Female CF-1 mice (20 g, 6–8 weeks old) were purchased from Charles River Laboratory and housed in a biosafety level 2 control facility. The mice were cared in accordance with animal care regulation and use protocols approved by the City of Hope Institutional Animal Care and Use Committee. Animal numbers (n, between 8 and 26) per group are indicated in the figure captions.

Aspergillus fumigatus strain AFCOH1 (Ito et al., 2006; Ito et al., 2009; Diaz-Arevalo et al., 2011) was used in all experiments and was cultured on potato dextrose agar (BD/Difco, Aspergillus fumigatusFranklin Lakes, NJ, USA) for 7 days at 37°C. Resting conidia were harvested by rinsing from the cultured fungus, and resuspended in Dulbecco’s phosphate-buffered saline without calcium and magnesium (DPBS, Mediatech, Inc., Manassas, VA, USA) prior to infection. Conidia were quantified with a Countess Automated Cell Counter (Invitrogen, Eugene, OR, USA), and their viability was measured by colony forming unit assay.

CF-1 mice were immunized twice, once per day on days 0 and 14, with subcutaneous injections of N-terminal truncated rAsp f3-based vaccine (15 μg) suspended in TiterMax (TiterMax, Inc., Norcross, GA, USA) as previously described (Ito et al., 2006). For neutrophil depletion assays, mice were injected with the vaccine a third time on day 60. As controls, mock vaccinated mice were injected with PBS plus TiterMax. Prior to infection, one of the following immunosuppression treatments was applied.

Vaccinated mice that survived were euthanized 5 days after A. fumigatus infection and their lungs collected and immediately fixed in 10% formalin and embedded in paraffin. The lungs of non-surviving mice were collected from four mice at the time of death. Lung sections were stained with hematoxylin and eosin (HE), and Gomori methenamine silver staining was used to stain hyphal mycelium and remaining conidia.

Immunohistochemistry was performed using anti-CD3 (Dako Inc., Carpinteria, CA, USA), anti-GR1, and anti-F4/80 (eBioscience, Inc., San Diego, CA, USA) primary antibodies and horseradish peroxidase (HRP) conjugated anti-rat IgG (Alpha Diagnostic Intl. Inc., San Antonio, TX, USA) as the secondary antibody. Imaging was performed on an Olympus AX70 model U-MPH microscope (Tokyo, Japan) with a QImaging RETIGA EXi camera. Data were acquired with Image ProPlus v5.1 software.

We first assessed the susceptibility of mice to invasive pulmonary aspergillosis after induction of either CA-induced immunosuppression or antibody-mediated neutrophil depletion. When challenged with three million VC, only 10–20% of CA-immunosuppressed mice survived (Figure 1A, curve Rat IgG/CA). In contrast, ∼67% of the neutrophil-depleted mice survived the infection (Figure 1A, curve anti-GR1). However, at an infectious dose of 30 million VC, the survival of neutrophil-depleted mice was comparable to that of CA-immunosuppressed mice that received only three million spores (6% survival vs. 10–20% survival, respectively; Figure 2C vs. Figure 1A). Therefore we concluded that, CA treatment made the mice more susceptible to A. fumigatus infection than neutrophil depletion. Because we depleted neutrophils with a rat anti-GR1 antibody, we treated CA-immunosuppressed mice with a non-specific rat IgG as a control, which had no significant effect on survival. In addition, the control group that received only rat IgG and no further immunosuppression survived the infection entirely (Figure 1A). We also combined CA-immunosuppression with anti-GR1-mediated neutrophil depletion, but this did not further increase the susceptibility of the mice to A. fumigatus infection (Figure 1A). To confirm that neutrophils were depleted, we monitored neutrophil counts by FACS analysis using Ly6C and GR1 as markers. Twenty-four hours after injection of anti-GR1 antibody, neutrophil counts were reduced to 0.033% of normal levels (Figure 1B). Treatment with non-specific rat IgG did not change the neutrophil population of the mice (Figure 1B).

When we attempted to deplete neutrophils in CA-immunosuppressed mice, we observed that a population of neutrophils was still present after treatment, but their staining with an APC-labeled anti-GR1 antibody had shifted to lower intensities in the FACS analysis (Figure 1C). Furthermore, the circulating neutrophils were still coated with the anti-GR1 antibody that was given to induce their depletion (Figure 1D). Others have observed a similar shift in the population of labeled neutrophils when attempting to antibody-deplete neutrophils after reduction of peritoneal macrophage numbers with Clodronate liposomes (Mircescu et al., 2009). No significant reduction of monocytes and CD8⁺ cells was observed after neutrophil depletion (Figure 1E).

To study the role of neutrophils in rAsp f3-vaccinated mice, we depleted neutrophils with anti-GR1 antibody (Figure 2A). Administration of the antibody led to systemic depletion of neutrophils, down to 0.004% of the population, while neutrophils composed 1.26% of the population in mice that were immunosuppressed with CA (Figure 2B). Neutrophil depletion partially affected the survival of vaccinated mice in that 40% of the mice remained protected after A. fumigatus infection. In contrast, non-immunized, neutrophil-depleted mice were highly susceptible to the infection (Figure 2C). Although the observed difference in survival lacks strict statistical significance (p = 0.056) a trend of partial vaccine protection of neutrophil-depleted mice is apparent.

As a control, we immunosuppressed rAsp f3-vaccinated mice with CA and then challenged them with 30 million conidia. Only 20% of these mice survived A. fumigatus infection, and no statistical difference in survival (p = 1) was observed between rAsp f3-immunized mice and non-vaccinated mice (Figure 2D). We attribute the high mortality of vaccinated mice to the very high number of conidia (30 M) used for challenge. Consistent with our previous observations (Ito et al., 2006; Diaz-Arevalo et al., 2011), vaccinated, CA-immunosuppressed mice challenged with three million conidia had significantly enhanced survival (p = 0.009) over that of non-vaccinated mice (Figure 2E).

Histological analysis using Gomori methenamine silver staining of the lungs of non-immunized, neutrophil-depleted mice showed extensive hyphal tissue invasion (Figure 3A vs. Figure 3B) characterized by hemorrhagic foci, with inflammatory infiltrates, necrotic tissue, and edema (Figure 3C vs. Figure 3D). The fungi were located in a few patchy areas in the lungs, and the mycelium contained multiply branched, densely grown hyphae (Figure 3B). In contrast, CA-immunosuppressed, non-vaccinated mice, had hyphal elements throughout the entire lung, but the mycelium consisted of only short hyphal fragments with very few branches (Figure 3F). In contrast, lungs of rAsp f3-vaccinated CA-immunosuppressed (Figure 3E) and neutrophil-depleted mice (Figure 3A) had no hyphae and only a small number of non-germinated conidia were present in the lung parenchyma.

Lungs of neutrophil-depleted, rAsp f3-vaccinated mice showed only mild infiltration of macrophages and CD3⁺ T cells in the peribronchial tissue (Figures 4A–C). Vaccinated CA-immunosuppressed mice had a similar type of peribronchial infiltrate that included neutrophils (Figures 4G–I). The infiltrate of immune cells in neutrophil-depleted, non-immunized mice was characterized by presence of CD3⁺ T cells and absence of neutrophils and alveolar macrophages (Figures 4D–F). The lungs of non-vaccinated, CA-immunosuppressed mice exhibited extensive infiltration of PMNs and CD3⁺ T cells, but almost completely lacked alveolar macrophages (Figures 4J–L), and exhibited edema and focal hemorrhage (Figure 3H).

In the current study, we demonstrated that rAsp f3-vaccination shows a protective trend in neutrophil-depleted mice (40% survival) challenged with pulmonary A. fumigatus infection. Consistent, with previous work by others (Mircescu et al., 2009; Ibrahim-Granet et al., 2010), our data suggest that neutrophils are essential for the innate immune protection of immunocompetent animals against A. fumigatus. Furthermore, the protective effect of the vaccine can only be observed in immunosuppressed mice, as it has not been possible to induce fatal pulmonary A. fumigatus infections in immunocompetent mice. However, neutrophils do not appear to be the key effector cells that provide vaccine protection against A. fumigatus, because neutrophil depletion did not abolish the protective vaccine effect entirely.

A secondary auxiliary role in antifungal protection was previously attributed to neutrophils in an in vitro study in which specific T cells in combination with antigen presenting cells and neutrophils enhanced the damage to cultured A. fumigatus hyphae (Beck et al., 2006). It must be pointed out that the immune system of immunosuppressed mice can be overwhelmed when the infectious dose is sufficiently large, regardless of vaccination status. However, the markedly higher susceptibility of CA-immunosuppressed mice as compared to that of neutrophil-depleted mice indicates that macrophages contribute substantially to innate antifungal protection, as CA is known reduce the killing potential of macrophages (Schaffner, 1985; Ibrahim-Granet et al., 2003; Philippe et al., 2003), and it elicits a broad effect on the immune system by reducing the numbers of immune cells and cytokine signaling (Almawi and Melemedjian, 2002).

The RB6-8C5 antibody is known to bind to the neutrophil cell surface antigens Ly6G and Ly6C, the latter with lower affinity (Fleming et al., 1993). Because Ly6C is not only present on neutrophils, but also on monocytes (Henderson et al., 2003), plasmacytoid dendritic cells (Nakano et al., 2001), and a subpopulation of CD8⁺ cells (Matsuzaki et al., 2003), we also analyzed PBMCs for a reduction of monocytes and CD8⁺ cells, but found it not to be statically significant (Figure 1E) and consistent with previous observations (Mircescu et al., 2009). Although the more specific anti-Ly6G antibody 1A8 can alternatively be used to deplete neutrophils (Daley et al., 2008), we chose the RB5-8C5 antibody, which is readily available, has been well characterized, and used extensively to study the role of neutrophils in A. fumigatus infection (Mehrad et al., 1999; Mircescu et al., 2009; Ibrahim-Granet et al., 2010).

The IA histopathology of CA-immunosuppressed mice was markedly different from that of neutrophil-depleted mice. The absence of a robust macrophage response in the non-immunized neutrophil-depleted mice may explain why A. fumigatus was able to invade the lungs and grow into dense patches with highly branched hyphal mycelium. Furthermore, the presence of a strong neutrophil infiltrate in CA-immunosuppressed mice, may explain why the hyphal elements appeared much shorter, because they may be constantly damaged by neutrophils, although this damage is not sufficient for fungal clearance. This observation is consistent with the suggestion that steroid-induced immunosuppression may lead to an inflammatory response that is destructive to lung tissue (Balloy et al., 2005; Stephens-Romero et al., 2005). However, neutrophil-depleted mice were not protected from IA either, and their inflammatory infiltrate at the time of death consisted mainly of CD3⁺ T cells and largely lacked macrophages. It is possible that, due to the larger infectious inoculum used, the macrophages were simply overwhelmed or their resident population was exhausted. Interestingly, the CD3⁺ T cell infiltrate was unable to recruit further macrophages, whereas in mice immunized with the rAsp f3-vaccine, T cell recruitment of macrophages to the point of infection appears to be associated with successful clearance of the fungal pathogen (Figures 3 and 4). Our observation does not exclude the possibility that other cell populations such as CD8⁺ T cells and natural killer (NK) cells could contribute to vaccine based antifungal protection; however, the selective depletion of CD4⁺ cells readily abolishes rAsp f3-based vaccine protection (Diaz-Arevalo et al., 2011). The involvement of NK cells in other type of vaccinations was demonstrated recently; for example the production of IL-2 by antigen specific CD4⁺ T cells enhanced NK activation after vaccination either with Plasmodium falciparum (Horowitz et al., 2010b; McCall et al., 2010), rabies virus (Horowitz et al., 2010a), or simian immunodeficiency virus (Vargas-Inchaustegui et al., 2012). Therefore, although the effector role of macrophages appears to be very likely, further investigation will be required to understand if CD8⁺ T or NK cells participate in the mechanism of the rAsp f3-based A. fumigatus vaccine.