The co-infection plate was synchronised for 5 min at 21 °C and subjected for 1 h incubation at 37 °C in a humidified CO2 incubator. After 1 h, the phagocytosis was stopped by washing with ice-cold PBS. Counter-staining of spores that are not phagocytosed was performed with 0.5 mg ml−1 CFW (calcofluorwhite; Sigma) in PBS for 15 min at room temperature. The cells were washed twice with PBS then fixed with 3.7% (vol/vol) formaldehyde/PBS for 15 min followed by another two washes BMN 673 supplier with PBS. Microscopic photographs were taken with Leica DM 4500B at a magnification of 40×. For statistical reproducibility, three biological replicates and in each case two technical replicates were performed

and analysed for each strain. The automated image analysis was performed by an algorithm that was previously implemented and rigorously validated in the context of phagocytosis assays for A. fumigatus conidia[16] and of invasion assays for Candida albicans.[20] The algorithm was developed within the Definiens Developer XD framework where the ruleset comprising all commands is written in a meta-language. Processing

the current image data of phagocytosis assays at a high level of performance was achieved by modifying this algorithm with regard to the second of its three main steps: (i) preprocessing, (ii) segmentation and (iii) classification. Each image is built of three distinct layers, one for each fluorescent label, and a schematic selleck chemicals representation of the ruleset acting on the three colour layers containing all spores (green layer), non-phagocytosed spores (blue layer) and macrophages (red layer)

is depicted in Fig. 1. Apart from a modification in the segmentation step, the original algorithm was applied for parameters values summarised in Table 1 that were adjusted to the images of size of 1600 × 1200 pixels with a pixel area of 0.0246 μm2 and a corresponding pixel-to-pixel AMP deaminase distance of 0.157 μm. After the ruleset-based image data analysis was performed, features obtained for all four labelled classes (macrophages, phagocytosed spores, non-phagocytosed spores that can be either adherent or non-adherent to macrophages), e.g. area in pixel, layer intensity and number of neighbours of each object as well as class membership of every object, were exported and used for subsequent analyses. Finally, the number of cells per class was calculated to perform statistical analyses and validation procedures. Images were preprocessed by smoothing the three distinct layers with a Gauss filter to reduce noise (split point 1 in Fig. 1). Afterwards an edge-detection filter was applied to enhance object boundaries. This filter assigns to every pixel the maximal intensity value of its pixel neighbourhood. No further preprocessing was necessary at split point 2 in Fig. 1 to optimise the segmentation and classification of regions of interest (ROIs) in the subsequent steps. As depicted in Fig.