Exploiting thioether reactivity to label mycobacterial glycans

Abstract

Mycobacterium tuberculosis (Mtb) is a leading cause of death worldwide. Mtb cell envelope glycans are potent virulence factors that play key roles in mediating infection of host tissues and modulating the host immune response. However, there are few ways to site-selectively modify and label these or any glycans to study their functions in disease. This gap arises because glycans generally lack functional groups amenable to bioconjugation strategies. Methylthioxylofuranose (MTX), a rare monosaccharide in select pathogenic mycobacteria, is an exception. MTX is appended to mannose-capped lipoarabinomannan (ManLAM), an antigenic glycolipid in the Mtb cell envelope implicated in downregulating the host immune system during infection. MTX is unique not only in its prevalence but also in its functionality-it contains a thioether not present in other glycans. We envisioned exploiting the MTX thioether to selectively label ManLAM with an oxaziridine probe. Here, we show that MTX-containing glycans can be labeled selectively in the test tube and live cells, highlighting the reactivity and accessibility of this motif. Our approach labels ManLAM efficiently despite the presence of protein methionine residues and can distinguish between different mycobacterial species. Using an oxaziridine equipped with a reporter, we could visualize ManLAM localization in live cells and a macrophage infection model, highlighting the stability of the label and the cell envelope in this environment. These studies will enable investigations of dynamic changes in a critical Mtb cell envelope component during infection. Moreover, the selective reactivity of thioethers can be leveraged to expand the repertoire of glycan bioconjugation strategies.

Keywords: Mycobacterium tuberculosis; bioconjugation; glycans; mannose-capped lipoarabinomannan; mycobacteria.

Fig. 1.

Overview of the mycobacterial cell wall (AG: arabinogalactan; MA: mycolic acids; LAM: lipoarabinomannan; LM: lipomannan; PIM: phosphoinositolmannoside), highlighting a key monosaccharide, methylthioxylofuranose (MTX). (A) Mannose-capped lipoarabinomannan (ManLAM) is an antigenic glycolipid present in the Mtb cell envelope. Mtb ManLAM is capped with the unique thioether-containing sugar, MTX. (B) Schematic of oxaziridine labeling of MTX on the bacterial cell surface. (C) The three oxaziridine probes used in this study: azidopropyl oxaziridine (AzOx), alkyne oxaziridine (AlkOx), and ethyl oxaziridine (EtOx).

Fig. 2.

Labeling purified LAM from Mtb and M. smegmatis with oxaziridines. (A) Mtb ManLAM is capped with mannose and MTX, while M. smegmatis (Msmeg) PILAM is capped with phosphoinositol residues. (B) Purified LAM from Mtb and Msmeg was incubated with AzOx (100 μM) followed by DBCO-biotin (250 μM) to label purified glycans. (C) Mtb and Msmeg LAM labeled with AzOx was separated by SDS-PAGE and transferred to a PVDF membrane. Glycans were visualized using an anti-LAM antibody or with streptavidin-HRP to detect biotinylation. The blot shown is representative of at least two independent experiments.

Fig. 3.

Labeling live cells with oxaziridines. (A) Representative histograms of flow cytometry of Mtb and M. smegmatis (Msmeg) treated with 100 μM AzOx and 250 μM DBCO-AF647 or DBCO-AF647 alone as a dye only control. Histograms are representative of at least three independent experiments. (B) Mean fluorescence intensity (MFI) of Mtb and M. smegmatis (Msmeg) treated with 100 μM AzOx relative to a dye-only control of two replicate experiments. Error bars represent the SEM for two replicate experiments.

Fig. 4.

Quantified fluorescence of LM/LAM isolated from Mtb treated with oxaziridine probes. (A) Live Mtb was incubated with AzOx followed by DBCO-AF647. Cells were heat-killed to remove from BSL3, lysed, and LM/LAM was purified from labeled cells. (B) Relative fluorescence units (RFU) of LM/LAM from Mtb treated with 100 μM AzOx versus Mtb treated with dye only as measured in solution by plate reader. Error bars represent the SEM of two replicate experiments. Significance was assessed by unpaired t test, *P < 0.05. (C) Purified LM/LAM from Mtb treated with 100 μM AzOx versus Mtb treated with dye only was spotted onto a nitrocellulose membrane (2 μL). Fluorescence of each dot was quantified after background correction as the mean integrated density and plotted in a bar graph. The nitrocellulose membrane was also stained with anti-LAM antibody to demonstrate the presence of LAM in the sample. Error bars represent the SEM of two replicate experiments. Significance was assessed by unpaired t test, *P < 0.05.

Fig. 5.

Visualizing MTX localization in live cells. (A) Microscopy of Mtb and M. smegmatis prelabeled with HADA and treated with or without 100 μM AzOx and with 250 μM DBCO-AF647 reveals distinct staining of Mtb but no labeling of M. smegmatis. (Scale bar, 3 μm.) Images are representative of at least two independent experiments. (B) Structure illumination microscopy (SIM) of Mtb prelabeled with HADA and treated with 100 μM AzOx and 250 μM DBCO-AF488 reveals distinct membrane staining of dividing cells. (Scale bar, 3 μm.) Images are representative of at least two independent experiments. (C) Quantifying cellular fluorescence across the longitudinal axis of the cell. A plot of normalized fluorescence intensity across cell length for one cell shows annular and polar labeling, with one pole significantly brighter than the other (Left). A heat map of fluorescence intensity across cell length for multiple cells organized by length, with shortest cells at the Top ranging to longest cells at the Bottom (Right) (n = 80 cells). The heat map demonstrates predominantly polar labeling of one pole over the other regardless of cell length.

Fig. 6.

Infection of host cells with oxaziridine-labeled Mtb. (A) Monocyte-derived macrophages were infected with Mtb prelabeled with 50 μM RADA, 100 μM AzOx, and 250 μM DBCO-AF647. Macrophages were stained with 2 μg/mL wheat germ agglutinin conjugated to AF405 to mark the macrophage membrane. (B) Macrophages infected with Mtb were fixed and imaged at four hours and 72 h postinfection. (Scale bar, 4 μm.) LAM labeling with oxaziridines persists for up to 72 h. Images are representative of at least two independent experiments.