Analysis tools to quantify dissemination of mycobacterial infection. Limitations of existing outcome measurements of infection, illustrated in schematic diagrams representing different distributions of fluorescent bacterial foci (green). Area of fluorescent signal and number of foci distinguish some distributions of infection (ai) but not others (aii–iii). We propose fluor₅₀, the number of foci that contribute 50% of the total fluorescence, to distinguish differences in the proportional distribution of the total burden of pathology at each site (aii) and parameters that quantify the spatial distribution of foci to differentiate localised dissemination from distant dissemination (aiii). (bi) Grid analysis divides the image into an array of squares, then quantifies the number of grid zones containing the centre point of ≥1 foci (highlighted blue). (bii) The area of a polygon (highlighted blue) encompassing the centre points of all foci. (biii) The maximum distance between the centre points of any two foci (IFD_max). (c–h) Quantitation of bacterial dissemination in images of zebrafish larvae four days post intravenous infection with 200–400 cfu Mycobacterium marinum expressing mWasabi, selected from three independent experiments, classified as having minimally, moderately or widely disseminated infection (n = 11, 8 and 8, respectively). (c) The relationship between cumulative percentage fluorescence and the number of foci which generate this signal. (d) Representative images from the data presented in (e–h) of zebrafish larvae with each category of dissemination. Scale bars, 500 µm. (e) Fluor₅₀, calculated by interpolation of the data shown in (c). Spatial measurements of dissemination, (f) grid analysis, (g) polygon area and (h) IFD_max. Lines (c) and data points (e–h) represent individual zebrafish larvae. Lines and error bars (e–h) represent median ± IQR. p values were derived from Kruskal-Wallis tests with Benjamini, Krieger and Yekutieli correction for multiple comparisons.

Quantitation of bacterial dissemination in response to a dose titration of intravenous Mycobacterium marinum (Mm) infection. Measures of dissemination in zebrafish larvae four days after intravenous infection with 25, 100 or 400 cfu Mm ± 400 µM isoniazid (I) (n = 24, 28, 22 and 25, respectively). (a) Representative images are shown for each inoculum dose. Scale bars, 500 µm. (b) The relationship between cumulative percentage fluorescence and the number of bacterial foci responsible for this signal. (c) Fluor₅₀, calculated by interpolation of the data shown in (b). The three spatial dissemination parameters, (d) grid analysis, (e) polygon area and (f) IFD_max. Lines (b) and data points (c–f) represent individual zebrafish larvae. Lines and error bars (c–f) represent median ± IQR. p values were derived from Kruskal-Wallis tests with Benjamini, Krieger and Yekutieli correction for multiple comparisons. Data are representative of three independent experiments.

Correlation matrices of relationships between integrated fluorescence, number of foci and parameters of dissemination. Spearman rank correlation matrices of the associations between existing measurements of outcome of bacterial infection; integrated fluorescence (IF) and number of foci and our new measures of dissemination; fluor₅₀, grid analysis (Grid), polygon area (Polygon) and IFD_max in (a) the images of Mycobacterium marinum (Mm) infected larvae used to develop our analysis tools (n = 27) and (b) the Mm dose titration experiment (n = 99). Data presented in (a) were derived from three independent experiments and (b) are representative of three independent experiments.

Comparison of localised and systemic Mycobacterium marinum (Mm) infection. (a) Representative images are shown for zebrafish larvae four days after either localised hindbrain ventricle (HBV) infection (n = 14) or systemic intravenous (IV) infection (n = 15) with 100 cfu Mm. Scale bars, 500 µm. Existing outcome measures, integrated fluorescence (IF), a surrogate for total bacterial burden (b) and the number of fluorescent bacterial foci (c) and dissemination parameters, fluor₅₀ (d), grid analysis (e), polygon area (f) and IFD_max (g) are presented. Data points represent individual zebrafish larvae. Lines and error bars represent median ± IQR. p values were derived from Mann-Whitney tests. Data are representative of three independent experiments. See also Supplementary Fig. S7.