key: cord-0023310-814zcrbb
authors: Mursalin, Md Huzzatul; Coburn, Phillip S.; Miller, Frederick C.; Livingston, Erin T.; Astley, Roger; Callegan, Michelle C.
title: C-X-C Chemokines Influence Intraocular Inflammation During Bacillus Endophthalmitis
date: 2021-11-16
journal: Invest Ophthalmol Vis Sci
DOI: 10.1167/iovs.62.14.14
sha: dbfbb30280a3d55b2403dc102c49df0b8e37e25c
doc_id: 23310
cord_uid: 814zcrbb

PURPOSE: The purpose of this study was to explore the C-X-C chemokines CXCL2 and CXCL10 as potential anti-inflammatory targets for Bacillus endophthalmitis. METHODS: Bacillus endophthalmitis was induced in C57BL/6J, CXCL2(−/−), and CXCL10(−/−) mice. At specific times postinfection, eyes were analyzed for Bacillus, retinal function, and inflammation. The efficacies of intravitreal anti-CXCL2 and anti-CXCL10 with or without gatifloxacin in B. cereus endophthalmitis were also assessed using the same techniques. RESULTS: Despite similar Bacillus growth in eyes of C57BL/6J, CXCL2(−/−), and CXCL10(−/−) mice, retinal function retention was greater in eyes of CXCL2(−/−) and CXCL10(−/−) mice compared to that of C57BL/6J mice. Neutrophil migration into eyes of CXCL2(−/−) and CXCL10(−/−) mice was reduced to a greater degree compared to that of eyes of C57BL/6J mice. Infected CXCL2(−/−) and CXCL10(−/−) mouse eyes had significantly less inflammation compared to that of C57BL/6J eyes. Retinal structures in infected eyes of CXCL2(−/−) mice were preserved for a longer time than in CXCL10(−/−) eyes. Compared to untreated eyes, there was less inflammation and significant retention of retinal function in eyes treated with anti-CXCL2 and anti-CXCL10 with or without gatifloxacin. CONCLUSIONS: For Bacillus endophthalmitis, the absence of CXCL2 or CXCL10 in mice resulted in retained retinal function and less inflammation. The absence of CXCL2 led to a better clinical outcome than the absence of CXCL10. The use of anti-CXCL2 and anti-CXCL10 limited inflammation during B. cereus endophthalmitis. These results highlight the utility of CXCL2 and CXCL10 as potential targets for anti-inflammatory therapy that can be tested in conjunction with antibiotics for improving treating Bacillus endophthalmitis.

Several components of the host and bacterium contribute to the pathogenesis of experimental Bacillus endophthalmitis. 12, 14 The toxicity of Bacillus secreted products, quorum sensing regulation of virulence factors, flagelladirected motility, pili-mediated adhesion, and S-layermediated inflammation, all impact the pathogenesis of Bacillus endophthalmitis to various degrees. [15] [16] [17] [18] [19] [20] [21] [22] From the host's perspective, activation of Toll-like receptors TLR4 and TLR2, but not TLR5, are responsible for rapid intraocular inflammation during infection. [23] [24] [25] During ocular infection, TLRs interact with microbial ligands, resulting in an acute and rapidly evolving inflammatory response in the eye. 24, 25 Activation of TLRs 2 and 4 and signaling through downstream adaptor proteins MyD88 and TRIF are critical in the response to Bacillus endophthalmitis. 26, 27 This signaling induces the expression of mediators that attract inflammatory cells to the infection site, primarily polymorphonuclear leukocytes (PMNs). These inflammatory mediators are often produced in parallel with a deteriorating clinical outcome in other endophthalmitis models and in the human eye infected with Bacillus. 28, 29 Numerous proinflammatory genes are expressed and mediators are synthesized during experimental Bacillus endophthalmitis. 22, [28] [29] [30] [31] The production of cytokines, including macrophage inflammatory protein 1 alpha (MIP-1α), interleukin-1 beta (IL-1β), IL-6, tumor necrosis factor alpha (TNF-α), and the chemokine KC (C-X-C motif ligand 1 [CXCL1]), have been reported to increase in parallel with blood retinal barrier (BRB) permeability and inflammatory cell influx. 22, [30] [31] [32] [33] To date, TNF-α and CXCL1 have been reported to be important to the immune response to experimental Bacillus endophthalmitis. 31 , 33 We also reported that macrophage inflammatory protein 2alpha (CXCL2 or MIP2-α) and interferon gamma-induced protein 10 (CXCL10 or IP-10) were highly upregulated during Bacillus endophthalmitis, and several of the aforementioned mediators were blunted when innate activation was blocked or absent. 22 CXCL2 and CXCL10 are highly homologous and share comparable functions to CXCL1. 34, 35 Detecting invading pathogens and clearing them as quickly as possible are the primary functions of the innate inflammatory response. 36, 37 Bacillus is highly inflammogenic and the host-mediated inflammatory response in the eye is often so robust that it is challenging to control, eventually resulting in significant loss of vision, blindness, or enucleation. Anti-inflammatory therapeutics and their use have not proven to be very effective at improving outcomes in this disease. [38] [39] [40] [41] Here, we explored the role of the C-X-C chemokines CXCL2 and CXCL10 in experimental Bacillus endophthalmitis. Previously, we reported improved and positive clinical outcomes in CXCL1-deficient mice and mice treated with anti-CXCL1 during Bacillus endophthalmitis. 31 Therefore, we tested the hypothesis that the absence or inhibition of CXCL2 or CXCL10 would also reduce the inflammation and infection severity, highlighting these mediators as potential therapeutic targets to treat Bacillus endophthalmitis. Our results suggested that the absence of CXCL2 or CXCL10 delayed overall pathogenesis and significantly minimized disease severity, as did the treatment of infected eyes with anti-CXCL2 or anti-CXCL10. These results identify CXCL2 and CXCL10 as potential anti-inflammatory targets for the treatment of Bacillus and possibly other types of endophthalmitis.

C57BL/6J (000664) mice and breeding pairs of CXCL2 −/− (C57BL/6NJ-Cxcl2 em1(IMPC) J/J, 029557) and CXCL10 −/− mice (B6.129S4-Cxcl10 tm1adl /J, 006087) were purchased from the Jackson Laboratory. CXCL2 −/− and CXCL10 −/− mice were bred on the C57BL/6J background. All breeders were kept on a 12-hours on/off light cycle in a barrier facility. Weaned or vendor-supplied animals were accustomed to housing under biosafety level 2 conditions and were co-housed for at least 2 weeks to equilibrate their microbiota. Mice were used in experiments at 8 to 10 weeks of age. All procedures were conducted according to guidelines and recommendations of the Guide for the Care and Use of Laboratory Animals, the University of Oklahoma Health Sciences Center (OUHSC) IACUC, and the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research.

Mouse primary neutrophils were isolated from bone marrow of euthanized C57BL/6J, CXCL2 −/− , and CXCL10 −/− mice by using a neutrophil isolation kit (MACS, Miltenyl Biotech, Gladbach, Germany) according to the manufacturer's instructions. Bone marrow from the C57BL/6J, CXCL2 −/− , and CXCL10 −/− harvested femurs was collected into 3 different 50 mL Falcon tubes containing RPMI media (GIBCO/Thermo Fisher Scientific, Waltham, MA, USA) with 10% FBS (Sigma Aldrich, St. Louis, MO, USA) using a 10 mL syringe. The bone marrow was centrifuged at 100 g for 10 minutes and washed with wash buffer (PBS, pH 7.2, 0.5% bovine serum albumin [BSA], and 2 mM EDTA). Cells were counted manually using a hemocytometer. Then, 200 μL of wash buffer and 50 μL neutrophil biotin-antibody cocktail were added to the tubes for every 5 × 10 7 cells. After mixing, cells were incubated for 10 minutes at 4°C. Cells were washed and the pellet resuspended in 400 μL of wash buffer and 100 μL of anti-biotin microbeads. Cells were mixed and incubated at 4°C for 15 minutes. Cells were washed and resuspended to a concentration 10 8 cells in 500 μL of buffer. For magnetic separation, a MACS column and separator of appropriate size was chosen according to the number of total cells and number of neutrophils. LS columns (Miltenyi Biotech) were placed inside the MACS separator. A 15 mL tube was placed under each LS column and the columns were washed with 3 mL of wash buffer. Once the wash buffer was completely removed, the buffer was discarded and new 15 mL tubes were placed under the columns. The entire sample (500 μL) was loaded onto the column. The columns were washed three times with 3 mL of wash buffer. The washed cells were collected, counted, centrifuged at 100 × g for 10 minutes, and resuspended in RPMI medium. 42

Primary neutrophils from C57BL/6J, CXCL2 −/− , and CXCL10 −/− mice were used in a gentamicin exclusion assay 42 to assess the impact of the absence of CXCL2 and CXCL10 on phagocytosis. Approximately 1 × 10 5 of cells from each group of mice were incubated at 20 MOI (∼2 × 10 6 ) with Bacillus cereus ATCC 14579 for 90 minutes. For each group of mice, one aliquot of cells was washed and then treated with 200 μg/mL gentamicin for 60 minutes to kill any extracellular bacteria. The gentamicin-treated cells were centrifuged, washed to remove residual antibiotic, and lysed with 0.5% Triton X-100. The gentamicin-treated group contained only intracellular bacteria. A second aliquot of cells from each group of mice was centrifuged, washed, and lysed with 0.5% Triton X-100. The second group contained both intra-and extracellular bacteria. Equal numbers of Bacillus cereus (approximately 2 × 10 6 in 2 mL) were incubated with 200 μg/mL gentamicin for 60 minutes as a control. CFU were enumerated by serial dilution and plating. 42

All in vivo experiments were performed with C57BL/6J, and CXCL2 −/− , CXCL10 −/− mice on the C57BL/6J background. Mice were housed in the animal facility as described above and were 8 to 10 weeks of age at the time of the experiments. Mice were sedated using a combination of ketamine (85 mg/kg body weight; Ketathesia, Henry Schein Animal Health, Dublin, OH, USA) and xylazine (14 mg/kg body weight; AnaSed, Akorn Inc., Decatur, IL, USA). C57BL/6J and CXCL2 −/− or CXCL10 −/male and female mice were infected with 100 CFU Bacillus cereus/0.5 μL BHI into the right eye using a sterile glass capillary needle, as previously described. 21, 28, 43 At various times postinfection, electroretinography was performed prior to euthanasia by CO 2 inhalation, and then eyes were harvested for quantitation of viable intraocular bacteria, retinal function, and PMN infiltration, and analysis of ocular architecture by histology, as described below.

Electroretinography (ERG) was used to quantify percent retinal function retained, as previously described. 20, 21, 24, 25, 28, 42 After infection, mice were dark adapted for 6 hours. Mice were then sedated as described above, and pupils were dilated with topical phenylephrine (Akorn, Inc.). Gold wire electrodes were placed onto each cornea. Reference electrodes were attached to the forehead and tail. Eyes were then stimulated by five flashes of white light (1200 cd s/m2) and retinal responses were recorded as A-wave (retinal photoreceptor cell function) and B-wave (Muller cell, bipolar cell, and second order neuronal function) amplitudes for infected eyes and compared with the uninfected eyes of the same animal (Espion E2 software; Diagnosys LLC, Lowell, MA, USA).

As previously described, harvested eyes from euthanized mice at specific time points were homogenized in 400 μL PBS with sterile 1-mm glass beads (BioSpec Products, Inc., Bartlesville, OK, USA), serially diluted 10-fold in PBS, and plated onto BHI agar plates. 20, 21, 24, 25, 28, 44 Histology Eyes were harvested from euthanized mice at specific time points and incubated in High Alcoholic Prefer fixative for 30 minutes, and then transferred to 70% ethanol. Paraffin-embedded eyes were sectioned and stained with hematoxylin and eosin (H&E). 20, 21, 25, 27, 28 

Inflammatory cell influx was assessed by measuring myeloperoxidase concentrations using a sandwich ELISA (Hycult Biotech, Plymouth Meeting, PA, USA), as previously described. At various time points, eyes were harvested, transferred into PBS supplemented with proteinase inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA), and homogenized using 1-mm sterile glass beads (BioSpec Products, Inc.). Uninfected eye homogenates were the negative controls. The lower limit of detection for this assay was 2 ng/mL. 24, 42 Neutralization of CXCL2 or CXCL10 In Vivo C57BL/6J mice were anesthetized using isoflurane and intravitreally infected with 100 CFU B. cereus, as described above. At 2 hours postinfection, mice were anesthetized again with isoflurane and randomly parsed into the following treatment groups: anti-CXCL2 monoclonal antibody (anti-CXCL2; 250 ng/0.5 μL anti-CXCL2/MIP-2 IgG, AF-452-NA clone 48415; R&D Systems), anti-CXCL10 monoclonal antibody (anti-CXCL10; 250 ng/0.5 μL anti-CXCL10/IP-10 IgG, clone 134013, MAB466; R&D Systems), 0.5 μL PBS containing 1.25 μg gatifloxacin (GAT; Zymaxid 0.5%; Allergan, Inc., Irvine, CA, USA), 0.5 μL PBS containing 1.25 μg gatifloxacin and 250 ng of anti-CXCL2 (GAT + anti-CXCL2), or 0.5 μL PBS containing 1.25 μg gatifloxacin and 250 ng of anti-CXCL10 (GAT + anti-CXCL10), or isotype control antibody (Isotype; 0.5 μg nonspecific control IgG2A/0.5 μL MAB 006, clone 54447l; R&D Systems). Another group of mice was left untreated to serve as controls. 31, 45 Bacterial growth, retinal function, and intraocular inflammation were assessed at 10 hours postinfection, as described above.

Chemokines and chemokine receptors are known to regulate neutrophil recruitment. Destruction of microorganisms as well as other foreign particles by phagocytosis is the primary function of neutrophils. 46 To determine whether the absence of CXCL2 or CXCL10 altered neutrophil function, a phagocytosis (gentamicin [Gen] exclusion) assay was used (Fig. 1 ). No differences in internalization of B. cereus were observed among neutrophils isolated from C57BL/6J, CXCL2 −/− (P = 0.1667), or CXCL10 −/− (P = 0.8254) mice. Incubation with gentamicin recovered no bacteria, indicating that B. cereus were susceptible to the antibiotic (data not shown). Together, these results demonstrated that an absence of CXCL2 or CXCL10 did not interfere with internalization of B. cereus by mouse primary neutrophils.

Intraocular B. cereus growth in CXCL2 −/− and CXCL10 −/− male and female mice was compared with that of C57BL/6J Overall, these results implied that the absence of CXCL2 or CXCL10 in mice did not alter the intraocular growth of B. cereus, suggesting that bacterial growth was independent of CXCL2 or CXCL10 production. Intraocular growth of B. cereus in these mice was also independent of sex at each time point. Fig. 4C ). To determine whether biological sex variability influenced these outcomes, we separated and analyzed data of male and female mice at each time point. Awave and B-wave retention was significantly higher in male CXCL2 −/− mice compared to CXCL2 −/− female mice at 8 and 12 hours postinfection only (P ≤ 0.05; see Figs. 3B, 3D). We did not observe any differences in A-wave and B-wave retention between male and female CXCL10 −/− mice or C57BL/6J mice (P ≥ 0.05; Figs. 4B, 4D). Together, these results demonstrated that the CXCL2 −/− and CXCL10 −/− eyes infected with B. cereus retained higher retinal function compared to C57BL/6J eyes, suggesting that absence of CXC chemokines influenced the retention of retinal function during experimental endophthalmitis.

Histological sections of infected and uninfected eyes of CXCL2 −/− , CXCL10 −/− , and C57BL/6J male and female mice are depicted in Figure 5 . Infected eyes were harvested, fixed, sectioned, and stained with H&E at 8, 10, 12, 14, and 16 hours postinfection. Uninfected CXCL2 −/− and CXCL10 −/− male/female mice eyes looked morphologically and architecturally similar to eyes of C57BL/6J male/female mice, with no obvious alterations in retinal or corneal cellular morphology and without signs of inflammation. At 8 hours postinfection in C57BL/6J male/female eyes, significant increase of infiltrating cells and fibrin in the posterior segment were detected. In these eyes, anterior chambers were obstructed with fibrin and retinas were also partially detached. On the other hand, ocular architectures were well-preserved in infected CXCL2 −/− and CXCL10 −/− male/female eyes. These eyes exhibited minimal inflammation and fibrin deposition, distinguishable retinal layers, and intact retinas. At 10 hours postinfection, inflammatory cell infiltration and significant fibrin deposition throughout the vitreous were observed in infected C57BL/6J male/female eyes. Indistinguishable retinal architecture, complete retinal detachments, and highly edematous corneas were observed in these eyes. In stark contrast, retinal layers were distinguishable and retinas were well-preserved in infected CXCL2 −/− and CXCL10 −/− male/female eyes. At 12, 14, and 16 hours postinfection, retinal architecture was completely lost and inflammation pervaded all ocular structures in infected C57BL/6J and CXCL10 −/− male/female eyes. In contrast, retinal layers and architecture remained well-preserved in infected CXCL2 −/− male/female eyes and these eyes looked similar to that of infected C57BL/6J eyes at 8 hours postinfection. By 16 hours postinfection, CXCL2 −/− male/female eyes were analogous to that of C57BL/6J and CXCL10 male/female eyes, as retinal architecture was completely lost. Overall, compared to C57BL/6J infected eyes during B. cereus endophthalmitis, neutrophil influx and the damage to retinal architecture in infected eyes of CXCL2 −/− mice was delayed. For infected FIGURE 2. Absence of CXCL2 and CXCL10 did not affect intraocular Bacillus growth. Intraocular B. cereus ATCC 14579 growth was not affected by absence of CXCL2 or CXCL10. Eyes of CXCL2 −/− , CXCL10 −/− , and C57BL/6J (C57) male and female mice were infected with 100 CFU B. cereus. Eyes were harvested at designated timepoints and B. cereus were quantified. No significant differences in B. cereus burden were observed among mouse strains (P ≥ 0.05) (A and C), and no difference was observed in intraocular bacterial burden between males and females (B and D). Values represent means ± SEM of n ≥ 6 eyes at each time point with at least 3 independent experiments. ns P ≥ 0.05. CXCL10 −/− mice, this delay was observed for a shorter period of time. In contrast to what was observed in wild type mice, ocular architecture was preserved for a longer period of time in the eyes of infected CXCL2 −/− mice than in CXCL10 −/− mice during B. cereus infection, supporting the contribution of both chemokines to inflammation and overall endophthalmitis pathogenesis.

We assessed the levels of inflammatory cell influx in C57BL/6J, CXCL2 −/− , and CXCL10 −/− eyes infected with B. cereus (see 

Because the absence of CXCL2 or CXCL10 in mice resulted in improved retinal function and resulted in reduced inflammation and inflammation-associated damage to the retina, we investigated the impact of CXCL2 and CXCL10 neutralization on disease severity. C57BL/6J mice were divided into five experimental groups. All groups were infected with B. cereus. At 2 hours postinfection, groups of infected eyes were treated with anti-CXCL2 or anti-CXCL10 antibody with or without GAT, or with nonspecific isotype IgG2A (isotype control). One group was left untreated to serve as controls. At 10 hours postinfection, eyes were analyzed for bacterial counts, inflammation, and retinal function (Figs. 7, 8) . GAT treatment completely sterilized the infected eyes in both CXCL2 and CXCL10 experimental groups. There were less B. cereus in the groups treated with anti-CXCL2 or anti-CXCL10 alone, but the differences were not significant (P ≥ 0.05; see Figs. 7A, 8A).The average A-and B-wave retention in both untreated and isotype treated experimental groups were approximately 14.5% (see Figs. 7B, 7C, 8B, 8C). Retention of A-wave and B-wave function was nearly 58% in infected eyes treated with anti-CXCL2 antibody alone (see Figs. 7B, 7C). A-wave and B-wave retention was approximately 85% when the eyes were treated with GAT alone (see Figs. 7B, 7C). In contrast, A-wave and B-wave retention was approximately 93% when infected eyes were treated with both GAT and anti-CXCL2 antibody (see Figs. 7B, 7C). Retinal function retention in these three groups was significantly higher than that of control eyes (P ≤ 0.05; see Figs. 7B, 7C). Similarly, in the anti-CXCL10 experimental groups, A-wave and B-wave retention in eyes treated with anti-CXCL10 alone, GAT alone, or GAT + anti-CXCL10-treated eyes was approximately 76%, 69%, and 94%, respectively, and were To assess whether PMN influx was altered following anti-CXCL2 and/or anti-CXCL10 treatment, MPO concentrations were measured in these eyes and compared with that of untreated eyes. MPO concentrations in untreated eyes were 29-fold higher compared to the uninfected eyes in both anti-CXCL2 and anti-CXCL10 experimental groups (see Figs. 7D, 8D). No differences in MPO concentrations between untreated and isotype control-treated eyes were observed. In contrast, MPO levels in infected eyes treated with anti-CXCL2 alone, GAT alone, or GAT + anti-CXCL2 were 52%, 64%, or 93% lower than that of untreated eyes, respectively (see Fig. 7D ). Compared to eyes treated anti-CXCL2 alone or GAT alone, the MPO level was reduced in GAT + anti-CXCL2-treated eyes by 86% and 82%, respectively (Fig. 7D) . Similarly, in anti-CXCL10 experimental groups, MPO concentrations in eyes treated with anti-CXCL10 alone, GAT alone, or GAT + anti-CXCL10 were approximately 64%, 81%, and 63% lower than that of untreated eyes, respectively (see Fig. 8D ). No differences in MPO levels between eyes treated with anti-CXCL10 alone, GAT alone, and GAT + anti-CXCL10 eyes were observed (P ≥ 0.05; see Fig. 8D ).

Histology revealed that control and isotype-treated eyes were greatly inflamed and severely damaged, as described above (see Figs. 7E, 8E). Retinal layers were detached from choroid and were indistinguishable. However, compared to untreated and isotype-treated eyes, eyes treated with GAT alone, anti-CXCL2 alone, GAT + anti-CXCL2, anti-CXCL10 alone, or GAT + anti-CXCL10 were similar in appearance, architecturally better-preserved, and had less inflammatory damage (see Figs. 7E, 8E). Taken together, these results demonstrated that treatment with an anti-CXCL2 or anti-CXCL10 antibody limited intraocular inflammation during B. cereus endophthalmitis. This suggests that CXCL2 and CXCL10 could potentially serve as possible anti-inflammatory therapeutic targets for endophthalmitis. 

Bacillus endophthalmitis is a devastating intraocular infection resulting in inflammation that frequently leads to blindness due to bystander damage to delicate ocular tissues. During Bacillus endophthalmitis, bacterial and host factors contribute to the development of a damaging inflammatory responses which interferes with ocular immune privilege by recruiting immune cells into the eye. 12, 29, 47, 48 Bacterial factors, such as cell wall components and secreted toxins, contribute to a loss of vascular permeability and overt damage to the retina. [15] [16] [17] [18] 21, 49 Host factors, such as innate receptors, their adaptors, and immune mediators, also significantly influence the extent of inflammation during this disease. [23] [24] [25] 27, [29] [30] [31] 50 With the increasing numbers of surgical procedures and intraocular injections for the treatment of ocular diseases, the incidence of cases of endophthalmitis will only increase. 4, 51, 52 Moreover, the current approaches to treat endophthalmitis are often unsuccessful in mitigating damaging inflammation. 40,53-56 Therefore, rational anti-inflammatory and anti-bacterial targets are needed to prevent the vision loss in this disease.

Among all the Gram-positive pathogens which cause endophthalmitis, Bacillus causes significant damage to ocular tissues in the shortest period of time. 12 Although members of the B. cereus sensu lato group (Bacillus thuringiensis and B. cereus) cause the most fulminant forms of bacterial endophthalmitis, other pathogens, such as Staphylococcus aureus, Streptococcus pneumoniae, and other streptococcal and enterococcal species can also cause sight-threatening endophthalmitis. 7,57-59 Cell wall peptido-glycan, S-layer, flagellar-driven motility, and secreted toxins play a key role in the pathogenesis of Bacillus endophthalmitis. [15] [16] [17] [18] 21, 49 TLR2 and TLR4 of the host also play a vital role in the pathogenesis of Bacillus endophthalmitis. 23, 25 We discovered that Bacillus S-layer protein was an activator of TLR2 and a novel activator of TLR4. 42 Downstream from TLRs, innate adaptors such as MyD88 and TRIF also impacted the pathogenesis of Bacillus endophthalmitis. 23 Bacterial products interact with and activate TLRs and their adaptors, resulting in the production of a plethora of immune mediators. 29, 37 Production of inflammatory mediators are linked with clinical signs, including anterior chamber inflammatory cells infiltration and fibrin deposition, vitreous exudate, and posterior synechiae (adhesion of the iris to the lens). 12, 29 We reported that inflammatory mediators CXCL1, TNFα, IL-1β, and IL-6 were detected as early as 4 hours postinfection, which correlated with fibrin accumulation in the anterior chamber and early neutrophil influx. 28 These reports showed that proinflammatory mediator synthesis was associated with neutrophil infiltration and subsequent decline of retinal function, which implied the importance of these inflammatory mediators as targets in directing ocular inflammation and overall endophthalmitis pathogenesis.

Chemoattractants are multifunctional mediators. Depending on the positioning of the first pair of N-terminal cysteine residues chemokines divided into two subfamilies (C-X-C and C-C chemokines). C-C chemokines tend to attract monocytes, whereas C-X-C chemokines recruit neutrophils. 34, 35 Chemokines are secreted by a broad range of cells, including immune cells such as neutrophils, macrophages, lymphocytes, natural killer cells, and thus play an important role in immune mediated inflammation. 60, 61 In the eye, these mediators are likely the products of receptor-agonist interactions from retinal pigment epithelium (RPE), astrocytes, corneal epithelium, iris epithelium, retinal microglia, and/or Muller cells. 29 We reported that the expression of C-X-C and C-C chemokines observed at 10 hours postinfection with Bacillus was TLR4-dependent. 22, 30 Of the 12 most highly upregulated chemokines reported in that study, 6 of them belonged to the C-X-C family. 22 S. aureus-mediated TLR2 activation of retinal microglia and triggering Muller glia to synthesize C-X-C chemokines. 62 To dissect the importance of these mediators and identify potential anti-inflammatory targets, we first investigated CXCL1. We reported that an absence of CXCL1, either by genetic knockout or treatment with neutralizing antibody, minimized inflammatory responses during endophthalmitis, but not completely. 31 In the current study, we investigated the role of CXCL2 and CXCL10 during Bacillus endophthalmitis, which were highly upregulated C-X-C chemokines during Bacillus endophthalmitis at 4 hours and 10 hours postinfection. 22, 30 Chemokines function in recruiting phagocytes and other immune cells to the area of infection, modulating inflammatory and wound healing responses, and directing cell differentiation, proliferation, and polarization. CXCL2 and CXCL10 mobilize cells by interacting with cell surface chemokine receptors CXCR2 and CXCR3, respectively. 60, 61, 63 Endogenous CXCL10 protected mice from viral infection by decreasing viral loads in corneal tissue, and thus the absence of this chemokine aggravated stromal keratitis due to reduced primary neutrophil influx in mice. 64 CXCL2 contributed to neutrophil recruitment and activation of their effector functions, including phagocytosis and production of reactive oxygen species in an immune complex-mediated cutaneous inflammation model. 65, 66 CXCR2-mediated (CXCL2 receptor) signaling is essential to margination of PMNs in S. aureus keratitis. 67 CXCL2 and CXCL10 also coordinate several types of cellular functions, including wound healing, angiogenesis, leucocyte recruitment, and internalization of foreign substances. [68] [69] [70] [71] Most studies on CXCL2 and CXCL10 have been focused on inflammatory cell recruitment, with very few focusing on bacterial internalization. To understand whether the absence of these chemokines altered neutrophil function, we compared the phagocytic capabilities of mouse neutrophils derived from wild type, CXCL2 −/− , and CXCL10 −/− mice. The absence of CXCL2 or CXCL10 did not alter Bacillus internalization, suggesting that the absence of either of these chemokines did not affect this function in neutrophils.

The vitreous humor contains a variety of organic and inorganic components that support bacterial growth, with no inherent antibacterial effects. 72 We reported that B. cereus grows faster than S. epidermidis, S. aureus, E. faecalis, and S. pneumoniae in the mouse eye, and reaches maximum concentrations within 12 hours postinfection. 12 During Bacillus endophthalmitis, the absence of innate receptors such as TLR2, TLR4, their adaptor MyD88, or immune mediators IL-6 or CXCL1 did not affect intraocular Bacillus growth. 23, 25, 27, 31 We reported that the absence of TNFα resulted in greater intraocular bacterial Bacillus growth, whereas the absence of the TLR4 adaptor TRIF resulted in significantly reduced intraocular Bacillus growth. 23, 33 Here, we found that in the absence of CXCL2 or CXCL10, Bacillus growth was not affected. Given the similar finding in CXCL1 −/− mouse eyes infected with B. cereus 31 it appears that these particular chemoattractants do not affect the control of Bacillus growth in the eye.

The retina is a multilayered tissue consisting of photosensitive cells and is responsible for phototransduction. Damage to this tissue is considered irreversible and can disrupt vision. 73 We and others have demonstrated that bacterial infection in the eye interferes with retinal structure and integrity, and negatively affects retinal function. 28, 49, 74 During Bacillus endophthalmitis, retinal function declines rapidly compared to endophthalmitis caused by other Grampositive ocular pathogens. 12 The absence of innate receptors TLR2 and TLR4, adaptor molecules MyD88 and TRIF, and chemokine CXCL1 prevented retinal damage and preserved retinal function. 23, 25, 27, 31 In the present study, compared to infected C57BL/6J eyes, CXCL2 −/− -infected eyes retained significant retinal function, even at 16 hours postinfection. However, CXCL10 −/− -infected eyes only retained retinal function up to 8 and 10 hours postinfection. Starting at 12 hours, retinal function in CXCL10 −/− infected eyes declined to approximately 25%. In contrast, retinal function in CXCL2 −/− eyes declined to 40% to 50% but remained elevated throughout 14 hours. We reported that an absence of CXCL1 preserved retinal structure and integrity, which corresponded with retinal function retention. 31 In the current study, retinal architecture was preserved and retinal layers were relatively intact in infected CXCL2 −/− eyes at 10, 12, and 14 hours postinfection. Ocular structures were also well preserved in infected CXCL10 −/− eyes at earlier time points. However, significant damage was observed after 12 hours postinfection, which correlated with the decline in retinal function in CXCL10 −/− eyes after 12 hours postinfection. These findings suggested that the absence of CXCL2 resulted in well preserved retinal structure and integrity for a prolonged period of time after infection, compared to that of the absence of CXCL10. The longer retention of retinal function and better ocular pathology observed in CXCL2 −/− mice could be because of the highly elevated CXCL2 expression in ocular environment during Bacillus endophthalmitis.

Chemokines play a crucial role in fighting bacterial infections by recruiting neutrophils to infected tissues. In mice, the chemokines KC (CXCL1) and MIP-2α (CXCL2) fulfill this role. 70, 75 Neutrophils function as effector cells that kill bacteria and destroy affected tissues mainly through the production of reactive oxygen species. 46, 48, 76 Cytokines TNFα, IL-6, and CXCL1 were detected as early as 4 hours postinfection during Bacillus endophthalmitis, which corre-lated with early neutrophil influx and fibrin accumulation in the anterior chamber. 28, 30, 33 High levels of myeloperoxidase, a microbicidal enzyme produced by PMNs, 77 have also been detected in the eye during Bacillus endophthalmitis. 25 Blocking the TLR2 and TLR4 pathways resulted in significantly reduced myeloperoxidase in infected eyes, which is a marker for reduced inflammation due to limited infiltration of PMN. 42 Inhibition of TLR2 and TLR4 activation significantly reduced the expression of chemoattractants that recruit PMNs to the infection site, which ultimately resulted in increased retinal function retention and improved ocular pathology. These studies suggest that the ocular damage and reduced retinal function observed in our current study are partly due to the uncontrolled infiltration of PMNs into the , and GAT + anti-CXCL10 (P = 0.0031) treated eyes compared to untreated eyes. (C) Compared to untreated eyes, B-wave function was significantly retained in anti-CXCL10 alone (P = 0.0031), GAT alone (P = 0.0016), and GAT + anti-CXCL10 (P = 0.0016) treated eyes. No differences in A-or B-wave retention was detected between untreated and isotype-treated eyes. (D) MPO concentrations in infected eyes of mice treated with anti-CXCL10 antibody with or without antibiotics were significantly less than in untreated eyes (*P ≤ 0.01). (E) Pathology analysis showed that eyes of untreated and isotype-treated mice were similar. Eyes in these groups were inflamed, with the loss of complete retinal architecture. In contrast, there was mild inflammation and undamaged retinal layers in infected eyes treated with anti-CXCL10 alone, GAT alone, and GAT + anti-CXCL10. Original magnification, × 10. Values represent means ± SEM for n ≥ 5 eyes per group per time point with at least 2 independent experiments. *P ≤ 0.05, **P ≤ 0.01, and ns P ≥ 0.05. eyes. Here, we observed that at all times PMN influx was significantly reduced in the absence of CXCL2. In infected CXCL10 −/− eyes, PMN influx was reduced at only early times, but was similar to that of infected C57BL/6J eyes at later times. These findings correspond to our retinal function and histology data. These findings suggest that, compared to CXCL10, CXCL2 may play a much broader role in the pathogenesis of Bacillus endophthalmitis.

Sex differences related to immune response and inflammation play a role in the susceptibility and pathogenesis of a variety of infections and diseases, [78] [79] [80] and are therefore an important biological variable. 81, 82 A recent report demonstrated that aging, not sex and genetic diversity, impacted the pathology of experimental S. aureus endophthalmitis. 83 In our study, we separately analyzed the data of male and female C57BL/6J, CXCL2 −/− , and CXCL10 −/− male and female mice to identify any potential sex differences in our model (see Figs. 2-4) . In general, we did not find consistent differences in Bacillus growth or retinal function loss between males and females in these groups, demonstrating that sex was not a significant biological variable for Bacillus endophthalmitis.

At present, treatment options for Bacillus endophthalmitis involve topical, systemic, and intravitreal antibiotics, which can effectively sterilize the infected eye if applied early during infection. 84, 85 Because severe inflammation is a hallmark of sight-threatening endophthalmitis, corticosteroid use for anti-inflammatory therapy is common, but their effectiveness in this disease is debatable. 39, 40, 54, 86, 87 Therefore, identifying viable anti-inflammatory targets is crucial in protecting the infected eye during Bacillus endophthalmitis. Proinflammatory mediator synthesis was associated with neutrophil infiltration and subsequent retinal function decline, which suggested the significance of these mediators as targets in controlling inflammation and overall pathogenesis of Bacillus endophthalmitis. 25, 28, 31, 33 Furthermore, targets that drive acute inflammatory responses in Bacillus endophthalmitis may be similar in slower-developing forms of bacterial endophthalmitis. During Bacillus endophthalmitis, in addition to CXCL1, we identified a cohort of inflammatory mediators whose expression in the retina is dependent upon a functional TLR4. 30 However, the role of these cytokines and chemokines in the pathogenesis of Bacillus endophthalmitis and as potential therapeutic targets has not been elucidated.

The proinflammatory chemokine CXCL2 possesses several physiological functions, including cell migration, adhesion, and innate immune responses. 75 A receptor antagonist of CXCL2 inhibited glioma growth during tumor initiation and progression, suggesting CXCL2 as a promising therapeutic target in glioma. 88 CXCL2 contributes to PMN infiltration in P. aeruginosa-infected corneas and administration of CXCL2 neutralizing antibody reduced PMN infiltration and corneal destruction during P. aeruginosa keratitis. 89 Here, we investigated the therapeutic potential of anti-CXCL2 antibody with or without routinely used antibiotics. Because intraocular Bacillus grow very rapidly and PMNs are recruited into the eye as early as 4 hours postinfection, we chose 2 hours postinfection as our time point for treatment. The concentration of GAT and antibody we chose was based on our previously reported findings. 90 Early administration of neutralizing antibodies against CXCL2 during experimental B. cereus endophthalmitis resulted in significantly improved retinal function, improved disease pathology, and reduced inflammation at 10 hours postinfection. Because intravitreal antibiotics are critical for eliminating bacteria, we also tested the therapeutic potential of CXCL2 antibody with or without GAT. The combination of GAT and anti-CXCL2 antibody treatment significantly reduced the inflammation compared to anti-CXCL2 antibody or GAT alone.

CXCL10 binds the CXCR3 receptor to induce apoptosis, cell growth, chemotaxis, and angiostasis. 91 CXCL10 is also recognized as a biomarker that predicts severity of various diseases. 92 Changes in CXCL10 levels have been associated with infectious diseases, immune dysfunction, and tumor development. 68, 91, 93 Administration of anti-CXCL10 antibody served as a promising therapeutic for experimental autoimmune encephalomyelitis (EAE) by decreasing the clinical and histological manifestation in mice. 94 CXCL10 neutralization, either by specific antibody or by genetic deletion, protected mice from cerebral malaria infection and inflammation. 95 In another study, targeted blocking of CXCL10 or CXCR3 receptor with antibodies attenuated inflammatory colitis in mice. 96 Here, administration of anti-CXCL10 antibody to Bacillus-infected eyes reduced inflammatory cell influx, resulting in retained retinal function and reduced disease pathology compared to that of untreated infected mice. However, there was no difference in inflammation or retinal function when anti-CXCL10 antibody with or without GAT or when GAT alone was administered. Although we observed better clinical outcomes in our study using a single antibiotic (gatifloxacin) and single antibodies alone at this early stage of infection, additional studies with different dosages of antibody and different antibiotics are necessary to better conclude the therapeutic potential of neutralizing the important chemokines analyzed in this model. Bacillus endophthalmitis is characterized by uncontrolled and robust acute immune responses that rapidly recruit inflammatory cells and damage host intraocular tissues. Identification of targets that reduce inflammatory cell infiltration into the eye could serve as the basis of novel therapeutic modalities. In this study, we showed that the absence of CXCL2 or CXCL10 by genetic deletion resulted in better retention of retinal function, better ocular pathology, and less inflammation compared to C57BL/6J eyes (Fig. 9) . In comparing the absence of CXCL2 with that of CXCL10, the absence of CXCL2 maintained a greater degree of retinal function and less pathology and inflammation. Furthermore, treatment aimed at neutralizing CXCL2 and CXCL10 reduced the inflammation and pathology. Adding gatifloxacin sterilized infected eyes, but synergistic effects with the antibodies were not clearly defined. Overall, our study presents a unique therapeutic approach of targeting chemokines in ocular environment as an anti-inflammation treatment modality that can be tested in conjunction with antibiotics for treating Bacillus endophthalmitis. Because the course of endophthalmitis caused by other Gram-positive pathogens is similar but delayed compared with that of Bacillus endophthalmitis, this therapeutic strategy could have broader applicability.

Postoperative Endophthalmitis: A Review of Risk Factors, Prophylaxis, Incidence, Microbiology, Treatment, and Outcomes

Post traumatic endophthalmitis: incidence and risk factors

Bacterial and Fungal Endophthalmitis

The pathogenesis of bacterial endophthalmitis

Bacillus cereus: Epidemiology, Virulence Factors, and Host-Pathogen Interactions

The Bacillus cereus Group

The Bacillus cereus Food Infection as Multifactorial Process

Endophthalmitis caused by Bacillus species

The cereus matter of Bacillus endophthalmitis

Endophthalmitis caused by Bacillus cereus: a devastating ophthalmological emergency

Role of bacterial and host factors in infectious endophthalmitis

Contribution of membrane-damaging toxins to Bacillus endophthalmitis pathogenesis

Role of hemolysin BL in the pathogenesis of extraintestinal Bacillus cereus infection assessed in an endophthalmitis model

Relationship of plcR-regulated factors to Bacillus endophthalmitis virulence

Bacillus endophthalmitis: roles of bacterial toxins and motility during infection

Role of swarming migration in the pathogenesis of Bacillus endophthalmitis

The role of pili in Bacillus cereus intraocular infection

S-layer Impacts the Virulence of Bacillus in Endophthalmitis

Innate Immune Interference Attenuates Inflammation in Bacillus Endophthalmitis

Unexpected Roles for Toll-Like Receptor 4 and TRIF in Intraocular Infection with Gram-Positive Bacteria

Role of TLR5 and flagella in Bacillus intraocular infection

Role of Toll-like receptor (TLR) 2 in experimental Bacillus cereus endophthalmitis

Toll-like receptor family and signalling pathway

Role of MyD88-dependent and MyD88-independent Pathways in Bacillus cereus Endophthalmitis

Acute inflammation and loss of retinal architecture and function during experimental Bacillus endophthalmitis

Targets of immunomodulation in bacterial endophthalmitis

TLR4 modulates inflammatory gene targets in the retina during Bacillus cereus endophthalmitis

CXCL1, but not IL-6, significantly impacts intraocular inflammation during infection

Bacillus cereus induces permeability of an in vitro blood-retina barrier

A role for tumor necrosis factor-alpha in experimental Bacillus cereus endophthalmitis pathogenesis

Structure, function, and inhibition of chemokines

Chemokine: receptor structure, interactions, and antagonism

Innate immune pattern recognition: a cell biological perspective

Toll-like Receptors and the Control of Immunity

Clinical profile and outcome in Bacillus endophthalmitis

Corticosteroid and antibiotic therapy for Bacillus endophthalmitis

Effects of intravitreal corticosteroids in the treatment of Bacillus cereus endophthalmitis

Efficacy of vitrectomy in improving the outcome of Bacillus cereus endophthalmitis

Bacillus S-Layer-Mediated Innate Interactions During Endophthalmitis

Bacillus cereus-induced permeability of the bloodocular barrier during experimental endophthalmitis

Disarming Pore-Forming Toxins with Biomimetic Nanosponges in Intraocular

A Novel Biomimetic Nanosponge Protects the Retina from the Enterococcus faecalis Cytolysin

The role of neutrophils in the immune system: an overview

The Bacillus virulome in endophthalmitis. Microbiology (Reading)

A Pyrrhic Victory: The PMN Response to Ocular Bacterial Infections. Microorganisms

Pathogenesis of gram-positive bacterial endophthalmitis

Expression of Bacillus cereus Virulence-Related Genes in an

Incidence of Endophthalmitis after Intravitreal Injections: Risk Factors, Microbiology Profile, and Clinical Outcomes

Incidence and characteristics of acute-onset postoperative bacterial and sterile endophthalmitis in dogs following elective phacoemulsification: 1,447 cases (1995-2015)

Treatment of Bacillus cereus endophthalmitis with endoscopy-assisted vitrectomy

Treatment of experimental Bacillus cereus endophthalmitis using intravitreal moxifloxacin with or without dexamethasone

Treatment and outcome of traumatic endophthalmitis in open globe injury with retained intraocular foreign body

Results in the management of Bacillus endophthalmitis

Diversity of Bacillus cereus sensu lato mobilome

Staphylococcus aureus endophthalmitis: antibiotic susceptibilities, methicillin resistance, and clinical outcomes

Endophthalmitis caused by Streptococcus pneumoniae

The crucial roles of inflammatory mediators in inflammation: A review

Chemokines and leukocyte traffic

TLR2 mediates the innate response of retinal Muller glia to Staphylococcus aureus

Structural basis of CXC chemokine receptor 2 activation and signalling

Absence of CXCL10 aggravates herpes stromal keratitis with reduced primary neutrophil influx in mice

Neutrophils Self-Regulate Immune Complex-Mediated Cutaneous Inflammation through CXCL2

Neutrophil chemoattractant receptors in health and disease: double-edged swords

The role of CXC chemokine receptor 2 in Staphylococcus aureus keratitis

Chemokines and Their Receptors Are Key Players in the Orchestra That Regulates Wound Healing

Chemokine Receptors and Phagocyte

Neutrophil recruitment by chemokines Cxcl1/KC and Cxcl2/MIP2: Role of Cxcr2 activation and glycosaminoglycan interactions

CXCL2/MIF-CXCR2 signaling promotes the recruitment of myeloid-derived suppressor cells and is correlated with prognosis in bladder cancer

Adult vitreous structure and postnatal changes

Functional architecture of the retina: development and disease

Role of Staphylococcus aureus Virulence Factors in Inducing Inflammation and Vascular Permeability in a Mouse Model of Bacterial Endophthalmitis

The chemokine system in innate immunity

Killing of Listeria monocytogenes by inflammatory neutrophils and mononuclear phagocytes from immune and nonimmune mice

Its role for host defense, inflammation, and neutrophil function

Sex differences in immune responses

Considering how biological sex impacts immune responses and COVID-19 outcomes

Characterization of Sex Differences in Ocular Herpes Simplex Virus 1 Infection and Herpes Stromal Keratitis Pathogenesis of Wild-Type and Herpesvirus Entry Mediator Knockout Mice. mSphere

Sex as an important biological variable in biomedical research

Considering sex as a biological variable in preclinical research

But Not Sex and Genetic Diversity, Impacts the Pathobiology of Bacterial Endophthalmitis

Bacterial endophthalmitis: epidemiology, therapeutics, and bacterium-host interactions

Bacterial endophthalmitis: therapeutic challenges and host-pathogen interactions

Failure of intravitreal dexamethasone to diminish inflammation or retinal toxicity in an experimental model of Bacillus cereus endophthalmitis

Toward improving therapeutic regimens for Bacillus endophthalmitis

The CXCR2/CXCL2 signalling pathway -An alternative therapeutic approach in high-grade glioma

Macrophage inflammatory protein-2 is a mediator of polymorphonuclear neutrophil influx in ocular bacterial infection

Biomimetic nanosponges augment gatifloxacin in reducing retinal damage during experimental MRSA endophthalmitis

CXCL10/IP-10 in infectious diseases pathogenesis and potential therapeutic implications

Plasma chemokines are biomarkers of disease severity, higher bacterial burden and delayed sputum culture conversion in pulmonary tuberculosis

CXCL9, CXCL10, CXCL11, and their receptor (CXCR3) in neuroinflammation and neurodegeneration

Neutralization of IFN-inducible protein 10/CXCL10 exacerbates experimental autoimmune encephalomyelitis

Pharmacologic inhibition of CXCL10 in combination with anti-malarial therapy eliminates mortality associated with murine model of cerebral malaria

CXCR3 axis: role in inflammatory bowel disease and its therapeutic implication

The authors thank Feng Li