Sequence and secondary structure of identified aptamers against AG as predicted with DNAMAN version 6.0.

H&E staining of lung sections from mice 4 weeks after intranasal infection with M. bovis BCG (D and E) or Mtb H37Rv (F and G) in the absence or presence of intranasally administrated AG aptamers (1 μg) once at a 1‐week interval. Quantification of lung inflamed regions shown in (E and G).

qPCR analysis of Mmps including Mmp9 (B), Mmp10 (C), and Mmp12 (D) mRNA from control or Galectin‐9 knockdown THP‐1 cells stimulated with AG (1 μg/ml) for indicated times.

Diagram showing the gRNA‐targeting genome sites.

Identification of galectin‐9 KO mice with PCR.

WT or Galectin‐9 KO mice were intraperitoneally treated with AG for 3 days in the absence or presence of the MMP inhibitor marimastat (10 mg/kg) given intraperitoneally prior to AG stimulation. Lung sections stained with H&E (B) and quantification of lung lesion burden from H&E‐stained sections (C).

WT or Galectin‐9 KO mice were intranasally infected with H37Rv for 4 weeks in absence or presence of intranasally administrated AG aptamers (1 μg) once at a 1‐week interval. Lung sections stained with H&E (D) and quantification of lung lesion burden from H&E‐stained sections (E).

CFU quantification of the bacterial titers of lung tissue homogenates from WT or Galectin‐9 KO mice intranasally infected with H37Rv for 4 weeks in the absence or presence of intranasally administrated AG aptamers (1 μg) once at a 1‐week interval.

Quantitative polymerase chain reaction (qPCR) analysis of Mmps including Mmp2, Mmp9, Mmp10, Mmp12, and Mmp13 mRNA from mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times.

Immunoblots of cell supernatants to analyze secreted MMP2, MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times; GADPH of cell lysates served as a loading control.

qPCR analysis of Mmps including Mmp2, Mmp9, Mmp10, Mmp12, and Mmp13 from the lungs of mice at indicated days post‐intraperitoneal administration of AG (100 μg).

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from mouse peritoneal macrophages stimulated with AG (1 μg/ml) for 24 h in the absence or presence of AG aptamers (0.5 μg/ml).

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times in the absence or presence of AG aptamers (0.5 μg/ml); GADPH of cell lysates served as the loading control.

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from the lungs of mice at 3 days post‐intraperitoneal administration of AG (100 μg) in the absence or presence of AG aptamers. AG aptamers (1 μg/mouse) were intranasally administrated per day.

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from mouse peritoneal macrophages infected with H37Rv for 24 h (MOI = 5) in the absence or presence of AG aptamers (1 μg/ml).

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages infected with H37Rv for indicated times (MOI = 5) in the absence or presence of AG aptamers (0.5 μg/ml); GADPH of cell lysates served as the loading control.

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from lungs of mice intranasally infected with H37Rv (2 × 10⁶ cfu/mouse) for 4 weeks in the absence or presence of AG aptamers. AG aptamers (1 μg/mouse) were intranasally administrated once at a 1‐week interval.

Coomassie blue‐stained SDS–PAGE of carbohydrate recognition domain CRD1 and CRD2 of galectin‐9. Data are representative of n = 3 independent experiments.

SPR assay of the interaction of AG with CRD2 of galectin‐9. Curve fittings to a 1:1 Langmuir‐binding model calculated with TraceDrawer are shown as smooth black lines.

FACS assay showing interactions of Mtb H37Rv with different galectins including galectin‐3, galectin‐8, and galectin‐9. The blank control was H37Rv staining with APC‐anti‐rabbit antibody alone.

FACS assay of interactions of galectin‐9 with mycobacteria including M. bovis BCG and M. smegmatis mc²155 as well as E. coli. The blank control was BCG staining with APC‐anti‐rabbit antibody alone. Data shown are representative of n = 3 independent experiments.

FACS assay of interactions of galectin‐9 with wild‐type M. marinum (Mm_WT) and MMAR‐5356 and MMAR‐5357 mutants of M. marinum treated with Tet. The blank control was Mm_WT staining with APC‐anti‐rabbit antibody alone. Data shown are representative of n = 3 independent experiments.

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from peritoneal macrophages left unstimulated (NT) or stimulated with AG (1 μg/ml) in the absence or presence of TAK1 inhibitor 5Z‐7‐OZ (1 μM) for 24 h.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times in the absence or presence of TAK1 inhibitor 5Z‐7‐OZ (1 μM); GADPH of cell lysates served as the loading control.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages infected with H37Rv for indicated times (MOI = 5) in the absence or presence of TAK1 inhibitor 5Z‐7‐OZ (1 μM); GADPH of cell lysates served as the loading control.

Immunoblots of cell lysates of HEK293T cells stimulated with AG (1 μg/ml) for the indicated time to analyze p‐ERK1/2 and p‐TAK1(T187). GADPH of cell lysates is shown as loading control.

Immunoblots of lysates of HEK293T cells stimulated with AG (1 μg/ml) for 3 h after transfection of the indicated plasmids for 48 h.

Immunoblots of cell lysates were performed to analyze p‐ERK1/2, p‐JNK, p‐P38, and p‐P65 by mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times left untreated or pretreated with AG aptamer (1 μg/ml), and GADPH as a loading control.

qPCR analysis of Mmps including Mmp9 (C), Mmp10 (D), and Mmp12 (E) from THP‐1 cells stimulated with AG (1 μg/ml) for 24 h in the absence or presence of different inhibitors targeting NF‐κB (PDTC), ERK (PD98059), JNK (SP600125), and p38 (SB203580) at the concentration of 10 μM.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages stimulated with AG (1 μg/ml) for indicated times in the absence or presence of inhibitor targeting ERK (PD98059) at the concentration of 10 μM; GADPH of cell lysates served as the loading control.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages infected with H37Rv for indicated times (MOI = 5) in the absence or presence of inhibitor targeting ERK (PD98059) at the concentration of 10 μM; GADPH of cell lysates served as the loading control.

In vitro glutathione S‐transferase (GST) precipitation assay purified histidine (His)‐tagged Galectin‐9 (+) with GST alone or GST‐tagged TAK1.

Confocal microscopy of mouse peritoneal macrophages left untreated (NC) (upper row) or stimulated with AG (1 μg/ml) for 2 h (middle row) or infected with H37Rv for 3 h (MOI = 5) (bottom row), staining with anti‐Galectin‐9 and anti‐TAK1 antibody. DAPI, nuclei, blue. Scale bar, 5 μm. Data in the right graph show mean ± SD of n = 12 fields from three independent experiments. The symbols indicate the colocalization ratio of at least 10 cells in each field.

Immunoblots of cell lysates to analyze phosphorylated TAK1 by mouse peritoneal macrophages isolated from WT or Galectin‐9 KO mice stimulated with AG (1 μg/ml) for indicated times; GADPH of cell lysates served as the loading control. Data are representative of at least n = 3 independent experiments.

Immunoblots of cell lysates of peritoneal macrophages isolated from WT or Galectin‐9 KO mice stimulated with AG (1 μg/ml) in the absence or presence of TAK1 inhibitor 5Z‐7‐OZ (1 μM) for indicated times. Data are representative of n = 3 independent experiments.

qPCR analysis of Mmps including Mmp9 (G), Mmp10 (H), Mmp12 (I), and Mmp13 (J) from WT or Galectin‐9 KO mouse peritoneal macrophages stimulated with AG (1 μg/ml) for 24 h in the absence or presence of TAK1 inhibitor 5Z‐7‐OZ (1 μM).

GO class of gene expressions in mouse peritoneal macrophages stimulated with AG (1 μg/ml) for 24 h as identified by RNA‐seq analysis.

KEGG class of gene expressions in mouse peritoneal macrophages stimulated with AG (1 μg/ml) for 24 h as identified by RNA‐seq analysis.

qPCR analysis of Mmps including Mmp9, Mmp10, and Mmp12 mRNA from THP‐1 cells stimulated with AG (1 μg/ml) for indicated times (D) or at indicated concentrations (μg/ml) for 48 h (E).

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 mRNA from mouse peritoneal macrophages stimulated with AG at indicated concentrations (μg/ml) for 24 h.

qPCR analysis of Mmps including Mmp9, Mmp10, Mmp12, and Mmp13 from the lungs of mice at indicated concentrations (μg) for 3 days post‐intraperitoneal administration of AG.

qPCR analysis of Mmps including Mmp9, Mmp10, and Mmp12 from THP‐1 cells stimulated with AG (1 μg/ml) for 48 h left untreated or pretreated with AG aptamers AA932 or AA835 (0.5 μg/ml).

qPCR analysis of Mmps including Mmp9 (B), Mmp10 (C), Mmp12 (D), and Mmp13 (E) from mouse peritoneal macrophages isolated from either wild‐type or Galectin‐9 KO mice stimulated with AG (1 μg/ml) for 24 h.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages isolated from WT or Galectin‐9 KO mice stimulated with AG (1 μg/ml) for indicated times; GADPH of cell lysates served as the loading control.

Immunoblots of cell supernatants to analyze secreted MMP9, MMP10, MMP12, and MMP13 by mouse peritoneal macrophages isolated from WT or Galectin‐9 KO mice infected with H37Rv for indicated times (MOI = 5); GADPH of cell lysates served as the loading control.

Immunoblot of lysates of peritoneal macrophages isolated from wild‐type and Galectin‐9 KO mice stimulated with AG (1 μg/ml) for indicated times. Data are representative of n = 3 independent experiments.

Immunoblot of lysates of shCtrl and shGalectin‐9 THP‐1 cells stimulated with AG (1 μg/ml) for indicated times. Data are representative of n = 3 independent experiments.

qRT–PCR detection of Mmp transcripts including Mmp9, Mmp10, and Mmp12 in wild‐type and Galectin‐9 KO peritoneal macrophages stimulated with AG (1 μg/ml) for 24 h in the absence or presence of ERK inhibitor PD98059 (10 μM).

qRT–PCR detection of MMP transcripts including Mmp9, Mmp10, and Mmp12 in shCtrl and shGalectin‐9 THP‐1 cells stimulated with AG (1 μg/ml) for 24 h in the absence or presence of ERK inhibitor PD98059 (10 μM).

Chemical structures and conformations of β‐galactofuranoside and β‐galactopyranoside.

SPR assay of interactions of AG with indicated galectins including galectin‐1 (C), galectin‐3 (D), galectin‐7 (E), galectin‐8 (F), galectin‐14 (G), LGASL (H), and a summary table of KD (I). Curve fittings to a 1:1 Langmuir‐binding model calculated with TraceDrawer are shown as smooth black lines. The binding affinity of galectin‐9 and CRD2 to AG is highlighted in (I) in red.

SPR assay of interactions of AG with CRD1 of galectin‐9.

Coomassie blue staining of galectin‐9(1–146) and CRD2 of galectin‐9 post–SDS–PAGE analysis.

SPR assay of interactions of AG with galectin‐9(1–146).