Regulation of the inflammatory response: enhancing neutrophil infiltration under chronic inflammatory conditions

Abstract

Neutrophil (PMN) infiltration plays a central role in inflammation and is a major cause of tissue damage. Thus, PMN infiltration must be tightly controlled. Using zymosan-induced peritonitis as an in vivo PMN infiltration model, we show in this study that PMN response and infiltration were significantly enhanced in mice experiencing various systemic inflammation including colitis and diabetes. Adoptive transfer of leukocytes from mice with inflammation into healthy recipients or from healthy into inflammatory recipients followed by inducing peritonitis demonstrated that both circulating PMN and tissue macrophages were altered under inflammation and that they collectively contributed to enhanced PMN infiltration. Detailed analyses of dextran sulfate sodium (DSS)-elicited colitis revealed that enhancement of PMN infiltration and macrophage function occurred only at the post-acute/chronic phase of inflammation and was associated with markedly increased IL-17A in serum. In vitro and ex vivo treatment of isolated PMN and macrophages confirmed that IL-17A directly modulates these cells and significantly enhances their inflammatory responses. Neutralization of IL-17A eliminated the enhancement of PMN infiltration and IL-6 production and also prevented severe tissue damage in DSS-treated mice. Thus, IL-17A produced at the chronic stage of colitis serves as an essential feedback signal that enhances PMN infiltration and promotes inflammation.

Introduction

PMN (polymorphonuclear leukocytes) infiltration into tissues during inflammation plays a central role in innate immunity. Upon pathogen infection or irritant infliction, local macrophages and other cells sense the insult and produce a panel of inflammatory mediators such as cytokines and chemokines that stimulate the nearby microvasculature and attract large numbers of PMN to migrate across the vascular wall and infiltrate into tissues. After arrival at the inflammatory site, PMN perform phagocytosis and also release powerful anti-pathogen and tissue-damaging reagents to kill pathogens and aberrant cells. While these forceful PMN activities are extremely important for host defense, their adverse effects are also apparent, as demonstrated in inflammatory bowel diseases (IBD), arthritis, some cardiovascular conditions, inflammatory pulmonary and renal diseases, and viral/bacterial infection-associated damage. In these cases, the powerful but relatively low target-specificity of PMN result in grave tissue/organ injuries, often leading to a rapid progression of the pathophysiological condition. Thus, effective control of PMN infiltration would much alleviate the inflammatory condition.

The mechanisms that control PMN response and regulate their infiltration involve multiple cell types and various inflammatory factors, and studies in this area are still inadequate. Some aspects of PMN function, such as chemotaxis and degranulation, have been studied in vitro by refined assays (1–3). However, these assays do not reveal the entire design of PMN inflammatory response and infiltration during inflammation and how these processes are dynamically regulated in vivo. In the study of PMN function in diabetes, we observed that PMN isolated from diabetic patients showed impaired transmigration by the in vitro chemotaxis assay. These observations, which are consistent with other reports (4–6), indicated an important fact that PMN from diabetic patients were different from those from healthy donors; however, they could not ascertain whether the activities of PMN of diabetes were truly abated or, on the contrary, were primed (activated to an extent). In the second possibility, the primed PMN were hypersensitive and may have become activated during the in vitro isolating procedure resulting in their inefficient chemotaxis in the later transmigration assays (3). We have also observed similar impaired PMN transmigration in vitro when testing PMN isolated from donors under other inflammatory conditions.

To further study PMN response and infiltration, we designed a two-layered in vivo inflammatory system in which a relatively chronic inflammatory condition, such as colitis, type I diabetes or other inflammatory condition, was first created in the experimental animals. After the condition was stabilized, a typical acute inflammatory response, such as zymosan-induced peritonitis, was induced and the immediate PMN infiltration into the newly established inflammatory site was assayed within a few hours. Using this design, we found that active inflammation at the post-acute/chronic phase in particular induces systemic alterations in both PMN and tissue microenvironments including macrophages. Instead of suppressing inflammation, these changes result in an enhancement of PMN infiltration, hence exaggerate the inflammatory reaction.

Materials and Methods

Mice and inflammatory mouse models

C57BL/6 mice of 6–10 weeks old (18–22 g) purchased from Jackson Laboratory (Bar Harbor, ME) were housed with free access to water and food in a specific pathogen-free facility. All experiments using animals and procedures of animal care and handling were carried out following protocols approved by the Institutional Animal Care and Use Committee (IACUC) of Georgia State University. To establish DSS-induced colitis (7–8), 2% DSS (MP Biomedicals) prepared in pure water was given to the mice as the drinking water. Mice were inspected daily for distress and colitis symptoms such as wet stool, diarrhea, bloody diarrhea and weight loss. To induce hyperglycemia/type-1 diabetes (9–10), the mice were fasted for 4 h and then injected with streptozotocin (STZ, Sigma) at a dose of 50 mg/kg body weight. STZ was freshly prepared in sodium citrate buffer (pH 4.5) prior to the injections, which were performed daily for five consecutive days. The mice were then maintained for an additional two weeks to allow the diabetic condition to be stabilized. Serum glucose levels were tested with the One-Touch glucose monitoring system, and in general, >90% of injected mice developed hyperglycemia. To induce bronchitis/lung inflammation, the mice were given LPS [E. coli serotype 026:B6 (Sigma)], 0.3 mg/kg body weight, in 70 μl PBS through the nares every other day for 15 days.

Zymosan-induced peritonitis

This model was used to assay the acute PMN response and infiltration in vivo (11). Briefly, mice were injected intraperitoneally with 0.5 mg zymosan A (Sigma) that was prepared in 0.5 ml PBS. Mice were euthanized at varying time points after the injection. PMN that transmigrated into the peritoneum were collected by peritoneal lavage with 3 ml cold Hank's buffered salt solution (HBSS) without Ca²⁺ and Mg²⁺ followed by analysis with myeloperoxidase (MPO) assay, cell counting and immunofluorescence labeling using a PE-Cy5-conjugated anti-mouse Ly-6G (Gr-1) antibody (eBioscience). Macrophages were analyzed using an Alexa Fluor488-conjugated anti-mouse F4/80 antibody (clone BM8, Biolegend). For leukocyte adoptive transfer experiments, bone marrow cells were obtained by flushing femur bone cavities followed by labeling with the fluorescent cell tracer carboxyfluorescein diacetate succinimidyl ester (CFSE) or 1,1′ - Dihexadecyl - 3,3,3′,3′ - tetramethylindocarbocyanine iodide (DiIC16(3)) (both from Invitrogen). In some experiments, the bone marrow cells were further separated by a Percoll (GE Healthcare) density gradient to obtain the population of mature PMN (12). In general, 5×10⁶ labeled cells (about 2.5 ×10⁶ were Gr-1⁺) were transfused into a recipient mouse via intravenous injection before inducing peritonitis using zymosan. To determine the role of IL-17A in PMN infiltration, ex vivo experiments were performed. In these experiments, freshly isolated bone marrow cells were labeled with CFSE followed by treatment with a recombinant mouse IL-17A (2ng/ml) (BioLegend), with or without the presence of a neutralization anti-IL-17A antibody (Low Endotoxin, Azide-Free (LEAF) purified anti-mouse IL-17A antibody (BioLegend)), or 50 μl serum collected from non-treated mice or mice treated with DSS for 12 days. After incubation for 2 h (37°C), the cells (about 2.5×10⁶ PMN) were combined with equal amount of DiIC₁₆(3)-labeled bone marrow PMN before transfer into healthy recipient mice. Zymosan-induced peritonitis was then performed to test the transferred PMN infiltration into the peritoneum.

Measurement of cytokines

Whole blood from mice was collected without anti-coagulation additives, and the serum was isolated. For ELISA, mouse serum of different dilutions was incubated in anti-cytokine antibody-coated wells in 96-well plates for 2 h (25°C). After washing, the wells were incubated with biotin-conjugated detecting antibodies and HRP-conjugated streptavidin followed by detection with o-Phenylenediamine dihydrochloride (Sigma) substrate. Both capture and detecting antibodies against different cytokines including IL-1β, IL-4, IL-6, IL-12 p40/70, IL-17A, IFNγ, TNFα, MCP-1, IL-10 and GM-CSF were purchased from BD Biosciences and Biolegend. Purified recombinant murine cytokines used as the standards were purchased from PeproTech. To detect cytokines produced by in vitro treated macrophages, peritoneal macrophages were obtained by lavage the peritoneal space and collection by centrifugation. The macrophages were then stimulated with zymosan, or recombinant IL-17A (2ng/ml) with or without the presence of the anti-IL-17A antibody (2mg/ml), for various time periods up to 48 h in DMEM with 8% FBS (37°C). After the treatment, the supernatants were collected and levels of IL-6 and other cytokines were detected by ELISA.

In vitro PMN chemotaxis assay

An in-vitro transmigration setup using transwell devices with collagen-coated transfilters (0.33cm², 5μm pore size) established previously (1–3) was used to assess PMN chemotactic capability. Briefly, bone marrow leukocytes were isolated from femur bones of non-treated, healthy mice and mice with DSS-induced colitis or STZ-induced diabetes by flushing the bone cavities using Hank's buffer devoid of Ca²⁺ and Mg²⁺. After lysis of RBC, percentages of PMN in cell populations were determined by MPO assay and immunofluorescence staining using PE-Cy5-conjugated anti-mouse Ly-6G (Gr-1) antibody. Aliquots of cell samples from each type of mice were also subject to further separation to obtain the population of mature PMN according to a protocol described by Dong and Wu (12). After normalization of the amount mature PMN, the cells (containing 2.5×10⁶ PMN) were added into the upper chamber of the chemotaxis setup in 150μl HBSS followed by inducing transmigration with 10μM chemoattractant formyl-methionyl-leucyl-phenylalanine (fMLF) adding into the lower chamber in 500μl HBSS. The setup was then incubated at 37°C and the PMN transmigrated across the filter into the lower chamber was then quantified at different time points by MPO assay. To test the effect of IL-17A on PMN, the freshly isolated cells loaded in the upper transmigration chambers were treated with the cytokine with or without neutralization antibody, or vehicle control buffer, for 2 h before inducing transmigration by fMLF.

Staining of tissue sections

For histochemical staining, mouse tissues including colon and lung were fixed in 4% paraformaldehyde, embedded in paraffin, and cut into 5–10 μm thin sections. Sections were stained with hematoxylin and eosin (H&E). For staining of infiltrated leukocytes and PMN, freshly sectioned tissues were blocked with 5% BSA followed by staining using a PE-Cy5-conjugated anti-mouse Ly-6G (Gr-1) antibody. Images were then taken using an Olympus BX21 fluorescence microscope.

Cell labeling and Western blot

To directly observe IL-17A binding to PMN and macrophages, recombinant murine IL-17A was labeled with FITC using an EZ-Label FITC protein labeling kit (PIERCE). The FITC-labeled IL-17A was then incubated with freshly isolated mouse bone marrow leukocytes and peritoneal macrophages for 30 min (25°C) followed by washing and analysis by fluorescence microscopy. To detect IL-17A-induced protein phosphorylation changes in macrophages, isolated murine peritoneal macrophages were treated with the recombinant IL-17A at concentrations of 0.5 and 2.0 ng/ml, or vehicle, for 48 h (37°C). The cells were then lysed in a buffer containing 100mM Tris (pH 7.5), 150mM NaCl, 1% Triton X-100, 1:100 dilution of a protease inhibitor cocktail (Sigma, Product Code: P8340) and 3 mM PMSF followed by SDS-PAGE and Western blot detection using the anti-phosphotyrosine antibody 4G10 (Millipore), an anti-phosphoserine antibody (Cell Signaling) and an anti-actin antibody (Sigma).

Results

Analysis of the acute PMN response in vivo by zymosan-induced peritonitis

In this study, we utilized zymosan-induced peritonitis (11), a commonly used self-resolving acute inflammation model, to assay the PMN immediate response and infiltration in vivo. For these experiments, healthy C57BL/6 mice that were maintained in a pathogen-free animal facility were injected intraperitoneally with zymosan A, followed by analysis for PMN recruitment into the peritoneum after various time periods. As shown in Table 1 and Figure 1 A, before zymosan injection, approximately 2.6×10⁶ cells were present in the peritoneum, and among these cells, ~30% were F4/80⁺ macrophages, and the remainder contained few (~5%) Gr-1⁺, which presumably were tissue granulocytes, and other tissue cells. Soon after zymosan challenge, PMN began to infiltrate into the peritoneum, resulting in drastic changes in the cell composition and numbers. As shown, at 2 h, small numbers of PMN (~1×10⁶) transmigrated across endothelium and arrived at the peritoneum. Meanwhile, majority of macrophages were seemingly “leaving” the area resulting in a significant decrease of the F4/80⁺ cells (Table 1). Massive PMN infiltration into the peritoneum occurred after 2 h, and nearly 10⁷ PMN and over 1.2×10⁷ PMN were recruited to the location at 4 h at 6 h, respectively. By 8 h, PMN infiltration slowed rapidly or stopped, resulting in a decrease in the number of PMN in the area. After the peak of PMN recruitment, monocyte infiltration began to increase at 6 h and continued to increase thereafter, suggesting that the acute inflammation proceeded to a resolving stage. The infiltrating monocytes and their-mediated clearance likely accounted for the decrease of PMN numbers in the peritoneum at later time points. Consistent with other reports (13–14), we detected a robust but transient increase in IL-6 in the peritoneal fluid and in the blood serum (Fig. 1B and Table 2) after zymosan challenge. As observed, the time course of IL-6 change tightly correlated with the dynamics of PMN infiltration, suggesting a potential role of this cytokine in controlling the episode of zymosan-induced peritonitis. Moderate increases in IL-1, IL-17A, and TNFα were also detected. The dynamics of IL-1 resembled that of IL-6, but IL-17A was only increased at later time periods, when the massive PMN transmigration was nearly completed (Fig. 1B and Table2).

Table 1.

Leukocytes in the peritoneum and their dynamic changes during zymosan-induced peritonitis. Mice injected with zymosan A were euthanized at various time points followed by collection cells from peritonea and staining for macrophages and PMN/granulocytes using an anti-F4/80 antibody (green) and an anti-Gr-1(red) antibody, respectively. Control cell staining used normal rat serum. The cells of different staining were counted and analyzed by FACS. The 0 h time point was determined without zymosan injection. The table shows the mean cell number ± SEM, n=6 mice/group.

Time	Cell Number (mean ±SEM)
	F4/80+(×105)	Gr-r(×105)	Total(×105)
0h	10.4±1.4	2.0±0.7	26.4±3.7
2h	1.5±0.2	10.7±1.5	15.3±2.2
4h	0.9±0.1	96.2±6.8	120.5±8.5
6h	1.5±0.2	124.3±10.4	146.2±12.2
8h	8.2±1.1	115.1±6.6	154.3±8.8

Figure 1.

Inflammatory response / peritonitis after zymosan challenging. Mice receiving injection (i.p.) of zymosan A were euthanized at the indicated time-points. A) Macrophages and PMN lavaged from the peritoneum were stained with anti-F4/80 (green) and anti-Gr-1(red) antibodies, respectively, and visualized under a fluorescent microscope. B) Serum concentrations of inflammatory cytokines during peritonitis were evaluated by ELISA. Data are expressed as means ± SEM; n=6 mice/group. **, p<0.01, versus respective serum cytokine levels at 0 h.

Table 2.

Time-course detection of serum cytokines during zymosan A-induced peritonitis. Concentrations of inflammatory cytokines in serum during peritonitis were evaluated by ELISA. Data are expressed as means ±SEM; n=6 mice/group.

Cytokine (pg/ml)	Time post zymosan injection
	0h	2h	4h	6h	8h
TNFα	49.2±10.3	118.1±20.6*	111.8±16.0*	96.4±12.8*	100.7±16.5*
IL-1β	31.0±6.6	159.3±26.5*	498.0±58.9**	113.7±32.5*	64.3±22.7
IL-6	65.6±22.8	2180.0±387.9**	4713.1±1333.4**	1023.7±322.4**	555.0±165.6**
IL-17A	109.6±10.8	126.9±12.7	370.3±94.5**	562.0±77.4**	517.3±64.6**

^*p<0.05

^**p<0.01, versus respective serum cytokine levels at 0 h.

Enhancing PMN infiltration under chronic inflammatory conditions

To investigate whether chronic inflammation affects PMN infiltration, we induced various inflammatory conditions in C57BL/6 mice before performing the zymosan-peritonitis assay. In these experiments, mice were treated with either 2% DSS to induce colitis (7–8), STZ to induce type 1 diabetes (9–10), or LPS to induce bronchitis/lung inflammation (15). A group of mice were also treated with one dose of emulsified complete Freund's adjuvant (CFA, 50% in PBS) to induce chronic inflammation. As shown in Figure 2A, progressive body weight loss, which was associated with diarrhea and colitis, was observed in DSS-treated mice. Immunofluorescence (see Fig. 4D) and histochemical staining of colon tissues after treating the mice with DSS demonstrated significant intestinal inflammation that was associated with aggressive leukocyte infiltration, severe mucosal damage and loss of epithelial structures. Treating mice with STZ and LPS also successfully induced diabetic and lung inflammatory conditions. As shown in Figure 2B, all of the STZ-treated mice developed hyperglycemia with the mean blood glucose level being increased to 431.7 ± 56.1 mg/dL, whereas the control group treated with vehicle (PBS) maintained a blood glucose level of 141.2 ± 14.0 mg/dL. Meanwhile, treating mice with low dose LPS via the nares for 15 days induced evident upper respiratory inflammation that was characterized by noticeable leukocyte infiltration around bronchi, bronchioles and in some area of the lung parenchyma (Fig. 2C).

Figure 2.

Mice with various chronic inflammatory conditions display enhanced PMN infiltration during zymosan-induced peritonitis. A) DSS-induced colitis. C57BL/6J mice were given 2% DSS in drinking water continuously and development of colitis was monitored by body weight loss, diarrhea, etc, and confirmed by dissection of the colonic tissues. B) STZ-induced type 1diabetes. Mice were consecutively injected with STZ or the vehicle (CTL) for five days. Serum glucose levels were assayed three weeks after the injections. C) Inducing upper respiratory inflammation by LPS. LPS or vehicle was given to mice via the nares every other day for 15 days. The figure shows the histology images (H&E staining) of lung tissue sections after vehicle (Ctl.) or LPS treatment. D–E) Enhanced PMN infiltration in mice with systemic inflammation. Control healthy mice and mice with colitis (DSS for 12days), or hyperglycemia (3 weeks post-STZ treatment) or lung inflammation (LPS) were tested for PMN infiltration by zymosan A-induced peritonitis. Mice received one dose of emulsified CFA (50% in PBS) were also tested in the assay. Panels D and E show the numbers of PMN infiltrated into the peritoneum at 2 h and the kinetics of PMN infiltration into the peritoneum, respectively. The data represent one of eight independent experiments and are expressed as means ± SEM; n=mice number/group. ***, p<0.001, **, p<0.01, versus respective controls.

Figure 4.

Enhancement of PMN infiltration in DSS-induced colitis is correlated with IL-17A production. A) Time-course analysis of serum cytokines during DSS-induced colitis (also see Table 3). B) Analysis of PMN infiltration by zymosan-induced peritonitis in mice treated with DSS for different days. Note that significantly enhanced PMN infiltration occurs after 6 days of DSS treatment and in parallel with increases in IL-17A. C) Analysis of peritonitis-associated IL-6 production in mice treated with DSS for different days. D) Tissue section of colon intestines and labeling for infiltrated PMN using an anti-Gr-1 antibody. The data represent one of at least three independent experiments, and are expresses as means ± SEM with at least five mice per group. *, p<0.05, **, p<0.01, and ***, p<0.001, versus respective controls.

We then examined PMN function in these mice with various inflammatory conditions by performing zymosan-induced peritonitis. For these experiments, we utilized the mice that had been treated with DSS for 12 days and manifested apparent colitic condition. As shown in Figure 2D, compared to the control healthy mice that PMN just began to infiltrate into the peritoneum at 2 h and only small numbers (~1×10⁶ PMN) were detected, significant increased numbers of PMN (> 6×10⁶) were recruited to the area in colitic mice at the same time point. Similar early promoted PMN infiltration into the peritoneum in response to zymosan challenge was also observed in STZ-induced diabetic mice and in LPS-treated mice. These mice were used to test for PMN infiltration by peritonitis after the STZ- and LPS-induced pathological conditions were stabilized. As shown in Figures 2D, at 2 h, 5.6×10⁶ PMN and approximately 3×10⁶ PMN were recruited to peritonea in STZ- and LPS-treated mice, respectively. We also induced peritonitis in the mice that were treated with CFA a month ago. As shown in Figure 2D, this form of chronic inflammation also significantly enhanced PMN response to zymosan, resulting in over 7×10⁶ PMN being recruited to the peritoneum by 2 h.

Kinetic analysis of PMN infiltration (Fig. 2E) in colitic mice by plotting the total numbers of PMN transmigrated as a function of time showed that the robust PMN infiltration occurred earlier than that in the control mice. Indeed, over 70% of the infiltrated PMN in mice with DSS treatment for 12 days had arrived at the peritoneum by 2 h, whereas in the controls, the majority of PMN were recruited by 4 h. Significant enhanced PMN infiltration in to the peritoneum at early time point (2 h) was also observed in mice with DSS treatment for 9 days (Fig. 2E). The kinetic results also indicated that the early enhancement of PMN infiltration in colitic mice was not due to inflammation-induced increases in PMN production but accelerated PMN response and transmigration. In fact, the accumulative PMN numbers that infiltrated into the peritoneum in colitic mice at 4 h were not higher, but tended to be lower, than those in the controls. Analyses of PMN numbers in the peripheral circulation and in bone marrows showed that despite the increase of GM-CSF in the serum (Table 3), continual DSS treatment (≥ 9 days) and colitis exhausted the entire system and the PMN reservoir, and also consumed PMN availability in the peripheral circulation.

Table 3.

Serum cytokine levels during DSS induced colitis. Inflammatory cytokine levels in serum during DSS-induced colitis were evaluated by ELISA. Data are expressed as means ±SEM; n=6–8 mice/group. ND, undetectable.

Cytokine (pg/ml)	DSS treatment
	Day 0	Day 3	Day 6	Day 9	Day 12
TNFα	52.8±5.6	71.0±5.8*	69.7±8.1	57.8±11.6	52.1±9.3
IL-1β	35.4±8.6	28.5±6.0	39.4±11.2	227.5±22.7**	144.6±16.4**
IL-6	70.2±17.6	78.7±10.4	132.2±42.4***	430.8±109.1**	173.4±24.0**
IL-17A	113.6±10.6	108.9±10.9	249.8±24.1	714.1±93.6** 2	2023.9±172.1**
MCP-1	30.3±8.6	37.2±9.4	156.8±16.2**	79.4±18.2*	22.5±4.3
IFN-γ	51.0±6.7	44.2±7.8	51.1±9.3	40.4±18.0	41.2±6.4
IL-12/23(p40)	55.6±7.3	57.8±6.5	77.4±11.9*	67.0±6.1	54.6±7.2
IL-4	39.5±4.2	43.6±7.3	ND	ND	ND
IL-10	59.2±20.9	133.6±16.4*	35.5±12.8	30.4±11.3	27.2±13.6
GM-CSF	ND	65.0±16.3*	29.4±9.2	21.0±8.4	ND

^*p<0.05

^**p<0.01

^***p<0.001, versus respective controls.

Inflammation-induced modulations of both circulating PMN and tissue microenvironments contribute to the promotion of PMN infiltration

The results of enhanced PMN infiltration in mice under various inflammatory conditions suggest a likelihood that alterations occurred in either PMN, which directly promote PMN transmigration, or in cells within the peritoneal vicinity, which then led to an earlier local response to zymosan and/or contributed to the facilitation when PMN transmigrated through. To test these possibilities, we designed two directional, adoptive leukocyte transfer and peritonitis experiments. In these experiments, donor PMN derived from mice with different inflammatory conditions were transfused into healthy, non-inflammatory recipient mice and were tested for response and infiltration in zymosan-induced peritonitis; conversely, donor PMN from healthy mice were transfused into recipient mice with ongoing inflammation and were then tested for infiltration in response to zymosan challenging. The former experiments were performed to determine whether chronic inflammation modulated PMN and converted them to be “faster” migrating cells even in a new host environment. The latter experiments were designed to examine the peritoneal tissue microenvironment and to determine whether chronic inflammation could induce changes that promoted PMN infiltration during peritonitis. Control experiments were performed by transfusion of PMN from healthy mice into other healthy mice followed by testing the donor PMN infiltration.

Figure 3A shows the results of the transfusion of PMN from inflammatory mice into healthy recipients and the infiltration of donor PMN when these mice were challenged with zymosan. As shown, although the host mice had no prior inflammation, donor PMN from colitic and diabetic mice (labeled with CFSE, green) still displayed much faster infiltration, resulting in more cells transmigrated into the peritoneum at 2 h than control donor PMN from other healthy mice. In parallel, we also performed in vitro chemotaxis assays and compared the transmigration of PMN derived from DSS-treated mice and from healthy mice. To avoid activation of PMN, especially those from mice with inflammation, we directly applied the obtained total bone marrow leukocytes without further isolation. As shown in Figure 3B, greater numbers of PMN transmigrating across matrix-coated filters towards the chemoattractant fMLF were observed in samples derived from DSS-treated mice than those from healthy mice. We further performed adoptive transfer experiments and compared these PMN for their infiltration in vivo. In these experiments, donor PMN or total bone marrow cells derived from DSS-treated mice and from healthy mice were labeled with different dyes (green and red), and simultaneously transfused into the same recipient mice (at 1:1 ratio of the PMN number). As shown in Figure 3C, zymosan induced more infiltration of PMN from colitic mice (green) than those from non-colitic, healthy mice (red) at 2 h, confirming that PMN in chronic inflammatory mice had an enhanced ability to respond and to migrate toward inflammatory stimuli.

Figure 3.

Systemic inflammation alters both PMN and tissue microenvironments. A) Testing donor PMN from inflammatory mice response in healthy recipients by zymosan-peritonitis. Bone marrow cells (2.5×10⁶ mature PMN) derived from colitic (DSS, 12 d), diabetic (STZ) or control healthy mice were labeled with CFSE and transfused into healthy recipients, respectively. Figure shows the microscopic analysis of infiltrated donor PMN in peritoneal lavage after inducing peritonitis for 2 h and the total numbers of donor PMN infiltrated at 2 h and 4 h. The result represents one of four independent experiments. B) PMN chemotaxis in vitro. Bone marrow leukocytes from healthy control and colitic (DSS, 12 d) mice were normalized for equal PMN contents by MPO assays before being applied into the upper chambers of the transmigration setups. Chemotaxis induced by fMLF was measured at 45 min. C) Co-transfer of leukocytes from healthy and inflammatory mice and testing by peritonitis. A mixture of bone marrow leukocytes derived from healthy mice (red, DiIC₁₆ labeling) and colitic mice (DSS, 12 d) (green, CFSE labeling) (2.5×10⁶ PMN/each) were transferred into the same healthy recipients. The images show zymosan-induced PMN infiltration into the peritoneum at 2 h in the peritoneal lavage. The result represents one of five independent experiments. D) Transfer of PMN from healthy mice into inflammatory recipients followed by testing PMN infiltration during peritonitis. Bone marrow leukocytes from five healthy mice were combined and labeled, followed by transfusion into different types of inflammatory recipient mice (3 mice per type, 2.5×10⁶ PMN/mouse). E) Peritoneal macrophages obtained by lavaging peritonea were stimulated in vitro using zymosan (1 mg/ml) in DMEM (containing 5% FBS) at 37°C. IL-6 secreted into the medium at different time points was assayed by ELISA. The data are expressed as means ± SEM. *, p<0.05, **, p<0.01, versus respective controls.

To investigate whether the tissue microenvironment also contributed to the enhanced PMN infiltration under inflammatory conditions, we transferred the PMN from healthy mice into DSS and STZ –treated, chronic inflammatory mice (recipients). Inducing peritonitis subsequently showed significantly faster PMN infiltration in mice with ongoing colitis or diabetes than in control, healthy recipients (Fig. 3D). These results suggest that considerable alterations had occurred in the peritoneal tissue environment in inflammatory mice and that these changes promoted PMN infiltration. Because peritoneal macrophages play an important role in the development of peritonitis, we collected these macrophages and directly tested their responses to zymosan in vitro. As shown in Figure 3E, in the cell culture dishes zymosan induced increased amounts of IL-6 production in macrophages from DSS- and STZ-treated mice when compared with macrophages isolated from non-inflammatory controls. Interestingly, we observed significantly increased numbers of macrophages in the peritoneum of STZ-induced diabetic mice (data not shown). Thus, these results confirmed tissue environment changes under chronic inflammation. Increases in cytokine production from the local macrophages certainly contributed to enhanced PMN infiltration during peritonitis.

Role of IL-17 in colitis-associated inflammatory modulation

The findings that circulating PMN and peritoneal macrophages had altered responsiveness under colitis and diabetes were striking given that both cells were distal to the original inflammatory sites in those conditions. We thus hypothesized that critical inflammatory factors were released into the circulation, through which it reached and regulated cells at a distance. To identify these important factors especially those that might contribute to enhance PMN infiltration, we carefully analyzed inflammatory cytokines associated with DSS-induced colitis in the serum. As shown in Table 3, while most of cytokines tested were reportedly detected by RT-PCR or by other means in colonic tissues (16–20), these cytokines were maintained at low or moderate levels in the serum throughout the course of the DSS treatment. The only cytokine that we found significantly elevated in the serum was IL-17A (Fig. 4A), which started to increase at day 6–8 of DSS treatment and continued increasing through the rest course of the experiment, eventually reaching >2ng/ml. We found that not only was its elevation in the serum extraordinary but also that the dynamics of IL-17A tightly correlated with the enhancement of PMN infiltration observed in peritonitis. As shown in Figure 4B, inducing peritonitis in mice that were treated with DSS for different days revealed that significant promotion of PMN infiltration occurred only after DSS treatment for 6 days and continued to be enhanced thereafter. Supporting this observation, assay of peritonitis-associated IL-6 release also demonstrated that enhancement of IL-6 release upon zymosan challenging occurred in mice treated with DSS for more than 6 days (Fig. 4C). Interestingly, significant body weight decreases in DSS-treated mice also started at approximately the same time (Fig. 2A and Fig. 5C). These results suggest that, although DSS-induced colitis started much earlier than day 6, as elevations of inflammatory cytokines and leukocyte infiltration into the colonic tissues were apparent on day 2–5 (Table 3, Fig. 4D), the quick collapse of the pathological condition at the later stage was likely associated with the enhancement of PMN infiltration (Fig. 4D) and the consequent severe tissue damage.

Figure 5.

Suppression of IL-17A function by a neutralization antibody eliminates the augmentation of inflammation and PMN infiltration at post-acute stage of DSS-induced colitis. Mice under DSS treatment were administrated (i.v.) with 30 μg LEAF purified anti-mouse IL-17A antibody on day 4, 6, 8, and 10. Control injections were performed by administration of an isotype-matched antibody (rat IgG1) (BioLegend). On day 12, mice with and without antibody administration were tested by zymosan-induced peritonitis. A) PMN infiltration into the peritoneum at 2 h and 4 h. B) Serum IL-6 after zymosan challenge. C) The body weight of mice treated with free water or DSS plus antibodies. The data are expresses as means ± SEM and represent one of at least three independent experiments with at least five mice per group. **, p<0.01, ***, p<0.001, versus respective controls.

To further explore the role of IL-17A in the enhancement of PMN infiltration, we treated the DSS mice with a functional neutralization antibody against IL-17A (21–22) (i.v.). As shown in Figures 5, suppression of IL-17A function by the antibody in DSS-treated mice nearly completely eliminated the enhanced PMN infiltration at 2 h during zymosan-induced peritonitis (Fig. 5A) and such event-associated increase in IL-6 (Fig. 5B). The colitic symptoms in these DSS-treated mice, such as body weight loss, diarrhea and intestinal tissue damage, were also largely alleviated (Fig. 5C). However, IL-17A neutralization had no effect on the basal level of PMN infiltration as large numbers of PMN were still recruited into the peritoneum at 4 h during peritonitis even with the antibody administration (Fig. 5A). PMN infiltration into intestinal mucosa was also present in DSS-treated mice with anti-IL-17A administration, but remained at a low level (data not shown). We further performed ex vivo experiments and tested whether exposure to IL-17A directly modulates PMN function. In these experiments, bone marrow leukocytes isolated from healthy mice were ex vivo treated with a functional recombinant IL-17A and then transferred into other healthy mice (recipients) followed by testing by zymosan-induced peritonitis. As shown in Figure 6A, IL-17A treatment markedly accelerated PMN infiltration into the peritoneum during peritonitis compared to non-treated, control PMN that were co-transferred into the same recipients. As expected, the presence of the anti-IL-17A antibody in the treatment completely eliminated the enhancement of PMN infiltration. Similarly, PMN that were treated with the serum of colitic mice (DSS treatment for 12 days) also manifested accelerated infiltration in recipient mice during peritonitis. Consistent with these results, assaying PMN chemotaxis in vitro (Fig. 6B) observed significantly augmented chemotactic transmigration for PMN treated with IL-17A, while neutralization of IL-17A function erased such augmentation. Since tissue macrophage responses were also altered under inflammatory conditions, we wondered whether IL-17A also directly modulates macrophage function. As shown in Figure 6C, IL-17A treatment of peritoneal macrophages isolated from healthy mice markedly elevated IL-6 production upon the subsequent stimulation with zymosan. Western blot analyses of IL-17A-treated macrophages indicated significantly increased protein tyrosine phosphorylations (Fig. 6D), suggesting that active signaling events were elicited by IL-17A. In addition, we also observed direct binding of recombinant IL-17A to the cell surfaces of PMN and macrophages (not shown), presumably to the receptors. In conclusion, these results strongly suggest that IL-17A, which was produced at a large amount in the post-acute/chronic stage of colitis, serves as a critical feedback factor and contributes to reinforcement of the inflammatory reaction through systemically modulating leukocytes in the circulation and tissues.

Figure 6.

IL-17A directly enhances PMN and macrophage inflammatory functions. A) PMN ex vivo exposure to IL-17A exhibit an accelerated infiltration during zymosan-induced peritonitis. Bone marrow leukocytes from healthy mice were treated for 2 h with a recombinant IL-17A (2ng/ml) or serum from mice with DSS-induced colitis (DSS, 12d) in the presence or absence of the neutralization anti-IL-17A antibody. After treatment, the cells were labeled with CSFE (green) and co-transferred with control, non-treated bone marrow leukocytes (red) into healthy recipient mice followed by testing for PMN infiltration during zymosan-induced peritonitis. B) IL-17A directly enhances PMN chemotactic transmigration. Bone marrow PMN with and without exposure to IL-17A were assayed in vitro for chemotaxis towards fMLF for 90 min. C) IL-17A promotes IL-6 production in macrophage upon stimulation. Peritoneal macrophages obtained from healthy mice were treated with recombinant IL-17A (2ng/ml) with or without the presence of anti-IL-17A antibody for 48 h (37°C). IL-6 production in cells with or without zymosan stimulation (2 h) was assayed by ELISA. D) Western blot analyses of protein phosphorylation of macrophages treated with IL-17A.

Discussion

PMN, which comprise more than two-thirds of peripheral leukocytes, are the first tier responsive cells during inflammation. As shown in zymosan-induced peritonitis, when a stimulatory signal is effectively produced, PMN rapidly transmigrate through the vasculature and tissues and arrive at the inflammatory site within a few hours. After a short period and vigorous activity, PMN infiltration is then promptly ceased, provided that the inflammatory causes are largely cleared, in order to confine and to limit tissue damage. However, when the inflammatory causes linger, as that in DSS-induced colitis, PMN infiltration continues and the inflammatory reaction prolongs and becomes chronic. In this study, we employed multiple mouse models of inflammation and studied PMN infiltration under various chronic conditions. We found that animals that were pre-conditioned with non-resolving colitis, type-1 diabetes or other inflammatory conditions had significantly enhanced PMN infiltration during zymosan-induced peritonitis. Cross-transfer of leukocytes from mice with ongoing inflammation to healthy recipients, or from healthy mice to inflammatory recipients, further revealed that both circulating PMN and tissue microenvironments were altered under chronic inflammatory conditions and that these changes collectively contributed to the promotion of PMN infiltration during peritonitis.

Detailed analysis of DSS-induced colitis found that PMN infiltration into intestinal mucosa occurred shortly after the starting of DSS treatment (day 2–3); concomitantly, a panel of inflammatory cytokines associated with colitis, including low levels of IL-17A, which are presumably released from the local γδ T cells and infiltrated PMN (23), were also produced (Table 3). However, when the DSS treatment continued over 6–8 days, we observed extraordinary high levels of IL-17A in the blood serum. Our studies suggest that this late produced strong wave of IL-17A, which was likely through the activation of the second tier inflammatory cells such as Th17 cells, is capable of reaching and regulating cells distant from the original inflammatory site through circulation, leading to systemic modulations of the entire inflammation and homeostasis. Indeed, our data showed that both macrophages and PMN in distant tissues are regulated by such released, circulating IL-17A. In fact, PMN appeared to be altered even before their release from the bone marrow, consistent with the microarray study by Theilgaard-Monch et al. (24–25) in which various cytokine receptors including IL-17R were identified in bone marrow myelocytes especially mature PMN. Strikingly, the time-course of elevation of IL-17A in serum during DSS treatment tightly correlated with the significant enhancement of PMN infiltration observed in zymosan-induced peritonitis. Subsequent in vitro and ex vivo experiments confirmed that IL-17A directly modulates PMN and significantly accelerated PMN chemotactic transmigration towards inflammatory stimuli. Not only modulating PMN, IL-17A also enhances the responses and sensitivities of macrophages around the body. In vitro exposure of macrophages to IL-17A prior to stimulation with zymosan resulted in much increased production of the pro-inflammatory cytokine IL-6. Peritoneal macrophages isolated from the late stage colitic mice also demonstrated enhanced responses upon inflammatory challenge. All these results together indicate that when inflammation prolongs, feedback signals are produced and trigger a systemic regulation of the entire inflammatory reaction. In the case of DSS-induced colitis, large amount of IL-17A produced at the post-acute stage primes circulating PMN and tissue macrophages and enhances their inflammatory functions. These regulations by IL-17A, which are presumably designed to accelerate the removal of the original inflammatory cause, nevertheless lead to exaggerated inflammation- associated tissue damage. Suppression of IL-17A–mediated propagation of inflammation mitigates the disease condition.

Priming PMN under inflammatory conditions and by inflammatory factors has been observed through years, and has been considered to be a major factor responsible for the pathogenesis of organ failure occurred under post-injury conditions such as trauma, burn and surgery, and other conditions such as adult respiratory distress syndrome, sepsis, cystic fibrosis, blood transfusion, etc. (26–33) Agents that prime PMN have been reported for platelet-activating factor (PAF) (34–35), granulocyte-colony stimulating factor (G-CSF) (36–37), granulocyte-monocyte colony-stimulating factor (GM-CSF) (38), IL-1β (39), IL-6 (40), IL-8 (36), etc. In this study, our data suggest that IL-17A is likely to be a critical priming agent for PMN and macrophages under inflammatory conditions, yet the mechanism underlying its effect and the phenotypic characteristics of PMN (41) after exposure to IL-17A need to be further determined. Although many studies have shown that IL-17A production is regulated by IL-6 and IL-23 (and other cytokines) (42–43), our study in DSS-induced colitis only detected slight increases of these cytokines in the serum, implicating that IL-6 and IL-23 –mediated activation of Th17 cells may be mainly restricted in the local intestinal tissues. In addition, previous studies by Schwarzenberger et al (44–45) and others (46–47) have shown that increases of IL-17A promote promyelocyte proliferation and PMN maturation. Since IL-17A was significant increased in our experiments of DSS-induced colitis, it is likely that such condition was associated with increased granulopoiesis and PMN production and these increases are necessary for the ongoing inflammation and continual PMN infiltration into intestines. However, increases of PMN production under colitis were not the reason for the enhanced PMN infiltration observed during peritonitis, which instead was due to PMN priming and accelerated recruitment. Indeed, we detected only minimal PMN in the peripheral circulation and no increase of mature PMN in bone marrows in mice treated with DSS for 9 days or longer, suggesting that incessant colitis has consumed majority of PMN and largely depleted the reservoirs.

Thus, through this study, we have learnt important mechanisms that regulate the dynamics of inflammation. Our results imply that when an inflammatory condition becomes unrelenting, such as in inflammatory bowel diseases (IBD), diabetes, and rheumatoid arthritis, significantly enhanced inflammatory responses and PMN infiltration can occur and lead to severe tissue injury and deteriorated pathophysiological condition. In addition, systemically enhanced inflammatory status also renders the chronic patients highly vulnerable when they face a new inflammatory challenge, such as viral / bacterial infection or cancer, during which exaggerated reactions can prevail and cause overstated consequences.