Presence of Mtb in human miliary tuberculosis patients:

(A) Hematoxylin and eosin (H and E) staining of the human autopsied liver tissue sections showing the presence of granuloma-like structure in the Mtb infected liver. (B) Acid-fast staining shows the presence of Mtb in the liver section of miliary TB patients. (arrows point to the presence of Mtb in the enlarged image) (C-D) Auramine O-Rhodamine B (A-O) staining and fluorescence in-situ hybridization (FISH) with Mtb specific 16s rRNA primer further confirms the presence of Mtb in human liver tissue sections. (E). Dual staining of β-actin (green) and Ag85B (red) using respective antibodies shows the presence of Mtb in hepatocytes of human biopsied liver sections. (F). Dual staining of β-actin (green) and Ag85B (red) using respective antibodies shows no distinct signal of Ag85B in the uninfected liver biopsied sample (control).

Involvement of the liver and hepatocytes in a mouse aerosol model of Mtb infection.

(A). Schematic denoting the flow of experimental set up of studying the liver at tissue level and cellular level. This panel was created using BioRender.com. (B) C57BL/6 mice were infected with 200 CFU of H37Rv through the aerosol route and the bacterial burden of the lung and liver was enumerated at different time points post-infection in lung and spleen (C). Alexa Fluor 488 Phalloidin and DAPI staining of uninfected and infected lungs at 8 weeks post-infection. Images were taken in 40 X magnification as mentioned (scale bar is 20 μm). (D). Immunofluorescence staining of Alexa Fluor 488 Phalloidin, DAPI, and Ag85B in the infected mice liver at 8 weeks post-infection shows the presence of Ag85B signals within hepatocytes (magnified image) (scale bar is 20 μm). (E). Isolated primary hepatocytes stained with phalloidin green and DAPI show the typical polygonal shape with binucleated morphology. (F). Anti-asialoglycoprotein receptor (ASGR1) antibody staining specifically stains primary hepatocytes isolated from infected mice as validated by the contour plot in flow cytometry. (G). Antiasialoglycoprotein receptor (ASGR1) antibody staining specifically stains primary hepatocytes isolated from mice as visualized through confocal microscopy. (H). Primary hepatocytes were isolated from the infected mice, lysed and CFU enumeration was done at the mentioned time points. (I). Ag85B staining of cultured primary hepatocytes, isolated from mice at 8 weeks post Mtb infection. (scale bar is 20 μm). (J). Dual immunostaining of Ag85B (red) and albumin (green) shows distinct signals of Mtb within the isolated hepatocytes from infected mice at 8 weeks post-infection. (n)= 6-7 mice in each group.

Mtb uses hepatocytes as a replicative niche:.

(A and B) Representative microscopic images showing the infection of primary hepatocytes (PHCs), AML-12, HepG2, and Huh-7 with labelled Mtb-H37Rv strains and subsequent quantification of percentage infectivity in the respective cell types. RAW 264.7, THP-1 and murine BMDMs were used as macrophage controls. The scale bar in all images is 10 µm except in HepG2, which is 5 µm (C). CFU enumeration of Mtb-H37Rv in different hepatic cell lines. (D and E). Representative confocal microscopy images of Mtb-GFP-H37Rv infected PHCs at the respective time points post-infection and bar blot depicting mean fluorescent intensity (MFI) / cell at the respective time points. (F and G). Representative confocal microscopy images of MtbH37Rv-GFP infected HepG2 at the respective time points post-infection and bar blot depicting MFI/cell at the respective time points, n= 4 independent experiments, with each dot representing 5 fields of 30-60 cells. Scale bar 10 µm. (H). Fold change of growth rate of Mtb within HepG2, PHCs, RAW 264.7, THP-1 and BMDMs at the mentioned time points post-infection.

RNA sequence analysis of infected and sorted hepatocytes at 0 hours and 48 hours post-infection:

(A). Experimental setup for infection of HepG2 with Mb-H37Rv-mCherry at 10 MOI with histogram of mCherry signals at 0- and 48-hours post-infection. Schematic depicting experimental set up was created using BioRender.com. (B). Principal component analysis (PCA) plot illustrating the HepG2 transcriptome, identified through global transcriptomic analysis of 0 hours uninfected and infected and 48 hours uninfected and infected. (C). GO pathway enrichment analysis was done for DEGs with adjusted p-value < 0.05 at 0- and 48-hours post-infection. The bubble plot depicts the enrichment of pathways on the mentioned time points post-infection, where the coordinate on the x-axis represents the gene ratio, the bubble size represents the gene count and colour represents the p-value. (D and E). Volcano plot depicting the fold change of different genes in 0 hours and 48 hours post-infection, the red dots represent the significantly upregulated and downregulated genes.

Increased fatty acid biogenesis drives Mtb growth in hepatocytes:

(A). Increased number of LDs/cells in HepG2 and PHCs post-Mtb infection as compared to the uninfected cells (B). Quantification of the number of lipid droplets in the infected HepG2 and PHCs with their uninfected control, 50-70 cells were analyzed from 3 independent experiments. (C). Increase in the BODIPY intensity at different days post-infection in infected hepatocytes (D). A high degree of colocalization of Mtb-H37Rv-GFP (green) with lipid droplets (red) within PHCs (E) Mass spectrometric quantification of the different species of DAGs, TAGs, and cholesterol esters in the Mtb infected hepatocytes. (F). Confocal images showing puncta of fluorescently labelled fatty acid in Mtb derived from metabolically labelled (with BODIPY 558/568 C12) HepG2. Relative CFU of Mtb under the administration of different inhibitors of the lipid metabolic pathway in (G). PHCs (H). HepG2 and (I). THP-1. Representative data from 3 independent experiments. Data were analyzed using the two-tailed unpaired Student’s t-test in B and one-way ANOVA in C, F, G, and H.*p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001. ns=non-significant.

PPARγ driven lipid biogenesis drives Mtb growth in hepatocytes:

(A) Heat map showing the fold change of the genes involved in the lipid biosynthesis and LD biogenesis in liver across weeks 2, 4 and 8, post infection. (B) Kinetic increase in the gene expression of Pparγ transcript levels across different weeks post infection (C) Immunoblot showing increased PPARγ in PHCs (MOI: 10) at 5-, 24- and 48-hours post-infection. (D) Bar plot showing the increased band intensity of PPARγ in the infected PHCs at the mentioned time points. (E) Representative confocal microscopy images showing HepG2 infected with Mtb-mCherry-H37Rv and treated with GW9662 and rosiglitazone (F) MFI/cell of in Mtb-mCherry-H37Rv DMSO treated, GW9662 and rosiglitazone treated HepG2 cells. (G) Representative confocal images of uninfected, infected, GW9662 and rosiglitazone treated infected HepG2 showing changes in the number of LDs. (H) Plot depicting the quantification of LDs/cell in the mentioned conditions. Representative data from n=4 biological replicates. Data were analysed by using the two tailed unpaired Student’s t test in D and by one way ANOVA in B, F and H. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001, ns=non-significant

Hepatocytes provide a drug-tolerant niche to Mtb:

(A). Experimental scheme of deducting the percentage drug tolerance in hepatocytes. This panel was created using BioRender.com. (B). Percentage tolerance of Mtb-H37Rv against Rifampicin within RAW 264.7, PHCs and HepG2 at different time points post-infection. (C). Percentage tolerance of Mtb-H37Rv against Isoniazid within RAW 264.7, PHCs, and HepG2 at different time points post-infection. Representative data from 3 independent experiments. (D). Percentage tolerance of Mtb-H37Rv against rifampicin within BMDMs, PHCs, and HepG2 as measured microscopically (E). Percentage tolerance of Mtb-H37Rv against isoniazid within BMDMs, PHCs, and HepG2 measured microscopically. Representative data from 3 independent experiments. Each dot represents a single field having more than 4 infected cells. 20 such fields were analyzed (F). Transcript levels of the various DMEs involved in Rifampicin and Isoniazid metabolism in mice liver, 8 weeks post-infection (n=4 mice), the fold change has been calculated by considering the expression in the uninfected mice to be 1. Data were analyzed by using 2-way ANOVA in (B) and (C), one-way ANOVA in (D) and (E) and the two-tailed unpaired Student’s t-test in (F) and Representative data from n=4 biological replicates *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0. 0001.ns=non-significant.

Histopathological changes in the liver biopsy samples of miliaryTB patients:

(A and B): A-O staining and FISH in the liver biopsy of the uninfected (control) group shows no Mtb-specific signals. (C). H & E staining shows enhanced immune cell infiltration in the liver biopsy of the pulmonary TB patients compared to the uninfected control. (D). Dual staining of β-actin (green) and Ag85B (red) using respective antibodies shows the presence of Mtb in hepatocytes and other cells in liver biopsy sections (indicated by yellow arrows) (E). H&E staining shows distinct granuloma in the lung section of pulmonary TB patient and (F). Acid-fast staining in the same lung section shows a high bacterial load (G). A-O staining in the same lung section further validates the presence of high Mtb load. Scale bars in A, B and C are 50 µm and in D is 20 µm. and E and F are 150 µm. Data shown in this figure is representative of 5 patients with pulmonary TB.

Histopathological changes in the murine liver post Mtb infection:

(A). 105 number of H37Rv were administered intra-peritoneally (IP) and the bacterial burden was enumerated in primary hepatocytes that were isolated from the infected mice. (B). Analysis of the serum parameters of albumin, ALT, AST and GGT in the uninfected and infected mice at the mentioned time points, mice were infected through the aerosol route with 200 CFU. (C). Representative H and E-stained liver sections show enhanced immune cell infiltration in Mtb-infected mice. (D). Multiplex immunostaining of infected mice liver with CD 45.2 and Ag85B indicates the presence of Ag85B positive signals in CD 45.2 cells. Sample size (n)= 5/6 mice in each group.

Involvement of the liver in the guinea pig aerosol infection model:

(A). Guinea pigs were infected with 200 CFU of H37Rv through the aerosol route and the bacterial burden of the lungs, liver and spleen was enumerated at different time points post-infection (B). H and E images of the lung and the liver at 4- and 8-weeks post infection showing distinct granulomas in both the organs. Sample size (n)= 3 guinea pigs in each group.

Standardization of multiplicity of infection (MOI) in PHCs and HepG2:

(A). Confocal microscopic images of Phalloidin Alexa fluor 555 stained PHCs infected with different multiplicity of infection (MOI) of Mtb-H37Rv-GFP. (B). Quantification of % infectivity at different MOIs in primary hepatocytes (PHCs) (C). Quantification of % infectivity at different MOIs in HepG2. Relative CFU of Mtb in (D). HepG2, (E). PHCs, (F). RAW 264.7, (G). THP-1, (H). murine BMDMs at the respective time points post-infection.

Comparison of significant pathways between Mtb infected THP-1 and HepG2, 48 HPI:

(A). GO pathway enrichment analysis was done for DEGs with adjusted p-value < 0.05 at 48 hours post-infection for Mtb-infected THP-1. (B). GO pathway enrichment analysis was done for DEGs with adjusted p-value < 0.05 at 48 hours post infection for Mtb-infected HepG2. The bubble plot depicts the enrichment of pathways on the mentioned time points post-infection, where the coordinate on the x-axis represents the gene ratio, the bubble size represents the gene count and colour represents the p value The bubble plot depicts the enrichment of pathways on the mentioned time points post-infection, where the coordinate on the x-axis represents the gene ratio, the bubble size represents the gene count and colour represents the p-value.

Increased lipid biogenesis provides a growth advantage to Mtb within hepatocytes:

(A). Confocal microscopy showing the LDs in both PHCs and HepG2 under inhibitors of fatty acid biosynthesis and TAG biosynthesis. (B). Mtb RNA isolated from DMEM grown Mtb and HepG2-derived Mtb,48 HPI, with the RNA expression pattern of key lipid biogenesis genes and TAG biosynthesis genes in Mtb as determined by qRT PCR, the fold change has been calculated by considering the expression in the DMEM grown Mtb to be 1 (C). Experimental plan for metabolic labelling of Hepg2 and subsequent infection by Mtb. The schematic of the experimental procedure was created using BioRender.com. (D). Fluorescently labelled fatty acid (BODIPY 558/568 C12) staining in DMSO treated, C75 treated and T863 treated HepG2. Data were analyzed by using the two-tailed unpaired Student’s t-test in B. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001, ns=non-significant.

Confocal microscopy images showing the puncta of fluorescently labelled fatty acid (BODIPY558/568 C12) in Mtb isolated from metabolically labeled HepG2:

(A). Confocal images showing puncta of labeled fatty acid in Mtb derived from DMSO, C75 and T863 treated metabolically labeled HepG2.

Enhanced lipid accumulation in the liver of Mtb infected mice:

(A). BODPIY staining in the liver of the uninfected and 8 weeks post Mtb infected murine liver (B). Dual staining of BODIPY (green) and Ag85B (red) shows Ag85B colocalization with BODIPY (highlighted in the zoomed image). (C). Representative images showing the mean lipidTOX Neutral Red staining in infected liver, 2-, 4- and 8-weeks post infection (D). Plot depicting the mean lipidTOX intensity at the mentioned time points post-infection (normalized to the area. Data were analyzed one-way ANOVA. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001. ns=nonsignificant in I. Representative data from experiments having n=4-6 mice/ group.

Mtb induced PPARγ expression in infected hepatocytes:

(A). mRNA levels of PPARγ and other lipid biosynthetic genes in infected HepG2, the fold change has been calculated by considering the expression in the uninfected cell to be 1 (B). Relative bacterial burden in DMSO, GW9662 and rosiglitazone treated HepG2 infected with Mtb H37Rv. (C). Multiplex immunostaining showed increased expression of PPARγ in the liver of infected mice, 8 weeks post-infection. (D). Bar plot showing PPARγ intensity (normalized to the area) in the uninfected and infected liver, 8 weeks post-infection. (E). Intensity of PPARγ per hepatocyte in uninfected and infected liver, 8-week post-infection. Data were analyzed by using the two-tailed unpaired Student’s t-test in (A, D, and E) and by one-way ANOVA in B. Representative data from n=4 biological replicates *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001. ns=non-significant.

Gene expression analysis of key DMEs related to rifampicin and isoniazid metabolism:

(A). Relative expression of CYP3A4, CYP3A43, and NAT2 genes in infected HepG2 post-Mtb infection, the fold change has been calculated by considering the expression in the uninfected cells to be 1. Data were analyzed using the two-tailed unpaired Student’s t test in C. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001, ns=non-significant. (B). KEGG pathway analysis shows the upregulation of NAT2 gene (N-acetyltransferase 2) in Mtb infected HepG2

Microscopic images of Mtb infected BMDMs, PHCs and HepG2 treated with rifampicin and isoniazid:

(A). Confocal microscopy images of Mtb-H37Rv-GFP infected BMDMs, PHCs, and HepG2. All the cells are treated with DMSO, 0.5 μg/ml rifampicin, and 0.5 μg/ml isoniazid post-infection. The cells were stained with Phalloidin Alexa fluor-555 to demarcate the cell boundary. Representative data from 3 independent experiments. Each dot represents a single field having more than 4 infected cells.20 such fields were analyzed. Figure

Mtb derived from hepatocytes show no changes in drug efflux pumps and antibiotic sensitivity:

(A). Resazurin microtitre assay of the three groups of Mtb in different concentrations of isoniazid and rifampicin. The schematic of the experimental procedure was created using Biorender.com. (B). Mtb was obtained from DMSO-treated, 0.5 μg/ml rifampicin, and 0.5 μg/ml isoniazid-treated HepG2 by differential lysis and RNA was isolated from those Mtb populations (C and D). qRT-PCR of some of the key efflux pumps genes in these three Mtb groups. Data were analysed using the two-tailed unpaired Student’s t-test in C and D. *p < 0.05, **p < 112 0.005, ***p < 0.0005, ****p < 0.0001, ns=non-significant.