Anaplasma phagocytophilum Infection Induces Protracted Neutrophil Degranulation

Abstract

Anaplasma phagocytophilum-infected neutrophil degranulation could exacerbate inflammation. Thus, the degranulation of infected neutrophils was assayed. Infected neutrophils expressed CD11b and CD66b, and supernatants of infected neutrophils showed more proMMP-9 and MMP-9 activity than controls and continued to do so for ≥18 h. Degranulation-related inflammatory tissue injury may account for some clinical manifestations in human granulocytic anaplasmosis.

Human granulocytic anaplasmosis (HGA) is a tick-borne disease caused by Anaplasma phagocytophilum, an obligate intracellular bacterium of neutrophils. Clinical manifestations range from inapparent to fatal disease, with occasional opportunistic infections (2, 14). A. phagocytophilum alters neutrophils by activating some functions (expression of adhesion molecules and chemokine secretion) (1, 4) and by deactivating others (reactive oxygen species, endothelial cell adhesion, and transmigration) (3-5, 12). Activated neutrophils contribute to the control of infectious agents through respiratory burst and degranulation (15, 16). Although respiratory burst is defective in A. phagocytophilum-infected neutrophils (3, 5), few studies of degranulation and the role that it plays in the pathogenesis of HGA have been conducted. The present study demonstrates that A. phagocytophilum infection stimulates neutrophil degranulation for at least 18 h, suggesting that prolonged degranulation, in the absence of intact phagocyte oxidase regulation, may contribute to proinflammatory tissue injury with HGA.

A. phagocytophilum strain Webster was cultivated on HL-60 cells as described previously (4, 5). Cell-free bacteria were prepared from heavily infected cells by syringe needle lysis, and cellular debris was removed by centrifugation (4, 5). These bacteria were washed and used immediately to infect 5 × 10⁶ human peripheral blood neutrophils isolated from healthy donors. Neutrophil purity was determined to be ≥95% by Romanowsky staining (Hema-3; Biochemical Sciences, Inc., Swedesboro, N.J.), and the viability of cells was determined to be >98% by trypan blue dye exclusion. Human neutrophils were obtained with the approval of The Johns Hopkins University School of Medicine Institutional Review Board and in compliance with all relevant federal guidelines and institutional policies.

Purified neutrophils were incubated for 3 h at 37°C alone, with 10 μg of lipopolysaccharide (LPS) from Escherichia coli 0111:B4 (Sigma, St. Louis, Mo.)/ml, or with cell-free A. phagocytophilum, washed, and then fixed for flow cytometry. Double staining was performed at 4°C (i) with fluorescein isothiocyanate-conjugated anti-human CD66b (BD Pharmingen, San Diego, Calif.), a glycosylphosphatidyl inositol-anchored membrane protein of specific granules that appears on cell membranes after degranulation, and (ii) with phycoerythrin-conjugated anti-human CD11b (BD Pharmingen), a β₂ integrin of specific and gelatinase-containing neutrophil granules also released to the cell membrane with degranulation (7). Differences between degranulation elicited by LPS-stimulated and A. phagocytophilum-infected neutrophils and that elicited by neutrophils in medium only, as reflected by expression of CD66b and CD11b, were analyzed by using two-sided, unequal-variance Student's t tests; P values of <0.01 were considered significant.

Neutrophil gelatinase granules also liberate matrix metalloproteases (MMPs) that can degrade basement membranes and interstitium during neutrophil extravasation and in disease (6, 10, 11). To demonstrate the degranulation of MMPs, purified neutrophils were incubated at 37°C for 3 and 18 h without any stimulation, with cell-free A. phagocytophilum, or with 0.1 μM N-formyl-methionyl-leucyl-phenylalanine (fMLP; Sigma), a proinflammatory stimulant for neutrophil activation and degranulation. Cells and supernatants were examined for proMMP-9 by immunoblotting or for MMP-9 activity by gelatin zymography as previously described (9). Cell lysates and supernatants were separated by sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis with a gel containing 0.1% gelatin (Sigma) and incubated in zymogram renaturing buffer and then in developing buffer overnight. Protease activity was identified as clear bands in the gel after Coomassie blue staining. To detect proMMP-9, proteins were transferred to nitrocellulose membranes, incubated with human MMP-9 monoclonal antibody (Gelatinase B; BD Pharmingen), reacted with alkaline phosphatase-labeled secondary antibody (KPL Inc., Gaithersburg, Md.), and visualized with 5-bromo-4-chloro-3-indolylphosphate (BCIP)-nitroblue tetrazolium.

After 3 and 18 h, increased expression of CD11b on A. phagocytophilum-infected neutrophils compared with that on either LPS- or mock-stimulated neutrophils was demonstrated (P < 0.001) (Fig. 1A). Similarly, significantly more CD66b was detected on the surfaces of A. phagocytophilum-infected neutrophils than on the surfaces of either LPS-stimulated or control neutrophils after 3 h (P < 0.001) (Fig. 1B). Increased CD11b expression in dimethyl sulfoxide-differentiated HL-60 cells as long as 3 days after infection was previously demonstrated (4).

FIG. 1.

A. phagocytophilum induces degranulation, as indicated by surface expression of CD11b and CD66b in neutrophils assessed by flow cytometry. (A) Purified neutrophils were incubated without or with LPS and A. phagocytophilum for 3 h. The expression of CD11b and CD66b was significantly increased on A. phagocytophilum-infected neutrophils compared with that on either LPS-stimulated or unstimulated neutrophils (P < 0.001). The bar graph shows the mean increases in fluorescence for A. phagocytophilum-infected and LPS-stimulated neutrophils relative to that for uninfected neutrophils in three replicated experiments in which neutrophils were assessed for degranulation and surface expression of CD11b at 3 h and of CD66b at 3 and 18 h (error bars represent standard errors of the means). (B) After 18 h of stimulation or infection, CD66b expression was significantly increased (P < 0.001) on A. phagocytophilum-infected neutrophils compared with that on LPS-stimulated neutrophils, LPS-stimulated A. phagocytophilum-infected neutrophils, and unstimulated neutrophils (shaded curves). CD11b expression was not analyzed at 18 h. The results presented are representative of three independent experiments.

A loss of active MMP-9 (83 kDa) and immunoreactive proMMP-9 (92 kDa) from neutrophils in the A. phagocytophilum-infected and fMLP-stimulated neutrophils after 3 h compared with levels in unstimulated neutrophils was demonstrated (Fig. 2A and B). Likewise, increases in MMP-9 activity and proMMP-9 immunoreactivity compared with those in unstimulated neutrophils was demonstrated in culture supernatants (Fig. 2A and B), corroborating degranulation and the activation of the proenzyme with infection.

FIG. 2.

A. phagocytophilum-infected neutrophils degranulate and release MMP-9 for at least 18 h in vitro. (A) Gelatin zymographic demonstration of changes in proteolytically active MMP-9 (83 kDa) in neutrophils and neutrophil culture supernatants. Neutrophils stimulated to degranulate with fMLP or infected with A. phagocytophilum (Ap) showed a decrease in active MMP-9 compared with the level for unstimulated neutrophils (N), whereas supernatants from these cultures showed an increase in active MMP-9 compared with the level for unstimulated neutrophils. (B) Immunoreactive (immunoblot) proMMP-9 was decreased in cell pellets and increased in neutrophil culture supernatants compared with that in uninfected controls after stimulation of degranulation with fMLP or after infection with A. phagocytophilum. +C, purified human MMP-9; N, unstimulated neutrophils; M, tissue culture medium only. Molecular masses (mw) are shown in kilodaltons.

Degranulation by neutrophils is a major mechanism by which these host defense cells limit pathogen replication and consequent tissue injury. The functional diversity of azurophilic and specific granule substances allows neutrophils to target bacteria in many different ways (8). However, poorly controlled degranulation is also the basis for some inflammatory diseases (13). A. phagocytophilum-infected neutrophils have a reduced ability to bind endothelial cells resulting from infection-induced degranulation of a sheddase that cleaves the adhesion molecule platelet selectin glycoprotein ligand 1 (4). These data suggest that A. phagocytophilum infection of neutrophils induces generalized persistent degranulation, including MMP-9 release. Bacteria-induced inflammatory injury from degranulation is usually contained through rapid neutrophil apoptosis (13). However, A. phagocytophilum extends neutrophil life by inhibition of apoptosis, prolonging exposure to the neutrophil degradative components that are likely to elicit at least focal tissue injury (17).

Despite the activated-deactivated phenotype of A. phagocytophilum-infected neutrophils, one paradoxical effect of defective phagocyte oxidase is increased inflammation, in part because of an inability to metabolize proinflammatory and chemoattractant molecules (13). A. phagocytophilum-infected neutrophils may also promote inflammation via expression of chemokines that recruit other inflammatory cells and exacerbate tissue injury (3, 4). Yet, infected cells are unable to emigrate or generate components for intracellular killing and contribute little toward infection control or regulation of inflammation (3, 4, 13).

How such derangements profit the bacterium is uncertain. Akkoyunlu et al. demonstrated increased bacterial burden with induced chemokine secretion (1). Sheddases from infected neutrophils cleave platelet selectin glycoprotein ligand 1 and diminish adhesion to endothelial cells (4). This process assures that most infected cells are released into the circulating blood pool for ready access to tick bites. It is possible that degranulation and inflammatory injury are the by-products of a system that evolved to prevent neutrophil adhesion. Since the absolute numbers of infected neutrophils in both animals and humans are relatively small, the net result is usually mild disease (2). However, infection could trigger widespread injury similar to that observed with toxic or septic shock or, if localized, organ damage such as that associated with respiratory distress, hepatitis, focal neuritis, and even opportunistic infections, among other possibilities (2). Further study will be required to demonstrate whether excess degranulation and neutrophil functional changes contribute to HGA pathogenesis and whether these changes also promote A. phagocytophilum transmission to ticks.