Experimental study design. Fertilized eggs at 0 days post fertilization (dpf) were harvested and injected with 200 CFU Mycobacterium marinum at 1 dpf. After 2 days of establishing the infection, fluorescence imaging was performed (orange arrow; n ≥ 20 larvae per group), and the treatment with isoniazid dissolved in the external treatment medium (0.25–10‐fold MIC, MIC = 15 mg·L⁻¹, and control) was started. Fluorescence imaging was repeated daily (orange solid arrows). In satellite larvae groups, destructive homogenate and blood samples for internal isoniazid exposure quantification were taken from 0 to 50 h of treatment (blue dashed arrows)

Internal isoniazid exposure over time in zebrafish larvae for increasing isoniazid doses. Internal exposure as pmol per larva in homogenate samples (left five panels) or as pmol·μl⁻¹ in blood samples (right panel) is shown on a semi‐logarithmic scale for waterborne doses in the external treatment medium, from 0.5‐ to 10‐fold MIC, as indicated in each graph (MIC = 15 mg·L⁻¹), for a constant treatment period of 50 h. Internal exposure linearly increases with dose, and steady state amounts increase with age, suggesting increased net absorption. Open symbols show observations below LLOQ

Bacterial burden in individual zebrafish larvae quantified by fluorescence imaging. (a) Representative images (brightfield [top], red fluorescence channel [middle], and overlay [bottom]) for control and 5‐fold MIC treatment groups at 2, 3, and 4 dpi (MIC = 15 mg·L⁻¹). (b) The bacterial burden in fluorescent pixel count quantified by automated image analysis for control and treatment groups with doses 0.25–5× MIC (at least n = 17 individual larvae, measured daily). Symbols represent observations, while lines represent median and quantiles with the inter‐quantile range as shaded area

Schematic representation of the pharmacokinetic–pharmacodynamic model quantifying the internal exposure of isoniazid and its response on the bacterial burden in zebrafish larvae. Compartments represent (a) drug concentration or (b) bacterial count inside the larva, solid straight arrows represent pharmacokinetic mass transfer, the curved arrow represents bacterial growth, and the dashed arrow represents drug response. The top compartment shows the pharmacokinetic component of the model for isoniazid (INH) with a first‐order absorption rate constant (k_a) from the external treatment medium, on which larval age is included as a covariate (Equation 10), distribution volume (V_d), and first‐order elimination rate constant (k_e). The bottom compartment shows the bacterial burden (Bac.) with exponential growth rate (k_g) as growth function and inoculum at treatment time point zero (INOC). The exposure–response relationship (EFF) is quantified with a linear model with a slope (SLP). dpf, days post fertilization; GI, gastrointestinal; hpf, hours post fertilization; k_a,0, absorption rate constant at the median age of 101 hours post fertilization (hpf); k_a,GI, discrete factor with which the absorption rate constant increased at 4 days post fertilization (dpf) due to opening of the gastrointestinal tract; k_a,hpf, constant in the exponential covariate relationship of age on absorption

Model‐based prediction of the internal isoniazid exposure in zebrafish larvae of the final pharmacokinetic–pharmacodynamic model. Visual predictive check with median (solid line) and 95% prediction interval (dashed lines, shaded area) from 500 simulations based on the pharmacokinetic component of the final model shows good prediction of the observed data (symbols, at least n = 3 with five larvae per sample) of internal exposure obtained after constant waterborne isoniazid treatment of 0.5‐ to 10‐fold MIC as indicated in each graph (MIC = 15 mg·L⁻¹)

Observed and model‐based prediction of the bacterial burden after isoniazid treatment in individual zebrafish larvae infected with Mycobacterium marinum. The bacterial burden as log₁₀‐transformed fluorescent pixel count is shown over treatment time of 50 h at isoniazid doses in the external treatment medium of 0.25×, 0.5×, 1×, 2×, and 5‐fold MIC, in addition to control. Symbols represent observed data (at least n = 17 individual larvae, measured daily), and lines connect model predictions. Biological variability as quantified by the model is relatively large in contrast to experimental variability (see Figure S3 for individual predictions)

Translation of isoniazid response in zebrafish larvae to humans, using a model‐based pharmacokinetic–pharmacodynamic approach. (a) isoniazid concentration–time profile (median: solid line, 80% prediction interval: shaded area) after 7 days of daily isoniazid doses of 150, 300, and 450 mg as simulated from a previously published pharmacokinetic model (Wilkins et al.,  2011). (b) Simulated median (solid line) and 80% prediction interval (shaded area, including biological and experimental variability) bacterial burden in CFU·ml⁻¹ sputum based on the human isoniazid concentration–time profile for 1,000 individuals per dose group. Concentration–time profiles were linked to the exposure–response relationship quantified in zebrafish larvae together with the translational factors on isoniazid sensitivity (MIC) and stage of infection (logarithmic vs. stationary). Translated response corresponds well to the observed bacterial burden in sputum, from Hafner et al., (1997; triangles), Johnson et al., (2006; squares) and Li et al., (2010; circles). Orange part of the prediction is extrapolated in time from the 48 h of treatment studied in the zebrafish, shown in black