types are associated with human disease and the aquatic environment. Neutrophil responses in the gut have been reported to peak at 24 h and contribute to the intestinal inflammation that is seen in many gastrointestinal (GI) diseases (32). In the current study, peak levels of neutrophil RNA at 24 h in zebrafish were also observed (Fig. 1A and B), which coincides with previously seen peaks of V. cholerae levels in zebrafish intestines (27, 28, 39). Intestinal infiltration was also witnessed via fluorescence microscopy, where neutrophils could be seen in V. cholerae biofilms (Fig. 2B and C). At the site of infection, neutrophils are known to release neutrophil extracellular traps (NETs), which are condensed chromatin that contains antimicrobial proteins to entrap and promote the extracellular killing of pathogens. It has been reported that V. cholerae induces NET formation and release upon contact with neutrophils and that V. cholerae utilizes two extracellular nucleases to degrade and evade these NETs (40, 41). Although the current study did not further investigate the relationship between zebrafish neutrophils and V. cholerae, future experiments similar to those seen here may provide further insight into this relationship.

One antimicrobial protein located on NETs is the nutritional immunity protein calprotectin. The antimicrobial activity of CP is well established, as is the mechanism through which CP inhibits microbial growth by transition metal chelation. Previously, only Ellis et al. have shown an increase in S100A8 protein during infection in cholera patients (31); however, no other work has been done to explore this relationship. In the current study, RNA and protein levels of calprotectin were observed to be significantly increased at 24 h and 72 h in fish infected with either V. cholerae El Tor or the non-O1/O139 environmental strains and returned to baseline levels by 120 h (Fig. 3A and B), providing new evidence of this host response during infection. After showing that purified CP protein was able to completely inhibit V. cholerae growth (Fig. 4A), we then established that it does so in part through the chelation of the transition metals zinc and manganese (Fig. 5A to C). Further evidence of this effect was also shown using the chemical metal chelator TPEN (Fig. 6A and B). It is well known that metals are tightly controlled and regulated by many biological systems, including V. cholerae, which uses metal transporters to tightly regulate intracellular levels (42). However, to our knowledge, this work shows for the first time the ability to inhibit V. cholerae growth through transition metal limitation. Further exploration into our understanding of how metal limitation inhibits growth may provide insight into exploiting this relationship as a potential target for therapeutics.

CP is also known to have downstream effects on the immune system. Using the zebrafish model, we showed that V. cholerae induces a significant CP response that leads to significant increases in the downstream receptor TLR4 (Fig. 7A and B), which then signals through the transcription factor NF-κB, a major regulator of both innate and adaptive immunity, and was also significantly increased in the fish model (Fig. 7C and D). This signaling pathway then leads to a positive-feedback loop, amplifying the immune response. This in turn leads to more neutrophil and leukocyte recruitment, as does CP in its inherent chemokine activity, which can be seen in increased CD45 RNA levels in the fish intestine (Fig. 7E and F).

The data in this study also show the importance of the cytokine IL-8 in the immune response to V. cholerae infection. By using morpholinos and drug inhibitors, IL-8 was found to be essential for neutrophil recruitment and, ultimately, the release of calprotectin (Fig. 8A to F). While the splice-blocking IL-8 morpholino prevents the production of the cytokine IL-8, the drug SB 225002 works as a potent chemokine receptor antagonist that inhibits IL-8 from binding its receptor CXCR2 (IL-8R) despite increased levels of IL-8 still being produced in response to infection. Using the IL-8 morpholino, IL-8, neutrophils, and calprotectin all returned to baseline levels after MO treatment. Drug inhibition of CXCR2 resulted in significantly decreased IL-8, neutrophil, and calprotectin levels compared to those in the untreated plus infection group; however, they were still significantly increased compared to those in the control untreated no-infection group. Because IL-8 is still produced in this experiment, it may be outcompeting the drug binding to its receptor CXCR2. In future experiments, increasing amounts of SB 225002 drug treatment may ameliorate these issues. de Oliveira et al. previously described that IL-8 was upregulated in zebrafish in response to acute inflammatory stimuli and was ultimately essential for neutrophil recruitment and the resolution of inflammation (13). Our data corroborate this, as IL-8 proved crucial in not just neutrophil recruitment but also calprotectin release, which comes primarily from neutrophils (14).

The use of the zebrafish model for studying V. cholerae infection is still relatively novel compared to other more commonly used mammalian animal models; however, its advantages are numerous. Among them are the availability of transgenic animals, transparent embryos, the relatively low cost, as well as several others (3). The work done here continues to strengthen the evidence for using this model for studying V. cholerae infection. In addition, this study expands the zebrafish model to the field of V. cholerae immunology, where a number of questions remain unanswered, such as how different strains of cholera cause differing levels of inflammation, how to create more efficacious vaccines, and how the gut microbiome affects pathogenesis (43).

Prior to this study, no published work had assessed the zebrafish calprotectin response. One reason for this may be due to a general lack of immunology-related tools in the zebrafish field, compared to other animal models. The data in this study provide evidence that the zebrafish is capable of mounting a calprotectin response similar to those seen in humans and other mammals. For the first time, a link between nutritional immunity and V. cholerae has also been established. The data presented here solidifying calprotectin responses to cholera call for more investigation into this relationship as well as other antimicrobial nutritional immunity proteins such as S100A12 (calgranulin C), which has been shown to be upregulated in response to other infectious diseases (44–46). One area of interest is the possibility of using calprotectin as a diagnostic marker. Although cholera can be diagnosed through rapid dipstick methods or simply diagnosed through symptoms in areas of endemicity, CP may provide another alternative for diagnosis and serve as an indicator of levels of inflammation in the patient. Due to its stability (stable in feces for up to 1 year when frozen), CP may be useful in epidemiological studies where samples must be revisited at a later date. Drawbacks to the use of this protein for diagnosis remain in that intestinal calprotectin release may not be specific for differentiating cholera from other bacterial infections or other inflammatory GI diseases, which can have increases as large as 10-fold and exceed 1 mg/mL (47, 48). In addition, the relationship between certain bacteria and CP needs further exploration, as the presence of CP can enhance some bacterial growth (49, 50).

In summary, the data presented here provide new approaches for the V. cholerae field using calprotectin as a potential diagnostic marker or therapeutic. This work has also established the ability to inhibit V. cholerae growth through metal limitation, giving further evidence that transition metals are essential for V. cholerae growth. The ability to alter or limit transition metal distribution and homeostasis in infected tissues may be an effective method in antibacterial development. Complete growth inhibition by CP would likely take excess metal binding capacity due to competition with high-affinity bacterial metal uptake systems (51). Further illumination of the pathogen factors affected, their intrinsic nutrient requirements, and the complex interplay between host and pathogen factors has the potential to open new avenues for the development of anti-V. cholerae therapeutics.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

V. cholerae El Tor strain N16961 (Sm^r [100 μg/mL]), V. cholerae classical strain O395 (Sm^r [100 μg/mL]), V. cholerae environmental non-O1/O139 strain 25493 (Sm^r [100 μg/mL]), V. cholerae El Tor strain E7946 (Sm^r [100 μg/mL]), and V. cholerae non-O1/O139 strain AM-19226 (Sm^r [100 μg/mL]) were used in this study. Bacterial strains were frozen in 20% glycerol in Luria-Bertani (LB) broth (Difco, NJ, USA) at −80°C. For experimentation, each strain was then grown in LB broth (Difco, NJ, USA) at 37°C under shaking conditions (180 rpm) or on plates in LB agar (Difco, NJ, USA) with the appropriate antibiotic(s). Thiosulfate-citrate-bile-sucrose (TCBS) agar (Difco, NJ, USA) was used as a selective medium for V. cholerae.

Zebrafish.

Wild-type AB strain zebrafish were used for all experiments, except for fluorescence microscopy experiments using Tg(mpx:Dendra2)^uwm4/AB zebrafish (52). For larval infections, zebrafish at 5 days postfertilization (dpf) were used. Zebrafish were housed in an automated recirculating tank system (Aquaneering, CA, USA) using water filtered by reverse osmosis and maintained at pH 7.0 to 7.5. The tank water was conditioned with Instant Ocean salt (Aquarium Systems, OH, USA) to a conductivity of 600 to 700 S. Zebrafish were euthanized in 100 mL of 32-μg/mL Tricaine-S (tricaine methanesulfonate) (catalog number MS-222; Western Chemical, WA, USA) for a minimum of 25 to 30 min after cessation of opercular movement.

Adult zebrafish infection procedure.

For experimental groups, 4 to 5 zebrafish were placed into a 400-mL beaker with perforated lids, containing 200 mL of tank water (autoclaved double-distilled water [ddH₂O] with 60 mg/L of Instant Ocean aquarium salts). Bacterial cultures were grown in LB broth at 37°C for 16 to 18 h with aeration. Bacteria were then washed once in phosphate-buffered saline (PBS) and diluted to a concentration of 10⁹ CFU/mL by measuring the OD at 600 nm. PBS-diluted bacteria were then added directly to beakers to an infection concentration of 2.5 × 10⁷ CFU and plated using serial dilutions for verification. Control fish were exposed to 1 mL of 1× PBS. Beakers containing fish were then placed in a glass-front incubator at 28°C for the duration of the experiment.

Larval zebrafish infection procedure.

At 5 days postfertilization, zebrafish larvae were placed into 50-mL beakers containing 20 mL sterilized tank water containing 2.5 × 10⁷ CFU/mL V. cholerae. For the designated drug treatment experiments, the selective nonpeptide inhibitor SB 225002 (Tocris, Minneapolis, MN) was added at a concentration of 5 μM directly to beakers containing fish for 1 h prior to infection. After 24 h, larvae were euthanized using a tricaine solution, added directly to 300 mL of TRIzol solution (Invitrogen, Waltham, MA), and homogenized using a pellet pestle (Fisher Scientific, Pittsburgh, PA). RNA was purified by ethanol precipitation and resuspended in RNase-free water. cDNA production and subsequent quantitative real-time PCR (qRT-PCR) were performed as described below.

Intestinal colonization assessment.

At the specified time points, adult zebrafish were euthanized using tricaine. Intestines were aseptically removed, placed into homogenization tubes (2.0-mL screw-cap tubes; Sarstedt, Nümbrecht, Germany) with 1.5 g of 1.0-mm glass beads (BioSpec Products, Inc., Bartlesville, OK) and 1 mL of 1× PBS, and held on ice. Homogenization tubes were loaded into a Mini-Beadbeater-24 instrument (BioSpec Products, Inc.). Serial dilutions of homogenized tissue were plated onto LB agar plates with the appropriate antibiotics.

RNA isolation and qRT-PCR.

Intestinal tissue from adult zebrafish was homogenized in 1 mL 1× PBS using homogenization beads as described above. RNA was then extracted using a Qiagen (Hilden, Germany) RNeasy minikit. Larval zebrafish were added directly to 300 mL of TRIzol solution (Invitrogen, Waltham, MA) and homogenized using a pellet pestle (Fisher Scientific, Pittsburgh, PA). RNA was purified by ethanol precipitation and resuspended in RNase-free water. Total RNA was resuspended in RNase-free water and quantified using a NanoDrop instrument. cDNA was then synthesized using an Invitrogen (Waltham, MA) SuperScript III first-strand synthesis system cDNA kit. qRT-PCR was performed using SYBR green (Applied Biosystems, Foster City, CA). Quantification of gene expression was performed using the comparative threshold cycle (ΔΔC_T) method. Gene expression was normalized to the endogenous reference β-actin level and is reported as the fold change relative to the reference gene.

Metal-reversible antimicrobial assays.

To assess the metal binding and inhibitory properties of calprotectin and the zinc chelator TPEN (Tocris, Minneapolis, MN) on V. cholerae, various concentrations of the protein or chelator were added directly to LB broth medium in a 96-well plate and incubated for 1 h at 37°C, and bacteria were then added and grown at 37°C with aeration. At the designated time points, the plate was then spectrophotometrically read at an OD₆₀₀. Mn²⁺ and Zn²⁺ were then added to the V. cholerae culture in increasing quantities to restore bacterial growth in calprotectin and TPEN experiments, and the OD₆₀₀ was read at 24 h. Recombinant wild-type and mutant CPs were expressed and purified as described previously (53).

IL-8 morpholino knockdown.

Morpholino (MO) microinjections were performed as previously described (54). Briefly, one- to four-cell-stage wild-type (AB) embryos were microinjected with the morpholino solution at 4 ng/embryo. For this study, two morpholinos were utilized, both obtained from Gene Tools (Philomath, OR): cxcl8-l1 E1/I1 MO (sequence, 5′-GGTTTTGCATGTTCACTTACCTTCA-3′), which is a previously established splice-blocking morpholino for IL-8 (13), and the standard control MO, which has no known target in zebrafish (55). At 48 h postfertilization (hpf), embryos from each group were pooled separately and harvested for qRT-PCR, as described above.

Imaging of infected zebrafish intestines.

Fish were euthanized using tricaine after 24 h of infection, and intestines were removed and then placed into 10% zinc formalin for 24 h. Next, intestines were placed into 70% ethanol and shipped to Reveal Biosciences, Inc. (San Diego, CA), for fluorescence imaging as follows. Samples were dewaxed in xylene three times and then cleared using 100% alcohol twice. Samples were then hydrated in 95% alcohol and rinsed in distilled water twice. Slides were mounted using ethyl methanesulfonate (EMS), stained using Fluoro-Gel II with DAPI and an anti-V. cholerae polyclonal antibody (KPL BacTrace), and then counterstained with a secondary antibody conjugated to Alexa Fluor 647. Images were uniformly adjusted for contrast and brightness in the figures.

Intestinal homogenate ELISA.

Infection procedures were performed as described above. Intestinal tissue was removed and placed into 100 μL 1× PBS and homogenized using pellet pestles. Next, 50 μL of radioimmunoprecipitation assay (RIPA) buffer was added, and the samples were centrifuged at 5,000 × g for 5 min to remove debris. Samples were then diluted 1:25 with 1× PBS, and a fish calprotectin enzyme-linked immunosorbent assay (ELISA) kit was run according to the manufacturer’s instructions (MyBioSource, San Diego, CA), as follows. Fifty microliters of the standard or sample was added to each well. One hundred microliters of a horseradish peroxidase (HRP)-conjugated reagent was added to each well, and the plate was incubated for 60 min at 37°C. Wells were washed 4 times, 50 μL chromogen solutions A and B was added to each well, and the mixture was incubated for 15 min at 37°C. Fifty microliters of stop solution was added to each well, and after 15 min, the plate was read spectrophotometrically at an OD₄₅₀.

Statistical analysis.

Experiments were performed in triplicate on separate occasions, unless otherwise specified. Data shown are presented as the means ± standard deviations (SD). All statistical analyses and Student’s t tests were performed using Prism version 7.0 for Windows (GraphPad Software, La Jolla, CA).

Ethics statement.

All animal procedures were approved by the Wayne State University IACUC, protocol number 18-10-0809. Zebrafish were euthanized in 100 mL of 32-μg/mL Tricaine-S (tricaine methanesulfonate) (catalog number MS-222; Western Chemical, WA, USA) for a minimum of 25 to 30 min after cessation of opercular movement.