Impact of Mycophenolic Acid on the Functionality of Human Polymorphonuclear Neutrophils and Dendritic Cells during Interaction with Aspergillus fumigatus

Abstract

Mycophenolic acid (MPA) is used clinically to prevent graft rejection but may increase the risk of fungal infection. We observed that MPA enhanced the Aspergillus fumigatus-induced oxidative burst of polymorphonuclear neutrophils, but without a corresponding increase in fungal killing. Furthermore, MPA inhibited the proinflammatory cytokine response and maturation of dendritic cells.

Mycophenolate mofetil is an antimetabolite immunosuppressant used in stem cell and solid organ transplant regimens for graft rejection and graft-versus-host disease prophylaxis and for treatment of acute or chronic graft-versus-host disease. Mycophenolic acid (MPA) is the active drug moiety and is a potent, selective, and reversible inhibitor of IMP dehydrogenase, leading to arrest of T- and B-lymphocyte proliferation (1, 2). The increasing application of immunosuppressive drugs is one reason for the growing incidence of invasive aspergillosis (6, 9). In this study, the impact of MPA on the functionality of polymorphonuclear neutrophils (PMN) and dendritic cells (DCs) in the immune response to Aspergillus fumigatus was analyzed.

To quantify reactive oxygen intermediates (ROI), PMN were separated from blood of healthy volunteers by dextran sedimentation (Nycomed, Norway) (15). ROI were quantified by measuring the conversion of dichlorofluorescein acetate (2.5 μM; Sigma-Aldrich, Germany) to green fluorescent dichlorofluorescein (15). PMN were stimulated at 37°C in a fluorescence reader (GENios; Tecan, Germany) recording in 5-min intervals (excitation wavelength, 485 nm; emission wavelength, 520 nm). Phorbol myristate acetate (PMA) (23 ng/ml; Sigma-Aldrich) was used as a positive control.

Killing assays with A. fumigatus germlings (ATCC 9197) (8 × 10⁴; multiplicity of infection [MOI] = 1) were performed as described previously (12). MPA (10 μM; Novartis, Germany) was administered for 3 h at 37°C. Serial dilutions were plated on Sabouraud agar and incubated for 24 h to quantify viable fungi. A two-tailed unpaired t test was used for statistical analyses.

To analyze the viability of PMN under MPA treatment, apoptosis and necrosis rates were determined by flow cytometry (FACSCalibur; Becton Dickinson), using a dual-color protocol quantifying phosphatidylserine by fluorescein isothiocyanate (FITC)-annexin V staining (Roche) and DNA of dead cells by propidium iodide staining (Sigma) (16).

For immature DC (iDC) generation, monocytes were isolated by magnet-associated cell sorting using anti-human CD14 antibodies (Miltenyi, Germany). Differentiation into iDCs (8) was achieved by cultivating cells for 7 days (4). Where indicated, MPA (10 μM) was added at either day 0, 2, or 7 of culturing. On day 7, cells were cocultivated with A. fumigatus germlings (1 × 10⁶; MOI = 1; 6 h), followed by RNA extraction (14). Fifteen micrograms of cRNA was incubated for 16 h at 45°C with an HG-U133 Plus 2.0 array (Affymetrix). Array data were analyzed using software from the Bioconductor project (www.bioconductor.org). Alternatively, LightCycler-based quantitative real-time reverse transcription-PCR assays were performed for selected genes (14).

Flow cytometry was performed using CD1a-FITC (Dako Cytomation), CD40-phycoerythrin (CD40-PE) (Immunotech), CD14-FITC, CD80-PE, CD83-PE, and CD86-PE (Becton Dickinson) antibodies.

Quantification of ROI generated by PMN during coincubation showed that A. fumigatus germlings triggered a strong oxidative burst. In accordance with the work of Hochegger et al. (11), we observed that 10 μM MPA had cumulative effects on ROI production by PMN (Fig. 1). PMN stimulated with germlings and MPA showed an increased oxidative burst (at 60 min, 3.8-fold [2.6- to 5.1-fold]; and at 120 min, 2.8-fold [2.0- to 4.2-fold]) compared to PMN stimulated with A. fumigatus germlings alone. MPA alone was able to induce the formation of ROI in human PMN to a small degree (Fig. 1). In contrast to MPA, cyclosporine, dexamethasone, cortisone, and everolimus did not induce ROI formation by PMN at concentrations of up to 10 μg/ml (data not shown).

FIG. 1.

(A) MPA triggers ROI generation in resting and activated human PMN. Resting PMN (medium) synthesize only background levels of ROI. In contrast, if MPA (10 μg/ml) is added (medium + MPA), significantly more ROI can be detected. After activation with PMA, PMN synthesize large amounts of ROI. This activation can be enhanced by the addition of MPA (10 μg/ml; PMA + MPA). Arithmetic means ± standard deviations for three independent experiments with PMN from different donors are shown. The upper detection limit of the fluorescence reader is 70,000 relative fluorescence units (RFU). (B) MPA enhances the oxidative burst triggered by A. fumigatus germlings. Af, A. fumigatus germlings at an MOI of 1; Af + MPA, A. fumigatus germlings at an MOI of 1 plus 10 μg/ml MPA. Arithmetic means ± standard deviations for three independent experiments with PMN from different donors are shown. (C) Killing of A. fumigatus germlings by human PMN (+PMN) is not affected by MPA (+PMN +MPA) but can be enhanced by PMA (+PMN +PMA). MPA does not affect the viability of A. fumigatus in the absence of PMN (MPA control). Bars represent arithmetic means plus standard variations for four independent experiments with PMN from different donors. Data sets for each experiment were normalized against the growth control.

To analyze if MPA-enhanced ROI production correlated with improved killing efficiency, plating assays were performed. Surprisingly, PMN killing of A. fumigatus was not affected by MPA, as surviving rates of fungi were not significantly changed (with MPA, 57% ± 12%; and without MPA, 53% ± 11%). To test whether costimulation of PMN during cultivation with A. fumigatus could result in improved killing, PMA, a potent activator of PMN, was tested in identical experimental settings. In contrast to the case for MPA, costimulation with PMA led to a significantly improved killing efficiency (fungal survival rate, 26% ± 6%; P < 0.05).

Apoptosis plays an important role in the interaction between PMN and microorganisms, and early cell death might preclude PMN from efficiently fighting pathogens. MPA has been suggested to induce apoptosis in activated T cells (3). To analyze whether MPA could induce apoptosis in PMN, flow cytometry analyses were performed. However, no difference in viability between PMN exposed to MPA and control PMN was detected.

Differentiation and maturation of DCs were reported to be influenced by MPA, leading to decreased expression of the surface markers CD80 and CD86 (5). We confirmed a down-regulation of CD80 and CD86 by MPA. Furthermore, MPA also decreased A. fumigatus-triggered up-regulation of costimulatory molecules (for CD1a, −68.8% ± 12.0%; for CD40, −39.3% ± 5.8%; for CD80, −46.7% ± 1.3%; for CD83, −21.1% ± 6.3%; and for CD86, −43.8% ± 2.5% [three replicates]) compared to A. fumigatus-stimulated DCs without the addition of MPA. These results indicate the decreased maturation capacity of DCs in the presence of MPA. Thus, the ability of DCs to stimulate T cells might be lower after MPA exposure (4). In accordance, Hesselink et al. reported that MPA affects reconstitution of DCs in renal transplant patients (10).

A. fumigatus germlings induced expression of cytokine (CCL20, CCL5, CXCL10, tumor necrosis factor alpha [TNF-α], and interleukin-12 [IL-12]) and immune receptor (Toll-like receptor 2 and PTX-3) genes in iDCs (www.ncbi.nlm.nih.gov/geo/) (Gene Expression Omnibus accession number GSE6965). MPA added simultaneously to fungal cocultures did not markedly affect this genomewide expression pattern (13). If DCs were treated with MPA on day 2 after monocyte isolation and cocultured with A. fumigatus on day 7, again, no alteration in CCL20, CXCL10, TNF-α, and IL-12 gene levels was observed compared to those in DCs without MPA supplementation. However, if MPA was administered from the beginning of the culture period (for 7 days), gene expression and secretion of proinflammatory cytokines were markedly reduced after exposure to A. fumigatus (Fig. 2). Expression of the housekeeping gene h-Alas showed no alleviation after MPA treatment. Similar impairment has been observed for lipopolysaccharide-induced TNF-α and IL-12 release after treatment of murine DCs with MPA (7). These results suggest that administration of MPA to DC progenitors affects differentiation, maturation, and cytokine responses of DCs in their interaction with A. fumigatus.

FIG. 2.

DCs were generated in the absence or presence (10 μM; administered on day 0) of MPA. After 7 days, DCs were washed, counted, and stimulated with A. fumigatus germ tubes (MOI = 1) for 6 h, followed by quantification of the TNF-α and IL-12 genes by quantitative reverse transcription-PCR. Relative expression (normalized to the h-ALAS housekeeping gene) is representative for three separate donors. I, unstimulated; II, cultured with MPA; III, cultured with A. fumigatus germ tubes; and IV, cultured with MPA and A. fumigatus germ tubes. MPA-induced reductions are depicted as means and standard deviations.

In conclusion, the risk of A. fumigatus infection with MPA therapy might be due to impaired function of T cells and B cells (5) and dysfunction of DCs. ROI production by PMN is increased but does not lead to enhanced fungal killing.