Human Intelectin-1 Promotes Cellular Attachment and Neutrophil Killing of Streptococcus pneumoniae in a Serotype-Dependent Manner

ABSTRACT

Human intelectin-1 (hIntL-1) is a secreted glycoprotein capable of binding exocyclic 1,2-diols within surface glycans of human pathogens such as Streptococcus pneumoniae, Vibrio cholerae, and Helicobacter pylori. For the latter, lectin binding was shown to cause bacterial agglutination and increased phagocytosis, suggesting a role for hIntL-1 in pathogen surveillance. In this study, we investigated the interactions between hIntL-1 and S. pneumoniae, the leading cause of bacterial pneumonia. We show that hIntL-1 also agglutinates S. pneumoniae serotype 43, which displays an exocyclic 1,2-diol moiety in its capsular polysaccharide but is unable to kill in a complement-dependent manner or to promote bacterial killing by peripheral blood mononuclear cells. In contrast, hIntL-1 not only significantly increases serotype-specific S. pneumoniae killing by neutrophils but also enhances the attachment of these bacteria to A549 lung epithelial cells. Taken together, our results suggest that hIntL-1 participates in host surveillance through microbe sequestration and enhanced targeting to neutrophils.

INTRODUCTION

Human immune lectins are known for their unique abilities to distinguish between self and non-self and play major roles during the innate immune response. During this response, lectins are used to initiate first contact with the invading microbes through recognition of specific carbohydrate structures. Subsequently, there are a plethora of effects the lectins can exert that lead to the elimination of the unwanted bacteria. For example, the binding of microbial surface mannans through the mannose-binding lectin (MBL), one of the best-studied lectins, initiates a cascade of binding events involving approximately 20 serum proteins prior to bacterial destruction by complement-mediated pore formation (1). The MBL and the later-discovered ficolin-2 are very similar in structure yet diverse in their target glycans (2). Nevertheless, the recognition of pathogens by either lectin leads to the recruitment of the MBL-associated serine proteases, which constitutes the activating step of the complement pathway (1, 3). Galectins are a distinct lectin family exhibiting a more direct mode of killing pathogens. These lectins are known for their agglutinating and membrane-disrupting activity through recognition of human blood group antigen-like structures. Galectins 4 and 8 specifically have been shown to bind the Escherichia coli strain 086, which mimics blood group B antigens on its surface (4). Lectin recognition causes interference with bacterial motility and disruption of membrane integrity (4). Apart from their killing mechanisms, lectins differ in their glycan binding specificity, playing different roles during the immune response. Both ficolin-2 and the galectins have very narrow carbohydrate specificities recognizing the O-acetyl group of carbon-6 of beta-galactose in S. pneumoniae 43 (also called 11A) and the blood group B antigen epitope displayed by E. coli 086, respectively (3–5). These ligands contribute to lectin specificity during the innate immune response, leading to recognition of their pathogen of choice down to certain serotypes within a species. A lectin with a much broader target range is RegIII or its human equivalent, HIP/PAP. RegIII recognizes peptidoglycan and forms pores in the membranes of Gram-positive bacteria (6, 7). This broad killing mechanism is associated with a role in host-microbe sequestration and general maintenance of homeostasis rather than specific pathogen surveillance and is likely the reason for the highly regulated manner of RegIII release and activation (6, 8). The mechanisms used by MBL/ficolin-2, the galectins, and RegIII to achieve bacterial clearance illustrate the diversity among secreted human innate immune lectins. Apart from the different processes employed by lectins for bacterial recognition and killing, many lectins show commonalities based on their structural characteristics. For example, the multimeric assembly of active forms of most lectins leads to increased avidity for their targets, since individual carbohydrate binding affinities commonly involve weak interactions. These lectins can also act in the same fashion as antibodies, in which their binding leads to the agglutination of microbes that alone can act as a physical barrier around the bacteria, blocking their access to the host epithelium and facilitating the removal of bacterial aggregates in the mucosal layer through the action of cilia and microvilli (9, 10). Alternatively, lectin binding, similar to antibodies, can also opsonize bacteria for phagocytic uptake by neutrophils and macrophages as part of the inflammatory processes that are activated during immune responses (3).

One lectin that has been proposed to be an important member of the innate immune response is human intelectin-1 (hIntL-1). hIntL-1 is part of the X-lectin, or intelectin, family, whose first member was identified in 1974 in the African toad Xenopus laevis (11, 12). hIntL-1 is a glycoprotein that is expressed by several cell types in many different tissues as well as in serum, where it is present in micrograms-per-milliliter concentrations (13). The highest hIntL-1 expression levels originate from goblet cells in the digestive system (stomach, duodenum, small intestine, and colon), where it is secreted into the mucus (14). It can further be found in the mesothelium, bladder, bronchial mucus glands, kidney, heart, testis, and basal stem cells, making it widely distributed throughout the human body (14–18). In the airways, in particular, the presence of hIntL-1 has been studied in connection with asthma and smoking, where its expression can be stimulated by interleukin-13 (IL-13) (19, 20). The active form of hIntL-1 consists of a disulfide-linked homotrimer complexing three calcium ions in each monomer, one of which is part of the glycan binding site and essential for ligand binding (21). The glycan binding specificity of hIntL-1 was originally analyzed by glycan arrays and showed that only microbial surface glycans, no mammalian glycans, are recognized (22). More specifically, β-linked d-galactofuranose (β-Galf), d-phosphoglycerol-modified glycans (G-1-P), d-glycero-d-talo-oct-2-ulosonic acid (KO), and 3-deoxy-d-manno-oct-2-ulosonic acid (KDO) are bound by the lectin, all of which have an exocyclic 1,2-diol moiety in common, constituting the ligand for hIntL-1 (21, 23). Despite this specificity, it was found that sialic acid, a sugar common to mammals, is not recognized by hIntL-1 due to anion-anion repulsion of sialic acid by one of the carboxyl groups in the binding site (21). Similarly, heptoses possess exocyclic 1,2-diols but are only weakly bound by hIntL-1 due to stereoelectronic effects (23). This demonstrates the degree of specificity in glycan recognition that allowed intelectins to differentiate self and non-self in phyla from placozoa to chordata (24, 25).

The first bacteria shown to be bound by hIntL-1 were S. pneumoniae serotypes 20, 43, and 70, which contain either β-galactofuranoses or glycerol-1-phosphates in their capsular polysaccharide (CPS) repeating units (21). This discovery was closely followed by a publication that noted recognition of a variety of Gram-negative (Vibrio cholerae, E. coli, Vibrio parahaemolyticus, and Salmonella enterica) and Gram-positive (Listeria monocytogenes and Staphylococcus aureus) human pathogens by hIntL-1 (26). The only bacterium for which lectin effects on host biology have been described is Helicobacter pylori. The binding by hIntL-1 led to the agglutination of the bacteria in a calcium- and concentration-dependent manner as well as to an increase in phagocytic uptake by macrophages (17). Notably, all of the bacteria bound by hIntL-1 are considered human pathogens, leading to the proposed role of hIntL-1 in pathogen surveillance.

Beyond the effects of hIntL-1 observed for H. pylori, it is unknown how hIntL-1 interacts with other microbes it recognizes or how the lectin protects from infection. We hypothesized that S. pneumoniae serotypes expressing exocyclic 1,2-diols would also be agglutinated by hIntL-1 and that this would be coupled to microbial clearance by host immune cells. Our results demonstrate that hIntL-1 specifically targets S. pneumoniae serotype 43 to polymorphonuclear leukocytes (PMNs) but not peripheral blood mononuclear cells (PBMCs) and also contributes to lung epithelial cell binding.

RESULTS

hIntL-1 binds and agglutinates S. pneumoniae serotype 43.

hIntL-1 was purified from the supernatant (conditioned medium) of transfected HEK-293F cells using a β-galactofuranose agarose column. Subsequently, the elution fractions shown in Fig. 1A were combined, concentrated, and used in the experiments described in this study. Previous studies have shown that hIntL-1 binds S. pneumoniae serotypes specifically (serotypes 20, 43, and 70), depending on their CPS structure (21). Therefore, we first verified this finding for S. pneumoniae serotypes 8 and 43 (Fig. 1B). The CPS repeating units of both serotypes are shown in Fig. 1C. For serotype 8, the CPS consists of glucose and galactose repeating units that do not include any exocyclic 1,2-diols. In contrast, S. pneumoniae serotype 43 displays a glycerol-1-phosphate moiety in its CPS that is recognized by hIntL-1. Fluorescent microscopy with 488 Alexa-Fluor-labeled hIntL-1 showed lectin binding only to S. pneumoniae serotype 43 (Fig. 1B). As previously observed for H. pylori, microscopy further indicated that serotype 43 is agglutinated by hIntL-1 (17). Agglutination was also verified in a round-bottom well agglutination assay in which an agglutinating agent cross-links bacterial cells, acting as a spacer between them and preventing the formation of a small pellet that is characteristic of the lack of agglutination in a well. As shown in Fig. 1D, the presence of hIntL-1 leads to bacterial agglutination (lack of pellet formation) in a concentration-dependent, serotype-specific, and calcium-dependent manner.

FIG 1.

hIntL-1 binds and agglutinates S. pneumoniae serotypes specifically. (A) Coomassie-stained SDS-PAGE of the hIntL-1 purification. hIntL-1 was purified from the supernatant of transfected HEK-293F cells using a β-galactofuranose agarose column and eluted with EDTA. Monomeric hIntL-1 has a molecular mass of 35 kDa. Molecular masses, in kDa, are shown. (B) S. pneumoniae was labeled with 4’,6-diamidino-2-phenylindole (DAPI, blue) and incubated with or without Alexa Fluor 488-labeled human intelectin-1 (hIntL-1, green) prior to viewing by fluorescence microscopy at ×40 magnification. (C) The capsular polysaccharide of S. pneumoniae serotype 8 consists of GlcA-1,4Glc-1,4Glc-1,4Gal-1,4 repeating units and does not contain an exocyclic 1,2-diol. Serotype 43 displays the hIntL-1 ligand, glycerol-1-phosphate (yellow circle), that is part of the GlcNAc3Ac[-4Gro-1-P]-1,4Gal-1,3Gal-1,4Glc-1,6 CPS repeating unit. (D) S. pneumoniae agglutination assay in calcium-containing buffer using indicated concentrations of hIntL-1. EDTA was used to inhibit hIntL-1 as a control. The assay was performed in biological triplicate, and representative results are shown.

hIntL-1 does not kill bacteria directly or through complement activation.

After the initial validation of the S. pneumoniae model, the direct biological effect of hIntL-1 on the bacteria was tested. The addition of hIntL-1 onto an S. pneumoniae soft-agar overlay did not affect bacterial growth (Fig. 2A). While some lectins, such as galectins 4 and 8, can kill bacteria directly through contact, others, like MBL and ficolin-2, require the complement cascade for the same effect (3, 4, 27). We therefore examined the ability of S. pneumoniae to survive in pooled normal human serum in the presence and absence of hIntL-1 and complement. To confirm the presence of specific antibodies for both S. pneumoniae serotypes in the pooled serum, S. pneumoniae whole-cell lysates were probed with the pooled human serum (used for the serum survival assay), followed by detection with an anti-human IgG4 antibody. As shown in Fig. 2B, several proteins were detected in the pooled serum. During the serum survival assay, both serotypes showed robust survival in the presence and absence of hIntL-1 (Fig. 2C), demonstrating that hIntL-1 does not promote killing by the complement cascade.

FIG 2.

hIntL-1 does not kill S. pneumoniae directly or in a complement-dependent manner. (A) Spot assay showing indicated S. pneumoniae serotypes in soft agar overlay with either hIntL-1, buffer alone (control), or lysis buffer (control) spotted onto the agar prior to overnight growth. Blood agar beneath agar overlay is visible in lysis buffer control. (B) Verification of the presence of S. pneumoniae-specific antibodies in pooled human complement-containing serum. Equal concentrations (OD₆₀₀ of 1) of bacterial whole-cell lysates of serotypes (indicated as 8 or 43) were Coomassie stained (left) or probed with pooled human complement serum, followed by detection by mouse-anti-human IgG4 Fc HRP antibodies in a Western blot (right). Molecular masses, in kDa, are shown. (C) In the serum bactericidal assay, S. pneumoniae was incubated with undiluted pooled human serum containing complement in the presence or absence of hIntL-1. Numbers of CFU per milliliter were determined after 2 h of incubation at 37°C, and percent survival was calculated relative to CFU numbers obtained with heat-inactivated serum. The error bars represent the standard errors of the means from quadruplicate independent experiments.

hIntL-1 increases bacterial attachment to host cells.

While hIntL-1 did not show direct killing of S. pneumoniae, the agglutinating phenotype caused by the lectin is likely to interfere with a variety of processes during bacterial infection. A major aspect of bacterial pathogenesis is the ability of the microbe to attach to host tissues and form biofilms. We first tested the effect of hIntL-1 on S. pneumoniae biofilm formation in a 96-well biofilm assay with spectrophotometric quantification. This assay showed no significant difference in biofilm formation in the presence of hIntL-1 (Fig. 3A). We further tested the human lung epithelial A549 cell line that is commonly used in S. pneumoniae host cell studies. Here, serotype 43 showed a significantly higher ability to attach to the host cells. This finding was verified by microscopy and Hema3 staining (Fig. 3C). The attachment of S. pneumoniae serotype 43, but not serotype 8, showed an unexpected increase in A549 cell attachment in the presence of hIntL-1 (Fig. 3B).

FIG 3.

hIntL-1 does not alter S. pneumoniae biofilm formation but increases A549 cell attachment. (A) S. pneumoniae with or without hIntL-1 was incubated in 96-well plastic plates for 24 h. After removal of planktonic bacteria, biofilms were stained with crystal violet and absorbance readings were measured as shown. Error bars represent the standard errors of the means from five replicates. (B) S. pneumoniae attachment to confluent A549 lung cells in the presence and absence of hIntL-1. The assay was done at an MOI of 100 for 2 h. Percent attachment was determined after washing unbound bacteria relative to wells without A549 cells. The error bars represent the standard errors of the means from four replicates, and an asterisk indicates significant difference (P < 0.05, Student's t test). (C) For microscopy, the attachment assay was modified by incubating S. pneumoniae with or without hIntL-1 with A549 cells grown to 75% confluence on glass slides. Hema3 staining was used to visualize the bacteria and A549 cells. Images are shown at ×100 magnification in each panel.

hIntL-1 does not cause increased bacterial killing by PBMCs.

We next sought to determine whether hIntL-1 was capable of opsonizing and subsequently targeting bacteria toward phagocyte elimination. In this phase of host defense, a variety of immune cells are recruited for pathogen elimination. One group of immune cells is the PBMCs containing the phagocytic cells, monocytes. To test this, we performed an opsonophagocytosis assay (OPA) by infecting freshly isolated PBMCs with S. pneumoniae in the presence or absence of hIntL-1. We also included pneumococcal pool B serotype 8-specific antisera and pool D serotype 43-specific antisera as assay controls (commonly used for S. pneumoniae serotyping) and verified their binding and agglutinating abilities through fluorescence microscopy with 4′,6-diamidino-2-phenylindole (DAPI)-stained bacteria (Fig. 4A). A low multiplicity of infection (MOI) of 0.05 (bacteria to PBMCs) was selected due to the extensive ability S. pneumoniae possesses to evade clearance strategies employed by the immune system, allowing the microbe to survive even when outnumbered by human immune cells (28–30). S. pneumoniae showed robust survival and even a trend toward increased proliferation when incubated with PBMCs. Thus, the addition of hIntL-1 did not show statistically significant changes in bacterial numbers, although incubation with serotype-specific antibodies showed a significant decrease in bacterial survival (Fig. 4B). The experiment was then repeated in the presence of 20% heat-inactivated human serum to supply autologous antibodies to accompany hIntL-1 during bacterial opsonization. The addition of serum did not lead to changes in bacterial killing in the presence of hIntL-1 (Fig. 4C).

FIG 4.

hIntL-1 does not promote S. pneumoniae phagocytosis by human PBMCs. (A) Labeled S. pneumoniae (blue) was incubated with pneumococcal antiserum pools B and D, or water, before imaging via fluorescence microscopy at ×64 magnification. (B) S. pneumoniae opsonophagocytosis assay (OPA) with or without hIntL-1. Human peripheral blood mononuclear cells (PBMCs) were freshly isolated from different donors for each of the 7 replicates shown. S. pneumoniae serotype-specific antisera were added as bacterial agglutination controls. Antiserum pool B and pool D bind to S. pneumoniae serotypes 8 and 43, respectively. The OPA was done at an MOI of 0.05 for 2 h. Percent survival was quantified by determining colony counts relative to samples with heat-inactivated PBMCs. Error bars represent the standard errors of the means, and the asterisks indicate significant differences (*, P < 0.05, Student's t test). (C) OPA as described in panel B with the addition of 20% fresh autologous human serum in each sample. Error bars represent the standard errors of the means for four replicates.

hIntL-1 causes a drastic increase in bacterial killing by neutrophils.

PMNs are another prominent class of immune cells during bacterial infection and are particularly abundant during S. pneumoniae infections. PMNs are equipped with a variety of mechanisms to kill bacteria, including phagocytosis, intracellular killing in the phagosome, and release of reactive oxygen species (ROS), as well as extracellularly through employment of neutrophil extracellular traps (NETs) and serine proteases (31, 32). To test the impact of hIntL-1 during PMN-mediated killing, S. pneumoniae and PMNs were incubated at a multiplicity of infection (MOI) of 0.01 (bacteria to PMNs) with 5% autologous serum and in the presence or absence of hIntL-1. As shown in Fig. 5A, while serotype 8 was unaffected by the presence of hIntL-1, the presence of the lectin led to a significant drop in serotype 43 survival, from 83% to 31% (Fig. 5A). The use of pneumococcal antiserum pools B and D as controls for bacterial clumping showed drastic and serotype-specific decreases in bacterial survival during the PMN killing assay. To assess the ability of hIntL-1 to serve as an opsonin during the PMN killing process, the assay was repeated with heat-inactivated serum and in the absence of serum. Through incubation of serum at 56°C, complement components were degraded, resulting in serum that contains only antibodies as the opsonizing agents. PMNs were unable to kill either serotype with hIntL-1 in the absence of complement. Furthermore, no decrease in bacterial survival could be noted in the presence of hIntL-1 in samples lacking serum altogether. These data suggest that hIntL-1 cannot serve as an opsonin itself. Rather, the increase in killing of S. pneumoniae serotype 43 in the presence of hIntL-1 is due to bacterial agglutination in conjunction with complement opsonization.

FIG 5.

hIntL-1 leads to increased S. pneumoniae serotype 43 killing by human PMNs. (A) PMNs and autologous serum were freshly isolated from the blood of healthy human donors (n ≥ 3). S. pneumoniae serotype-specific antisera were added as bacterial agglutination controls. Pool B and pool D antisera bind to S. pneumoniae serotypes 8 and 43, respectively. During the neutrophil killing assay, cells and bacteria were incubated at an MOI of 0.01 for 2 h in the presence of 5% autologous serum, 5% heat-inactivated autologous serum (56°C for 1 h for heat-inactivated complement, HIC), or no serum. Percent survival was determined by comparing colony counts relative to samples without PMNs as controls or to samples with inactivated PMNs for samples without serum. The error bars represent the standard errors of the means, and the asterisks indicate significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001 by Student's t test). (B) Neutrophil extracellular trap (NET) production using the Sytox Orange nucleic acid stain. PMNs isolated from five different donors were incubated with and without hIntL-1. Percent fluorescence was calculated relative to measurements with phorbol 12-myristate 13-acetate (PMA)-activated PMNs. The error bars represent the standard errors of the means, and an asterisk indicates significant differences (*, P < 0.05, Student's t test). (C) Measurement of reactive oxygen species production using the Diogenes-based chemiluminescence kit. PMNs from different donors for each of 7 replicates were incubated with and without hIntL-1. Percent luminescence was calculated relative to measurements with phorbol 12-myristate 13-acetate (PMA)-activated PMNs. (D) The enzymatic activity of neutrophil elastase released from PMNs in the presence of hIntL-1 was measured using the neutrophil elastase activity assay kit. PMNs isolated from six different donors were incubated with hIntL-1. Neutrophil elastase activity, in milliunits per milliliter, was calculated using a neutrophil elastase standard curve.

We aimed to further investigate the mechanism of killing leading to hIntL-1-mediated bacterial elimination. Direct incubation of PMNs with hIntL-1 led to a decrease of DNA release by the cells, indicating that hIntL-1 does not affect PMN viability or increased NET formation (Fig. 5B). Similarly, the production of ROS (H₂O₂) and the release of enzymatically active neutrophil elastase from PMNs were not stimulated by hIntL-1 (Fig. 5C and D).

To establish whether bacterial phagocytosis by PMNs is the primary killing mechanism, two approaches were utilized. The percentage of phagocytosing PMNs was assessed using flow cytometry. Bacteria were stained using the pH-sensitive dye pHrodo, which fluoresces green only in the acidic environment of the phagosome. A representative image of the flow gating strategy is shown in Fig. 6A. The collected data from four replicates showed the same amount of PMN phagocytosis for both serotypes in the presence and absence of hIntL-1 (Fig. 6B). Since the data only represent the number of PMNs that have phagocytosed bacteria but not the amount of bacteria phagocytosed per PMN, we further employed fluorescence microscopy. To quantify the bacteria phagocytosed by PMNs, we used pHrodo-labeled S. pneumoniae to visualize phagocytosis. Using ImageJ software, 150 PMNs were analyzed for their green fluorescence intensity, resulting in a significant increase in bacterial uptake of S. pneumoniae serotype 43 in the presence of hIntL-1 (Fig. 6C). The same trend can be seen in the representative fluorescence microscopy image in Fig. 6D showing hIntL-1-mediated serotype-specific phagocytosis.

FIG 6.

PMNs display increased bacterial uptake of S. pneumoniae serotype 43 in the presence of hIntL-1. (A) Flow cytometry was used to quantify the percentage of PMNs that have phagocytosed S. pneumoniae (Spn) in the presence and absence of hIntL-1. The gating strategy for the percentage of PMNs positive for phagocytosis (Spn+ PMN) is shown for one representative sample (S. pneumoniae serotype 43 in the presence of hIntL-1). (B) Flow cytometry analyses for phagocytic PMNs were executed as four replicates with isolated PMNs from four different donors. Means and standard errors are shown. FACS, fluorescence-activated cell sorting. (C) pHrodo-labeled S. pneumoniae (green) was incubated with PMNs at an MOI of 5 for 2 h in the presence or absence of hIntL-1. Samples were fixed with paraformaldehyde, applied to microscopy slides, and stained with DAPI (blue). Using the ImageJ software, the green fluorescence intensity per PMN was quantified for a total of 150 cells from three independent experiments. Means and standard errors are shown, and the asterisks represent significant difference (****, P < 0.0001, Student's t test). (D) Representative images of fluorescence microscopy showing the phagocytosis of pHrodo-labeled S. pneumoniae by PMNs in the presence and absence of hIntL-1. Images were obtained using fluorescence microscopy at ×100 magnification.

DISCUSSION

Our research aimed to better understand the proposed role of hIntL-1 in pathogen surveillance as a part of the human innate immune response. As shown previously for S. pneumoniae and H. pylori, hIntL-1 is unable to kill bacteria directly, and our data further demonstrated that the presence of the lectin does not lead to bacterial killing through the complement-mediated membrane attack complex (17, 21). hIntL-1 is composed of a homotrimer containing a glycan binding site in the fibrinogen-like domain with three bound calcium ions in each monomer, but it lacks the collagen-binding domain present in MBL and ficolins. Thus, it is not surprising that hIntL-1 is unable to recruit MBL-associated serine proteases (MASPs) to engage the complement cascade (21). This activity has only been shown for lectins such as MBL, ficolins, and other C-type lectins (collectins), all of which contain the appropriate fibrinogen-like carbohydrate binding domains in their N-termini and collagen-like domains for association with MASPs (2, 33, 34). Nevertheless, it was important to exclude the possibility of hIntL-1 association with the complement cascade, particularly since MBL was long believed to be the only lectin to activate the lectin complement pathway and now several other lectins have been demonstrated to possess this ability (33, 34).

Previous research accompanied by the findings presented here shed further light on the role hIntL-1 plays within the large group of human innate immune lectins. While lectins like RegIII and the galectins demonstrate direct killing effects, others engage various immune proteins and host cells to assist in bacterial elimination (4, 6, 7). Especially in the context of Gram-positive bacteria with thick peptidoglycan layers, for which the membrane attack complex of the complement pathway is only minimally effective, the most potent method of microbial killing involves phagocytosis (29). Lectins have evolved to utilize phagocytic cells in different ways. The presence of soluble lectins like MBL and ficolin-2 lead to increased phagocytosis upon bacterial binding through opsonization with recruited proteins of the complement cascade (3). A direct method for lectins to facilitate phagocytosis is observed with membrane-bound lectins, such as the mannose receptor present on macrophage surfaces, where pathogen binding to the lectin leads to direct phagocytic uptake (35, 36). In the case of hIntL-1, our data show no killing through complement alone but enhanced bacterial killing by PMNs in the presence of serum. This suggests bacterial phagocytosis is initiated through opsonization by complement but increased through the agglutinating effect of the lectin binding (Fig. 7). Agglutination is a common feature for lectins involved in innate immunity due to their multimeric structures, which increase binding avidity and allow for increased bacterial clearance. Bacterial aggregates are typically hampered in their ability to attach to and invade the host epithelium and are more easily removed by the mucociliary system (10, 37, 38). Combined, we suggest the primary function for hIntL-1 is in bacterial agglutination and facilitating killing by specific immune cells.

FIG 7.

Proposed role of hIntL-1 during S. pneumoniae serotype 43 infection and immune-mediated clearance. hIntL-1 (red square) recognizes S. pneumoniae serotype 43 (Spn43, dark blue circle) through exocyclic 1,2-diols displayed on its CPS, leading to bacterial agglutination and increased attachment to lung A549 host cells. The presence of hIntL-1 does not result in an increase in bacterial killing by PBMCs, and complement (green triangle) alone does not kill Spn43 in the presence or absence of hIntL-1 through a membrane attack complex (MAC) mechanism. Rather, complement opsonization together with hIntL-1-mediated Spn43 agglutination leads to PMN uptake. The presence of hIntL-1 does not increase production of neutrophil extracellular traps (NETs), reactive oxygen species (ROS), H₂O₂, or neutrophil elastase (NE) by PMNs; instead, the increased killing is caused by phagocytosis.

The exocyclic 1,2-diol glycan ligand is widely distributed among bacteria and is present as surface polysaccharides, including glycerol-1-phosphates, β-galactofuranose, KO, and KDO (21). This gives hIntL-1 the potential to bind Gram-positive and Gram-negative bacteria, commensals, and pathogens alike. While MBL and RegIII have broad glycan binding specificities, the galectins and ficolin-2 display narrow ligand ranges (1, 3, 6). Ficolin-2 has been shown to recognize S. pneumoniae serotype 43 (also called 11A), leading to the activation of the complement cascade and increased phagocytosis. The authors of this study note the comparatively low invasive potential of serotype 43 among other S. pneumoniae serotypes and speculate that the recognition through ficolin-2 is the cause of this lowered pathogenicity (3). This phenotype can be further explained by the specific targeting of this bacterium through not only ficolin-2 but also hIntL-1. This demonstrates the intricacies employed by our immune system to ensure effective and redundant defenses.

hIntL-1 is secreted into intestinal and airway mucus, where it can exercise its agglutinating effects on a variety of bacteria that can penetrate the mucosal layer and present a risk to the host (14, 19, 20). It is anticipated that the resulting bacterial aggregates will be more easily removed through the mucociliary system or phagocytosed by immune cells. Although our data indicate an unexpected increase in S. pneumoniae serotype 43 attachment to A549 cells, this does not necessarily negate the role for hIntL-1 in host protection in vivo. One possible explanation for the increase in host cell attachment may be a direct result from bacterial agglutination while still leaving some of the bacterial surfaces exposed for host contact. This would lead to attachment of large bacterial aggregates rather than short bacterial chains or diplococci. This phenomenon has been observed before with S. pneumoniae mutants that are capable of forming elongated bacterial chains showing increased attachment to A549 cells compared to the wild-type (WT) strain, with predominantly diplococcal cells (39). This relationship also held true in a mouse model where longer-chain-forming mutants exhibited higher degrees of colonization (39). Additionally, S. pneumoniae is well known for its ability to adapt to and evade human defense mechanisms (29, 30). Thus, it is also possible that bacteria like S. pneumoniae have evolved methods to use immune strategies of their hosts to their own benefit. Nevertheless, although the A549 cell line is a commonly used pulmonary epithelial model, it does not possess a thick mucus layer or the complete architecture of the more complex lung environment, so the increased attachment of agglutinated S. pneumoniae serotype 43 to host cells should be explored further. Future work could test this hypothesis using in vitro cultures of human bronchial epithelial cells, lung organoids, or mouse models.

Our data demonstrate that hIntL-1 enhanced killing of S. pneumoniae serotype 43 by PMNs in the presence of complement, suggesting the main mechanism of microbial elimination is hIntL-1-mediated bacterial agglutination followed by elimination through PMN phagocytosis. Previous studies have shown increased phagocytosis of H. pylori by bone marrow-derived macrophages in the presence of hIntL-1 (17). For S. pneumoniae, we found that hIntL-1 did not increase killing by freshly isolated human PBMCs. This discrepancy could be due to a variety of factors, including the differences in the bacterium, experimental conditions, and immune cells. Pneumococcal infections are controlled by host neutrophils, which kill this pathogen via opsonophagocytosis in collaboration with the complement system (28). Given the use of S. pneumoniae as our model pathogen, it is not surprising that we observed PMNs as the primary mechanism for bacterial killing together with hIntL-1 (rather than PBMCs). During the immune response to invading S. pneumoniae, all three pathways of complement are activated, resulting in the deposition of C3b on the bacterial surface (40, 41). While in Gram-negative bacteria the formation of the membrane attack complex would play a major role in bacterial elimination, this effect excludes Gram-positive bacteria due to the thick layer of impenetrable peptidoglycan (29). Instead, the deposited C3b is cleaved to iC3b, which is recognized by complement receptor 3 on neutrophils, stimulating phagocytosis (42–44). Therefore, killing of S. pneumoniae is largely facilitated through neutrophil uptake. We have observed similar bacterial killing mechanisms in our studies in the presence of hIntL-1. We observed no increase in NET, hydrogen peroxide, or neutrophil elastase release. Instead, we noted hIntL-1-mediated increases in killing of S. pneumoniae serotype 43 only in the presence of complement and an enhanced uptake of bacteria, while the number of phagocytic PMNs remained unchanged. The presence of hIntL-1 did not increase the number of phagocytic events; rather, the bacterial count for each phagocytosed unit was increased through bacterial clumping. Our data further suggest that hIntL-1 acts as an agglutinating agent, but not an opsonin, leading to an increase of bacterial killing through bacterial clumping followed by complement opsonization and phagocytic uptake. This finding is supported by studies demonstrating enhanced phagocytosis of S. pneumoniae mutants with increased chain lengths. To investigate this, the authors used an S. pneumoniae mutant with long-chain morphology and showed increased complement deposition as well as uptake by human neutrophils compared to the wild type. The fitness disadvantage observed for long bacterial chains of S. pneumoniae can be phenotypically compared to the bacterial aggregation observed in the presence of hIntL-1 (28). It can be speculated that hIntL-1, in interaction with neutrophils, has a similar killing effect on other S. pneumoniae serotypes containing the exocyclic 1,2-diol ligand. This could include the serotypes 20 and 70 that have been shown to be bound by hIntL-1 as well as the 21 other serotypes (10, 13, 18, 22, 31, 34, 35, 41, 44, 53, 55, 56, 61, 69, 72, 74, 78, 10B, 11D, 11E, and 20B) that contain β-galactofuranoses or glycerol-1-phosphates in their CPS repeating unit (21, 45, 46).

The full potential for intelectin function as part of an immune system is best exemplified by looking at one of arguably the simplest free-living animals on earth. As a placozoan, Trichoplax measures 1 to 2 mm in diameter and consists of two epithelial layers that sandwich multinucleated fiber cells (47, 48). The organism only contains a small number of cell types, none of which appear to be immune cells, suggesting an adaptive immune response is not possible (49, 50). Instead, Trichoplax employs a wide variety of Toll-like and NOD-like extracellular scavenger receptors as well as lectins, composing its innate immune system. Computational analyses identified 31 intelectins in the genome of Trichoplax, which, in combination with the other pattern recognition molecules, ensure differentiation between self and non-self (24). Humans, in comparison, possess only two intelectins. While the field has been working on elucidating the role of hIntL-1, much less is known about hIntL-2, which shows a more restricted glycan binding specificity and is only expressed by Paneth cells in the small intestine (18). Nevertheless, our data serve to exemplify the evolution of intelectins from a potentially key player in the Trichoplax innate immune response to a microbial pattern recognition receptor involved in pathogen surveillance.

Collectively, we suggest the main mechanism of bacterial killing by hIntL-1 lies in its ability to agglutinate followed by targeting microbes for elimination through neutrophil phagocytosis. Considering the prevalent nature of the ligand for hIntL-1 on various microbes, along with our data showing enhanced neutrophil phagocytosis, hIntL-1 is most likely involved in general host defense, preventing the breach of host barriers.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The S. pneumoniae serotypes 8 (ATCC 6308) and 43 (ATCC 10343) were grown on tryptic-soy agar (Hardy Diagnostics) plates enriched with 10% defibrilized sheep blood (TSA plates) or grown statically in liquid Bacto Todd Hewitt broth (Becton, Dickinson, and Company) with 0.5% yeast extract (THY) at 37°C under aerobic conditions.

Expression and purification of native human intelectin-1.

Standard protocols were used for the generation of the hIntL-1 expression plasmid restriction enzyme digests, DNA ligations, PCR, and other recombinant DNA procedures. cDNA encoding the open reading frame of hITLN-1 was cloned by reverse transcription-PCR. Primers were designed on the basis of a reported hITLN-1 sequence, 5′-CCACTAGTATTACAATGAACCAACTCAGCTTCC-3′ and 5′-CCTCTAGACTCTCAACGATAGAATAGAAGCACA-3′ (18). cDNA was synthesized using previously isolated hITLN-1, cloned from human small intestinal total RNA (18). A single band was amplified by PCR using these primers, cDNA, and Platinum high-fidelity Taq DNA polymerase (Life Technologies) (35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 60 s). The PCR product was digested with the enzymes SpeI and XbaI and then cloned into the pTracer vector (Life Technologies). The DNA sequence was confirmed by the University of Georgia Molecular Genetics Instrumentation Facility.

Large amounts of the pTracer-hIntL-1 expression vector were generated in Escherichia coli and purified using the PureLink HiPure Expi Plasmid Megaprep kit (Invitrogen). The pTracer_hIntL-1 vector was then introduced into HEK-293F suspension cultures (FreeStyle 293-Fcells; Thermo Fisher Scientific) maintained at 2.5 × 10⁶ to 3.0 × 10⁶ cells/mL in a humidified CO₂ platform shaker incubator at 125 rpm at 37°C and 50% humidity. Transient transfection was performed using HEK293-F cells in the expression medium comprised of a 9:1 ratio of Freestyle 293 expression medium (Thermo Fisher Scientific) and EX-Cell expression medium including GlutaMAX (Sigma-Aldrich). Transfection was initiated by the addition of plasmid DNA and polyethyleneimine as the transfection reagent (linear 25-kDa polyethyleneimine; Polysciences, Inc.) at a ratio of 4:9. Twenty-four hours posttransfection, the cell cultures were diluted with an equal volume of fresh medium supplemented with valproic acid (2.2 mM final concentration), and protein production was continued for an additional 5 days at 37°C. The cell cultures were harvested, clarified by sequential centrifugation at 1,200 rpm for 10 min and 3,500 rpm for 15 min at 4°C, and passed through a 0.8-μm filter (Millipore).

For the purification of hIntL-1, the culture supernatant was treated with protease inhibitor tablets (cOmplete, Mini, EDTA-free; Roche Diagnostics GmBH) and DNase/RNase (Turbonuclease from Serratia marcescens; Sigma-Aldrich) and adjusted to 10 mM CaCl₂. The purification was performed at 4°C using a β-galactofuranose agarose column equilibrated with 20 mM Tris, pH 7.4, 150 mM NaCl, 0.1% Tween 20, and 10 mM CaCl₂. hIntL-1 in the culture supernatant was allowed to bind to the column resin during a 30-min batch incubation before being slowly passed through the column. The column was washed using 20 mM Tris, pH 7.4, 150 mM NaCl, and 10 mM CaCl₂ and hIntL-1 was eluted in 20 mM Tris, pH 7.4, 150 mM NaCl, and 10 mM EDTA. The resulting eluent was concentrated using a 30-kDa molecular-weight-cutoff Ultra-15 centrifugal filter unit (Amicon) and subsequently dialyzed into 20 mM HEPES, pH 7.4, 150 mM NaCl, 500 mM d-glucose (control buffer for samples without hIntL-1). The purity of hIntL-1 was assessed by SDS-PAGE and Coomassie staining, and the concentration was measured using absorbance at 280 nm with a hIntL-1-specific extinction coefficient of ε = 237 400 cm⁻¹ M⁻¹ and molecular mass of 101,400 Da (21).

Agglutination assay.

Bacterial strains were grown on their respective media and harvested into binding buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 10 mM CaCl₂) or EDTA buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 10 mM EDTA). Next, 100 μL of the bacterial suspension adjusted to an optical density at 600 nm (OD₆₀₀) of 2 was added per well of a round-bottom 96-well plate (Cooke Microtiter). hIntL-1 at various concentrations, or equivalent volumes of control buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 500 mM d-glucose), were added. The plate was incubated for 2 h at 4°C under agitation before allowing the bacteria to settle overnight at 4°C.

Fluorescence microscopy with labeled hIntL-1.

The fluorescent labeling of the purified hIntL-1 was done using the Alexa Fluor 488 microscale protein labeling kit (Thermofisher Scientific), resulting in a final concentration of labeled hIntL-1 of 0.2 μg/μL at a calculated average of two molecules of Alexa Fluor per trimer of hIntL-1.

For the preparation of microscopy samples, bacteria were grown on their respective media. The harvested bacteria were washed once in binding buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 10 mM CaCl₂) before adjusting their OD₆₀₀ to 0.2. Next, 100 μL of bacterial suspension was incubated with 1.5 μg hIntL-1 or control buffer (20 mM HEPES, pH 7.4, 150 mM NaCl, 500 mM d-glucose) for 2 h at 4°C in the dark and under agitation. Subsequently, the bacteria were centrifuged at 8,000 rpm for 4 min, washed three times with binding buffer, and resuspended to 15 μL. The resulting bacterial suspension was applied to a coverslip, dried at 37°C, and heat fixed. The slides were stained with DAPI (2.5 μg/mL), rinsed in H₂O, and mounted onto a microscope slide with 5 μL Vectashield antifade mounting medium (Vector Laboratories). The margins of the coverslip were sealed with nail polish before storage at 4°C in the dark. Confocal microscopy was executed on a Zeiss LSM 710 at ×40 magnification.

Biofilm formation assay.

The conditions for the biofilm formation assay with S. pneumoniae were adapted from a previously described method (51). S. pneumoniae serotypes 8 and 43 were grown in THY to mid-exponential phase (OD₆₀₀ of approximately 0.6). The cultures were diluted at 1:20 with THY. In a 96-well plate, 200 μL of bacteria suspension per well was incubated with 45 μg/mL of hIntL-1 or its control buffer for 24 h at 37°C. The bacterial suspensions were removed from the wells by vigorous inversion of the plate, followed by a wash in H₂O and drying of the plate. All washes in the assay were executed by submerging the 96-well plate in H₂O entirely, followed by the emptying of all wells simultaneously via vigorous inversion of the plate. The biofilms were stained with 250 μL/well of a 1% crystal violet solution for 15 min at room temperature (RT). Subsequently, the plate was washed twice and left to dry. To dissolve the crystal violet stain, 200 μL of 30% acetic acid was added per well and incubated for 15 min. Next, 150 μL of the stained acetic acid solution was transferred into a fresh 96-well plate for photometric analysis using the Synergy H1 microplate reader at an absorbance of 550 nm (BioTek Instruments, Inc.). The assay was executed as five biological replicates; means and standard errors of the means are displayed in the graph.

Attachment to A549 lung cells.

The A549 lung epithelial cell line was cultured in Dulbecco’s modified Eagle medium (DMEM) containing 4.5 g/liter glucose and l-glutamine, without pyruvate (Corning) replaced by 10% fetal bovine serum (FBS) (VWR), 100 U/mL of penicillin, and 100 μg/mL of streptomycin (Gibco).

The attachment assay was performed as described previously (52). Briefly, A549 cells between passage numbers 6 and 20 were used. The cells were seeded into 24-well plates with approximately 1 × 10⁵ cells/well at 95 to 100% confluence. Two hours before the assay, the cells were washed twice with phosphate-buffered saline (PBS) before 500 μL DMEM with 5 mM CaCl₂ was added. S. pneumoniae was grown to mid-exponential phase and adjusted to 1 × 10⁹ CFU/mL in DMEM with 5 mM CaCl₂. To start the attachment assay, an appropriate amount of medium was removed from each well to yield a final reaction volume of 200 μL/well. Next, 45 μg/mL hIntL-1 and 1 × 10⁷ CFU of bacteria were added for an MOI of 100:1 (bacteria to A549 cells). The plate was incubated for 2 h at 37°C with 5% CO₂. Each well was washed 3× with PBS, followed by the addition of 100 μL trypsin-EDTA solution (Sigma) and incubation at 37°C for 5 min. To detach the cells, the suspension was pipetted up and down vigorously before serial diluting and plating for determination of number of CFU per milliliter. Samples without A549 cells served as 100% attachment controls. These samples were diluted and plated directly after the 2-h incubation period. Shown are means and standard errors of the means for 4 biological replicates.

Microscopy of A549 cells with Hema3 stain.

The attachment assay for microscopy was executed as described above, with the following changes. A549 cells were seeded into 24-well plates with sterile glass slides at the bottom. The attachment assay was done at 75% confluence for better visualization of the cellular margins. After the attachment period of 2 h at 37°C, paraformaldehyde was added to a final concentration of 5% and incubated at RT for 30 min. The wells were washed 3× with PBS before the glass slides were removed, rinsed with H₂O, and dried. The Hema3 Staining System (Fisher Scientific) was used by following the manufacturer’s instructions. To stain the samples, three dips in each staining reagent were performed. The slides were imaged using a Zeiss AxioScope A1 at ×100 magnification.

Serum survival assay.

A serum survival assay was performed to assess the ability of S. pneumoniae to resist complement-mediated killing in the presence of hIntL-1 (53, 54). Pooled human complement serum was purchased from Innovative Research. Prior to the serum survival assay, the presence of human antibodies specific to S. pneumoniae serotypes 8 and 43 was verified by Western blotting. Briefly, bacterial whole-cell lysates were separated by SDS-PAGE and transferred onto a nitrocellulose membrane. The membrane was blocked overnight at 4°C in 5% bovine serum albumin (BSA). The pooled human serum was used as the primary antibody (1:1,000) and mouse-anti-human IgG4 Fc, horseradish peroxidase (HRP) (1:2,000; Invitrogen), as the secondary antibody. The serum survival assay was largely executed as described by Lees-Miller et al. (54), with some modifications. Briefly, S. pneumoniae was grown to mid-exponential phase in THY and was brought to an OD₆₀₀ of 1. In a noncoated 96-well plate, 15 μL bacterial suspension (for a final OD₆₀₀ of 0.1) was combined with hIntL-1 or buffer (for a final concentration of 45 μg/mL hIntL-1) and brought to a final volume of 150 μL with normal human serum (NHS) or heat-inactivated human serum (HIS). HIS was prepared by incubating NHS for 1 h at 56°C. Tenfold serial dilutions of all samples were prepared in 150 μL NHS or HIS. The 96-well plate was incubated at 37°C for 2 h before plating on TSA plates. The survival of S. pneumoniae in NHS in the presence and absence of hIntL-1 was calculated as a percentage of its survival in HIS. Four biological replicates of the experiment were performed, and the resulting mean and its standard error are displayed. The serum survival assay was verified using the Acinetobacter baumannii 5075 wild-type and A. baumannii 5075 ΔpglC strains for which survival and complete killing, respectively, were published previously (53). Under the exposure to normal human complement serum for 2 h as described for the assay above, 37.2% survival was observed for A. baumannii WT and complete killing for the ΔpglC mutant.

Isolation of fresh human PBMCs.

Whole fresh human blood was drawn from adult volunteers at the Health Center or the Clinical Translational Research Unit of the University of Georgia under informed consent according to procedures approved by the Institutional Review Boards at the University of Georgia (UGA number 2012-10769-06). For cell isolation, blood was drawn into EDTA-coated tubes. Peripheral blood mononuclear cells (PBMCs) were isolated using a gradient of Histopaque-1077 Hybri-Max (H8889-100ML; Sigma Life Science). Briefly, equal amounts of blood were layered onto Histopaque 1077 and centrifuged at 350 × g for 30 min with the centrifuge deceleration set to 1. The resulting PBMC layer was collected and washed twice with PBS before resuspending in DMEM. The cells were stained with methylene blue for quantification via hemocytometer.

Isolation of autologous serum.

Whole fresh human blood was drawn from adult volunteers at the Health Center or the Clinical Translational Research Unit of the University of Georgia under informed consent according to procedures approved by the Institutional Review Boards at the University of Georgia (UGA number 2012-10769-06). Next, 10 mL of blood was drawn into a silicone-coated tube without anticoagulant, and the blood was allowed to clot at RT for 30 min. The tube was centrifuged twice at 2,000 × g for 5 min, and the resulting serum was stored on ice until use.

Fluorescent microscopy of S. pneumoniae with pneumococcus antiserum pools.

S. pneumoniae was grown fresh to mid-exponential phase in THY. The bacteria were harvested and brought to an OD₆₀₀ of 0.35. Next, 40 μL of bacterial suspension was incubated with 2 μL pneumococcal antiserum pool B or D or H₂O for 20 min at RT under agitation (Cedarlane Laboratories). Next, 20 μL was applied to a microscopy coverslip, air dried, and heat fixed. The bacteria were stained with DAPI (2.5 μg/mL final concentration) in 5 μL Vectashield antifade mounting medium (Vector Laboratories) and sealed to a microscopy slide using nail polish. Confocal microscopy was executed on a Zeiss LSM 710 at ×100 magnification.

Opsonophagocytosis assay.

S. pneumoniae was grown to mid-exponential phase in THY prior to the assay. Bacterial stock solutions were prepared in DMEM containing 5 mM CaCl₂. The OPA was executed at an MOI of 0.05 in an uncoated 96-well plate. A total of 3.3 × 10⁴ CFU, 45 μg/mL hIntL-1, or its buffer and 6.6 × 10⁵ PBMCs were combined and filled to a reaction volume of 100 μL with DMEM containing 5 mM CaCl₂. Where indicated, 1.3 μL Pneumococcus antiserum pool B and pool D (Cedarlane Laboratories) were added as a control for S. pneumoniae agglutination. During the assay, samples were incubated at 37°C for 2 h under 5% CO₂ before serial dilution in DMEM containing 5 mM CaCl₂ and plating for CFU quantification. The percentage of survival was calculated relative to samples to which heat-killed PBMCs were added (15 min at 95°C). For the OPA in the presence of 20% serum, 20% autologous serum was added to all samples and controls. All experiments were executed at least in triplicates; means and standard errors of the means are depicted.

Neutrophil killing assay.

Whole fresh human blood was drawn from adult volunteers at the Health Center or the Clinical Translational Research Unit of the University of Georgia under informed consent according to procedures approved by the Institutional Review Boards at the University of Georgia (UGA number 2012-10769-06). For cell isolation, blood was drawn into EDTA-coated tubes. For the neutrophil killing assay, human PMNs were isolated from the peripheral blood of healthy, consenting volunteers using the EasySep Direct Human PMN isolation kit (Stem Cell Technologies). Isolated PMNs were >98% viable as assessed by Trypan blue staining and >95% pure as determined by CD66b flow cytometry. The assay was adapted from previously described methods (55, 56).

S. pneumoniae serotypes 8 and 43 were grown to mid-exponential phase in THY and adjusted to 1 × 10⁷ CFU/mL in 1× Hanks’ balanced salt solution (HBSS) containing 5% autologous serum, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂. Purified human PMNs were washed and resuspended in assay medium (1× HBSS containing 5% [vol/vol] autologous serum of the PMN donor, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂). For samples with differing serum conditions, the PMNs were washed 3 times in serum-free assay medium and resuspended in assay medium containing either no serum or 5% heat-inactivated serum (1 h at 56°C for complement inactivation). The assay was executed as a batch-killing assay with no prior opsonization of the bacteria. Eppendorf tubes (1.5 mL) were blocked with 500 μL 5% BSA in PBS followed by a wash with PBS. A total of 4 × 10⁶ PMNs were combined with 45 μg/mL hIntL-1 or an equivalent volume of hIntL-1 buffer (20 mM HEPES, 150 mM NaCl, 500 mM glucose, pH 7.4). For the agglutination controls, 1.3 μL Pneumococcus antiserum pool B and pool D (Cedarlane Laboratories) was used. Assay medium was used to reach a volume of 95 μL before the addition of 5 μL bacterial suspension containing 4 × 10⁴ CFU to start the assay at an MOI of 0.01. The samples were incubated for 2 h at 37°C, during which the tubes were mixed every 5 to 10 min to prevent the settling of the PMNs. To lyse the PMNs, 20 μL of each sample was incubated with 180 μL ice-cold saponin (1 mg/mL in HBSS) for 5 min, followed by a dilution series in HBSS on ice. Samples were plated onto TSA plates containing 5% sheep blood and incubated at 37°C overnight before the quantification of survival via CFU count. The percentage of survival was calculated relative to control samples that did not contain PMNs. For samples containing no serum, the percentage of survival was calculated relative to control samples that contained heat-killed PMNs (20 min at 65°C). All experiments were executed at least in triplicate; shown are means and standard errors.

Measurement of NET release.

NET production by PMNs was determined using 0.2% Sytox Orange nucleic acid stain (Thermo Fischer Scientific). The assay was adapted from a method described previously (55, 57). In a 96-well black transparent-bottom plate, 250,000 PMNs were combined with 45 μg/mL hIntL-1 and assay medium (1× HBSS containing 5% autologous serum, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂). Fluorescence measurements were taken for 8 h at 37°C using a Varioskan Flash microplate luminometer (Thermo Scientific, Waltham, MA, USA) (excitation, 530 nm; emission, 590 nm). The relative fluorescence units (RFU) were accumulated over the entire 8 h, normalized to PMNs activated with PMA (100 nM, positive control), and are shown as percent. The experiment was performed in five biological replicates.

Measurement of ROS release.

Production of ROS by PMNs was analyzed using the Diogenes-based chemiluminescence kit (National Diagnostics, Atlanta, GA). The assay was adapted from a method described previously (55, 58, 59). A total of 250,000 PMNs and 45 μg/mL hIntL-1 were combined in assay medium (1× HBSS containing 5% autologous serum, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂) in a solid white 96-well plate. Luminescence measurements were taken for 2 h at 37°C using a Varioskan Flash microplate luminometer (Thermo Scientific, Waltham, MA, USA). The relative luminescence units (RLU) were accumulated over the entire 2 h and normalized to PMNs activated with PMA (100 nM) and are shown as percent. The experiment was performed in seven biological replicates.

Measurement of neutrophil elastase release.

The release of enzymatically active neutrophil elastase by PMNs was measured using the neutrophil elastase activity assay kit (Cayman Chemical) by following the manufacturer’s instructions. First, 250,000 PMNs and 45 μg/mL hIntL-1 were combined in assay medium (1× HBSS containing 5% autologous serum, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂) in a black clear-bottom 96-well plate. Fluorescence measurements were then taken for 2 h at 37°C using a Varioskan Flash microplate luminometer (Thermo Scientific, Waltham, MA, USA). Relative fluorescence units were accumulated over the entire 2 h, and activity of neutrophil elastase (in microunits per milliliter) were calculated using a standard curve. The experiment was performed in six biological replicates.

Flow cytometry.

S. pneumoniae was stained with 5 mM pHrodo iFL green STP ester (Thermofisher) in assay medium (1× HBSS containing 5% autologous serum, 5 mM glucose, 10 mM HEPES, and 5 mM CaCl₂) with 45 μg/mL hIntL-1 or buffer for 1 h at 37°C. Bacteria were added to 1 × 10⁶ PMNs in assay medium at an MOI of 0.1 (bacteria to PMN) in the presence or absence of 45 μg/mL hIntL-1. The infection mixture was incubated for 1 h at 37°C with regular perturbation to avoid settling. Samples were washed and resuspended in PBS before staining with the Zombie Aqua Fixability kit (BioLegend) at a final dilution of 1:5,000 for 15 min at RT. The samples then were washed and resuspended in 1% BSA in PBS before incubation with Fc blocking reagent (Miltenyi Biotec) at a final dilution of 1:100 for 10 min at RT, followed by the addition of Alexa Fluor 647 mouse anti-human CD66b antibody (BD Biosciences) to a final concentration of 1 μg/mL and further incubation at RT for 25 min. In the final step, all samples were washed and resuspended in 300 μL 1× BD stabilizing fixative. During the entire preparation, samples were kept in the dark. For flow cytometry, the same number of PMNs was analyzed for all conditions. The following lasers and filters were used during analysis: Zombie Aqua at 405 nm with the 530/30 filter, CD66B Alexa Fluor at 647 nm with the 660/20 filter, and pHrodo green at 488 nm with the 530/30 filter. The flow cytometry experiments were performed at the University of Georgia College of Veterinary Medicine Cytometry Core Facility using a NovoCyte Quanteon 4025 with NovoSamplerQ utilizing NovoExpress software v.1.4.1. The gating strategy is shown for one representative sample. For each sample, an identical control was included that was left on ice during infection to inhibit phagocytosis, and all controls showed less than 20% PMN positive for bacterial uptake. The assay was performed in four biological replicates; means and standard errors are shown.

Microscopy and quantification of bacterial phagocytosis.

To visualize and quantify bacterial phagocytosis by PMNs, the neutrophil killing assay was executed as described above, with the following changes. S. pneumoniae was stained with the pHrodo green AM intracellular pH indicator (Thermo Fisher Scientific) for 30 min at 37°C by following the manufacturer’s instructions. Next, 4 × 10⁷ bacteria and 4 × 10⁶ PMNs were incubated at an MOI of 5 (bacteria to PMN) for better visualization of the bacteria. After the 2-h incubation period at 37°C, paraformaldehyde was added to the samples to a final concentration of 5% for 15 min at RT. Samples were applied to a microscopy slide, dried at RT, rinsed in H₂O, and stained with DAPI. During the assay and microscopy preparation, samples were kept in the dark. Under equal exposure conditions, neither pure PMNs nor pHrodo-labeled S. pneumoniae showed any green fluorescence. Confocal microscopy was performed on a Zeiss LSM 710 at ×100 magnification. For quantification of bacterial phagocytosis, the microscopy images were analyzed for green fluorescence intensity with ImageJ software. A total of 150 PMNs for each condition from three independent replicates were analyzed; shown are means and standard errors.