Fig. 4.
Screen of in silico-predicted chemical compounds active against intracellular Mtb. a Schematic overview of the in silico predictive model. BLoad = bacterial load z-score, CViab = cell viability z-score. b Chemical compound primary screen (left panel) and rescreen (middle panel) at 10 µM in the MelJuSo-Mtb infection model using Mtb constitutively expressing stable DsRed, expressed as mean z-scores ± standard deviation. Horizontal dashed lines indicate a hit cutoff at a z-score of 2 or −2. The average z-score and standard deviation of the controls (DMSO and H-89) are displayed separately. To indicate the extent of bacterial inhibition, rescreen results are expressed both as z-score and as a percentage of control value ± standard deviation in the right panel. c CFU assay of the MelJuSo (left panel) and human primary Mφ1 (middle panel) and Mφ2 (right panel) Mtb infection models treated with the validated hit compounds from b at 10 µM. Mφ1 and Mφ2 models have been described by Verreck et al.78. Representative data out of three independent experiments (MelJuSo) and data from a representative donor (Mφs) out of two (Mφ1) or five (Mφ2) different healthy blood bank donors are shown. To indicate the extent of bacterial inhibition, the results are expressed as a percentage of control value ± standard deviation. Number of replicates in the MelJuSo model: AT9283 and ENMD-2076: n = 6; dovitinib, VEGFR KI I, and DAPH 2: n = 5; DMSO and H-89: n = 9. Number of replicates in the Mφ models: AT9283, ENMD-2076, dovitinib, VEGFR KI I, and DAPH 2: n = 3; DMSO and H-89: n = 5. Statistically significant difference compared to DMSO was tested using a one-way ANOVA (MelJuSo: F(6,39) = 16.35; Mφ1: F(6,18) = 10.88; Mφ2: F(6,18) = 5.23; *p value < 0.05, **p value < 0.01, ***p value < 0.001). d CFU assay of the human primary Mφ1 and Mφ2 models infected with two different MDR-Mtb strains (Beijing family China 16319 and Dutch outbreak 2003-1128) and treated with the validated hit compounds from c at 10 µM. Data (n = 3 technical replicates) from a representative donor out of four different healthy blood bank donors, displayed as a percentage of the DMSO control ± standard deviation are shown. Statistically significant differences compared to DMSO were tested using a one-way ANOVA (Mφ1 Beijing: F(4,10) = 11.43; Mφ2 Beijing: F(4,10) = 3.72; Mφ1 Dutch outbreak: F(4,10) = 29.09; Mφ2 Dutch outbreak: F(4,10) = 8.81; *p value < 0.05, **p value < 0.01, ***p value < 0.001). e Six-day treatment of Mtb broth cultures with the hit compounds at 10 µM. The Mtb antibiotic rifampicin (20 μg/ml) was used as a positive control. The average bacterial density ± standard deviation of three replicates is shown, expressed as a percentage of the DMSO control. Representative results out of three individual experiments are displayed. Statistically significant difference compared to DMSO was tested using a one-way ANOVA (F(6,25) = 101.4; **p value < 0.01, ***p value < 0.001)