Fish and mammalian phagocytes differentially regulate pro-inflammatory and homeostatic responses in vivo

Abstract

Phagocytosis is a cellular mechanism that is important to the early induction of antimicrobial responses and the regulation of adaptive immunity. At an inflammatory site, phagocytes serve as central regulators for both pro-inflammatory and homeostatic anti-inflammatory processes. However, it remains unclear if this is a recent evolutionary development or whether the capacity to balance between these two seemingly contradictory processes is a feature already displayed in lower vertebrates. In this study, we used murine (C57BL/6) and teleost fish (C. auratus) in vitro and in vivo models to assess the evolutionary conservation of this dichotomy at a site of inflammation. At the level of the macrophage, we found that teleost fish already displayed divergent pro-inflammatory and homeostatic responses following internalization of zymosan or apoptotic bodies, respectively, and that these were consistent with those of mice. However, fish and mice displayed significant differences in vivo with regards to the level of responsiveness to zymosan and apoptotic bodies, the identity of infiltrating leukocytes, their rate of infiltration, and the kinetics and strength of resulting antimicrobial responses. Unlike macrophages, significant differences were identified between teleost and murine neutrophilic responses. We report for the first time that activated murine, but not teleost neutrophils, possess the capacity to internalize apoptotic bodies. This internalization translates into reduction of neutrophil ROS production. This may play an important part in the recently identified anti-inflammatory activity that mammalian neutrophils display during the resolution phase of inflammation. Our observations are consistent with continued honing of inflammatory control mechanisms from fish to mammals, and provide added insights into the evolutionary path that has resulted in the integrated, multilayered responses that are characteristic of higher vertebrates.

Figure 1. Apoptotic bodies significantly reduce inflammatory respiratory burst responses of goldfish primary macrophages.

(A) Goldfish PKM were separately incubated with apoptotic bodies or zymosan (5∶1, particle: cell ratio) for 2 h. Cells were then harvested and respiratory burst was assayed using DHR. n = 3. (B) Goldfish PKM were cultured in the presence of both apoptotic bodies and zymosan for 2 h (5∶1 ratio for both). Respiratory burst was then analyzed in cells based on phagocytic capacity as follows: non-phagocytic cells, cells containing only apoptotic bodies, cells containing only zymosan or cells containing both apoptotic bodies and zymosan (white bars). Representative images are shown for cells in each of these groups. To investigate the effects of pre-incubation with apoptotic bodies, apoptotic bodies were added to PKM simultaneously to zymosan (‘−0 h’), 2 h before zymosan (‘−2 h’) or 4 h before zymosan (‘−4 h’). n = 4. For all, * p<0.05; ** p<0.01; For (c), + p<0.05 and ++ p<0.01 compared to (‘−0 h’). No- no internalized particle; AB- apoptotic body; Zy- zymosan.

Figure 2. In vivo administration of zymosan induces a marked infiltration of leukocytes that is linked to high levels of respiratory burst.

Goldfish (left) and C57BL/6 mice (right) were injected intraperitoneally with 2.5 mg of zymosan. Cells were harvested by peritoneal lavage at 0 h (saline alone), 8, 24 and 48 h and counted (top row). Injection of zymosan resulted in a marked increase in cell numbers isolated from the peritoneum that peaked at 48 h for mice and 24 h for goldfish. Respiratory burst in isolated cells at these time points was determined with DHR (bottom row). n = 4; * p<0.05; ** p<0.01.

Figure 3. Pro-inflammatory (zymosan) and homeostatic (apoptotic bodies) stimuli differentially impact leukocyte infiltration profiles in goldfish and mice.

Goldfish and C57BL/6 mice were injected intraperitoneally with saline, apoptotic bodies (5×10⁶) or zymosan (2.5 mg). Apoptotic bodies were also pre-injected 4 h before zymosan injections. Goldfish leukocyte populations were defined by imaging flow cytometry (area, internal complexity, and morphology) and staining patterns with Sudan Black and an anti-CSF-1R antibody (Figure S3). Murine cells were defined based on surface markers for neutrophils (F4/80⁻/Gr1⁺/CD11b⁺), monocytes (F4/80^lo/Gr1^+/−/CD11b⁺), macrophages (F4/80^hi/Gr1^+/−/CD11b⁺) and lymphocytes (F4/80⁻/Gr1⁻; CD3, B220, NK1.1; Figure S4A). n = 4; * p<0.05 and ** p<0.01 compared to control; + p<0.05 and ++ p<0.01 compared to zymosan. No- no internalized particle; AB- apoptotic body; Zy- zymosan.

Figure 4. In vivo administration of apoptotic bodies leads to a more dramatic reduction of pro-inflammatory respiratory burst responses in teleost fish compared to mice.

Goldfish and C57BL/6 mice were injected intraperitoneally with saline, apoptotic bodies (5×10⁶) or zymosan (2.5 mg) and incubated for 24 h. Apoptotic bodies were also pre-injected 0, 2, or 4 h before zymosan injections to assess the contributions of kinetics to these responses. Cells from injected animals were harvested by peritoneal lavage and respiratory burst was assayed with DHR. n = 4; * p<0.05 and ** p<0.01 compared to PBS (saline) control; + p<0.05 and ++ p<0.01 compared to zymosan. No- no internalized particle; A.B.− apoptotic body; Zy- zymosan.

Figure 5. Goldfish myeloid cell respiratory burst responses are most affected by the presence of apoptotic bodies.

(A) Goldfish were injected intraperitoneally with saline, apoptotic bodies (5×10⁶) or zymosan (2.5 mg). Apoptotic bodies were also pre-injected 0, 2, or 4 h before zymosan injections. Cells from injected goldfish were harvested by peritoneal lavage and respiratory burst was assayed with DHR in peritoneal cell subpopulations based on forward scatter and side scatter profiles. n = 4; * p<0.05 and ** p<0.01 compared to control; + p<0.05 and ++ p<0.01 compared to zymosan. No- no internalized particle; A.B.- apoptotic body; Zy- zymosan. (B) Representative histograms show a single peak in DHR responses.

Figure 6. Murine neutrophil respiratory burst antimicrobial responses are most greatly affected by the presence of apoptotic bodies.

(A) C57BL/6 mice were injected intraperitoneally with saline, apoptotic bodies (5×10⁶) or zymosan (2.5 mg). Apoptotic bodies were also pre-injected 0, 2, or 4 h before zymosan injections. Cells from injected mice were harvested by peritoneal lavage and respiratory burst was assayed with DHR in peritoneal cell subpopulations based on forward scatter and side scatter profiles (Figure S4B). n = 4; * p<0.05 and ** p<0.01 compared to control; + p<0.05 and ++ p<0.01 compared to zymosan. No- no internalized particle; A.B.- apoptotic body; Zy- zymosan. (B) Histograms show representative DHR responses for total leukocytes, neutrophils and monocytes/macrophages. We found that the high responders were predominantly neutrophils (>90%). Pre-incubation with apoptotic bodies resulted in a preferential switch in the neutrophil population from high responders to mid/low responders.

Figure 7. Apoptotic bodies downregulate murine neutrophil ROS production in a contact dependent manner.

Goldfish (left) and C57BL/6 mice (right) were injected intraperitoneally with zymosan (2.5 mg). Activated peritoneal cells from were harvested by peritoneal lavage and subpopulations were isolated by density centrifugation. (A) Separated neutrophil or mononuclear populations were incubated with labeled apoptotic bodies for 2 h and internalization was analyzed. n = 4; * p<0.05 and ** p<0.01; Mφ/M  =  macrophage/monocyte. (B) Isolated populations were cultured for 2 h in the presence of the indicated stimuli. Conditions denoted within brackets were contained within a 4 μm transwell. After 2 h, responder cells outside the transwells were harvested and respiratory burst was assayed using DHR.